Magnetic Resonance Imaging - sprawls.org Magnetic Resonance Ima… · Magnetic Resonance Imaging Principles, Methods, and Techniques Perry Sprawls, Ph.D., FACR, FAAPM, FIOMP Distinguished

Magnetic Resonance Imaging

Principles, Methods, and Techniques

Perry Sprawls, Ph.D., FACR, FAAPM, FIOMP

Distinguished Emeritus ProfessorDepartment of Radiology

Emory UniversityAtlanta, Georgia

Medical Physics PublishingMadison, Wisconsin

Copyright 2000 by Perry SprawlsAll rights reserved.

ISBN 13: 0-944838-97-6ISBN 10: 0-944838-97-9

12 10 08 06 5 4 3 2

Special printing for Sprawls Educational Foundation

Library of Congress Cataloging-in-Publication Data

Sprawls, Perry.Magnetic resonance imaging : principles, methods, and techniques / Perry Sprawls.

p. ; cm.Includes index.ISBN 0-944838-97-9 1. Magnetic resonance imaging. I. Title.[DNLM: 1. Magnetic Resonance Imaging--methods. 2. Health Physics. WN 185S767m 2000]RC78.7.N83 S68 2000616.07'548--dc21

00-034862

Perry Sprawls grants permission for photocopying for limited personal use and making slides for educational presentations if the source is fully acknowledged. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. For information, address Perry Sprawls, at [email protected].

Every reasonable effort has been made to give factual and up-to-date information to

the reader of this book. However, because of the possibility of human error and the

potential for change in the medical sciences, the author, publisher, and any other

persons involved in the publication of this book cannot assume responsibility for

the validity of all materials or for the consequences of their use.

Medical Physics Publishing4513 Vernon BoulevardMadison, WI 53705-4964

Printed in the United States of America

Dedication

The power of knowledge is realized when it is shared with others and used to benefit all peoplewith improved health and quality of life.

This book is provided

By the

SPRAWLS EDUCATIONAL FOUNDATIONhttp://www.sprawls.org

In Partnership with Institutions and Faculties Worldwide to Enhance the Learning and Teaching of Medical Imaging Physics, Engineering,

and Technology

Preface ..............................................................................................xvAcknowledgments ...........................................................................xviMind Maps ......................................................................................xvii

Magnetic Resonance Image CharacteristicsIntroduction And Overview..................................................................1The MR Image .....................................................................................2

Tissue Characteristics and Image Types.........................................3Proton Density (PD) Images ..................................................3Magnetic Relaxation Times—T1 and T2 Images....................3

Fluid Movement and Image Types ................................................3Vascular Flow ........................................................................3Perfusion and Diffusion .........................................................3

Spectroscopic and Chemical Shift.................................................3What Do You See In An MR Image?......................................................4

Radio Frequency Signal Intensity ..................................................4Tissue Magnetization....................................................................4Protons (Magnetic Nuclei)............................................................4Hydrogen.....................................................................................5Tissue Characteristics....................................................................5

PD (Proton Density) ..............................................................6T1.........................................................................................6T2.........................................................................................6

Spatial Characteristics ..........................................................................6Slices ............................................................................................6Voxels...........................................................................................6Image Pixels .................................................................................6

Control Of Image Characteristics..........................................................7Contrast Sensitivity.......................................................................7Detail ...........................................................................................8Noise............................................................................................9Artifacts........................................................................................9Spatial ........................................................................................11

Image Acquisition Time......................................................................11Protocol Optimization........................................................................11Mind Map Summary..........................................................................12

Contents

1

vii

Magnetic Resonance Imaging System ComponentsIntroduction And Overview................................................................13

Tissue Magnetization..................................................................13Tissue Resonance........................................................................14

The Magnetic Field ............................................................................14Field Direction ............................................................................14Field Strength.............................................................................15Homogeneity .............................................................................15

Magnets ...........................................................................................15Superconducting .......................................................................15Resistive .....................................................................................16Permanent .................................................................................16

Gradients ...........................................................................................16Gradient Orientation ..................................................................17Gradient Functions .....................................................................17Gradient Strength ......................................................................18Risetime and Slew-Rate...............................................................18Eddy Currents ............................................................................18

Shimming..........................................................................................18Magnetic Field Shielding....................................................................19

Passive Shielding ........................................................................19Active Shielding..........................................................................20

The Radio Frequency System..............................................................20RF Coils ......................................................................................20Transmitter.................................................................................20Receiver......................................................................................21RF Polarization............................................................................21RF Shielding ...............................................................................21

Computer Functions ..........................................................................21Acquisition Control.....................................................................21Image Reconstruction.................................................................22Image Storage and Retrieval .......................................................22Viewing Control and Post Processing ..........................................22

Mind Map Summary..........................................................................23

Nuclear Magnetic ResonanceIntroduction And Overview................................................................25Magnetic Nuclei ................................................................................25

Spins ..........................................................................................26RF Signal Intensity ......................................................................26Relative Signal Strength..............................................................27Tissue Concentration of Elements ...............................................27

viii MAGNETIC RESONANCE IMAGING

2

3

Isotopic Abundance....................................................................28Relative Sensitivity and Signal Strength.......................................28

Radio Frequency Energy.....................................................................28Pulses .........................................................................................28Signals .......................................................................................28

Nuclear Magnetic Interactions ...........................................................28Nuclear Alignment .....................................................................28Precession and Resonance ..........................................................29Excitation ...................................................................................29Relaxation ..................................................................................30

Resonance .........................................................................................30Larmor Frequency ......................................................................31Field Strength.............................................................................31Chemical Shift ............................................................................32

Mind Map Summary..........................................................................33

Tissue Magnetization And RelaxationIntroduction And Overview................................................................35Tissue Magnetization .........................................................................36

Magnetic Direction.....................................................................36Magnetic Flipping ......................................................................36

Flip Angle ...........................................................................36The 90˚ Pulse, Saturation and Excitation .............................36

Longitudinal Magnetization And Relaxation .......................................37T1 Contrast ................................................................................39Molecular Size ............................................................................39Magnetic Field Strength Effect....................................................40

Transverse Magnetization And Relaxation...........................................40T2 Contrast ................................................................................41Proton Dephasing ......................................................................43

T2 Tissue Characteristics .....................................................43T2* Magnetic Field Effects...................................................43

Magnetic Susceptibility ......................................................................44Contrast Agents..........................................................................45Diamagnetic Materials................................................................45Paramagnetic Materials ..............................................................45Superparamagnetic Materials .....................................................46Ferromagnetic Materials .............................................................46

Mind Map Summary..........................................................................47

CONTENTS ix

4

The Imaging ProcessIntroduction And Overview................................................................49

k Space ......................................................................................49Acquisition .................................................................................49Reconstruction ...........................................................................50

Imaging Protocol ...............................................................................50Imaging Methods ..............................................................................50The Imaging Cycle .............................................................................51

TR ..............................................................................................52TE...............................................................................................52Excitation ...................................................................................53The Echo Event and Signals ........................................................53

Contrast Sensitivity ............................................................................53T1 Contrast ................................................................................53Proton Density (PD) Contrast .....................................................55T2 Contrast ................................................................................56

Mind Map Summary..........................................................................58

Spin Echo Imaging MethodsIntroduction And Overview................................................................59The Spin Echo Process........................................................................59

RF Pulse Sequence ......................................................................61The Spin Echo Method.......................................................................61

Proton Density (PD) Contrast .....................................................62T1 Contrast ................................................................................62T2 Contrast ................................................................................63Multiple Spin Echo .....................................................................63

Inversion Recovery .............................................................................64T1 Contrast........................................................................................65Mind Map Summary..........................................................................67

Gradient Echo Imaging MethodsIntroduction And Overview................................................................69The Gradient Echo Process .................................................................69Small Angle Gradient Echo Methods ..................................................71

Excitation/Saturation-Pulse Flip Angle.........................................71Contrast Sensitivity.....................................................................73

T1 Contrast.........................................................................73Low Contrast ......................................................................75Proton Density (PD) Contrast ..............................................75T2 and T2* Contrast ...........................................................75

x MAGNETIC RESONANCE IMAGING

6

7

5

Contrast Enhancement ...............................................................75Mixed Contrast...................................................................76Spoiling and T1 Contrast Enhancement ..............................76

Echo Planar Imaging (EPI) Method .....................................................76Gradient And Spin Echo (GRASE) Method ..........................................77Magnetization Preparation.................................................................78Mind Map Summary..........................................................................80

Selective Signal SuppressionIntroduction And Overview................................................................81T1-Based Fat And Fluid Suppression ...................................................81

STIR Fat Suppression...................................................................82Fluid Suppression .......................................................................83

SPIR Fat Suppression ..........................................................................84Magnetization Transfer Contrast (MTC) .............................................85

Free Proton Pool.........................................................................86Bound Proton Pool .....................................................................86Magnetization Transfer ...............................................................86Selective Saturation ....................................................................86

Regional Saturation............................................................................86Mind Map Summary..........................................................................88

Spatial Characteristics of the Magnetic Resonance ImageIntroduction And Overview................................................................89Signal Acquisition...............................................................................91Image Reconstruction ........................................................................91Image Characteristics .........................................................................91Gradients ...........................................................................................91Slice Selection ....................................................................................91

Selective Excitation.....................................................................92Multi-Slice Imaging ....................................................................92Volume Acquisition.....................................................................93

Frequency Encoding ..........................................................................94Resonant Frequency ...................................................................95

Phase-Encoding .................................................................................96The Gradient Cycle ............................................................................98Image Reconstruction ........................................................................99Mind Map Summary........................................................................101

CONTENTS xi

8

9

Image Detail And NoiseIntroduction And Overview..............................................................103Image Detail ....................................................................................104Noise Sources ..................................................................................106Signal-To-Noise Considerations ........................................................106

Voxel Size .................................................................................106Field Strength...........................................................................106Tissue Characteristics................................................................107TR and TE.................................................................................107RF Coils ....................................................................................108Receiver Bandwidth..................................................................109Averaging.................................................................................109

Mind Map Summary........................................................................111

Acquisition Time And Procedure OptimizationIntroduction And Overview..............................................................113Acquisition Time ..............................................................................113Cycle Repetition Time, TR ................................................................115Matrix Size.......................................................................................115

Reduced Matrix in Phase-Encoded Direction.............................116Rectangular Field of View..........................................................116

Half Acquisition................................................................................116Signal Averaging ..............................................................................117Protocol Factor Interactions..............................................................118Developing An Optimized Protocol ..................................................119

Contrast Sensitivity...................................................................119Image Detail.............................................................................119Spatial Characteristics and Methods .........................................119Image Noise .............................................................................119Artifact Reduction.....................................................................120

Fast Acquisition Methods .................................................................120Mind Map Summary........................................................................122

Vascular ImagingIntroduction And Overview..............................................................123Time Effects .....................................................................................124

Flow-Related Enhancement (Bright Blood)................................124Flow-Void Effect (Black Blood) ..................................................125

Selective Saturation..........................................................................126Phase Effects ....................................................................................126

Intravoxel Phase .......................................................................128Flow Dephasing........................................................................129

xii MAGNETIC RESONANCE IMAGING

11

12

10

Flow Compensation .................................................................129Even-Echo Rephasing................................................................129Intervoxel Phase .......................................................................130

Phase Imaging ..................................................................130Artifacts ............................................................................130

Angiography....................................................................................131Phase Contrast Angiography ....................................................131

Velocity.............................................................................132Flow Direction ..................................................................132

In-Flow Contrast Angiography..................................................133Three-Dimension (3-D) Volume Acquisition.......................133Two-Dimension (2-D) Slice Acquisition .............................134

Image Format...........................................................................134Maximum Intensity Projection (MIP) .................................134Surface Rendering.............................................................135

Contrast Enhanced Angiography ..............................................135Mind Map Summary........................................................................136

Functional ImagingIntroduction And Overview..............................................................137Diffusion Imaging ............................................................................137

Brownian Motion and Diffusion ................................................137Diffusion Coefficient .................................................................138Apparent Diffusion Coefficient (ADC) .......................................138Diffusion Direction ...................................................................139The Imaging Process ................................................................139Diffusion Sensitivity ..................................................................140Diffusion-Weighted Images.......................................................141ADC Map Images .....................................................................141

Blood Oxygenation Level Dependent (BOLD) Contrast ....................141Perfusion Imaging............................................................................142Mind Map Summary........................................................................144

Image ArtifactsIntroduction And Overview..............................................................145Motion-Induced Artifacts .................................................................146

Phase-Encoded Direction..........................................................147Cardiac Motion ........................................................................147

Triggering.........................................................................147Flow Compensation..........................................................147

Respiratory Motion...................................................................147Averaging .........................................................................148

CONTENTS xiii

13

14

Ordered Phase-Encoding ..................................................148Regional Presaturation ......................................................149

Flow.........................................................................................149Regional Presaturation ......................................................149Flow Compensation..........................................................149

Aliasing Artifacts...............................................................................150Chemical-Shift Artifacts ....................................................................150

Field Strength...........................................................................151Bandwidth ...............................................................................151

Mind Map Summary........................................................................153

MRI SafetyIntroduction And Overview..............................................................155Magnetic Fields................................................................................155

Physical Characteristics .............................................................156Biological Effects.......................................................................156Internal Objects........................................................................157External Objects .......................................................................157Magneto-electrical Effects.........................................................157Activation of or Damage to Implanted Devices .........................158

Gradients .........................................................................................158RF Energy.........................................................................................159

Specific Absorption Rate (SAR)..................................................160Magnetic Field Strength....................................................161Pulse Flip Angles ...............................................................161Pulses per Cycle ................................................................161TR.....................................................................................161

Determining the SAR for a Patient ............................................161Surface Burns ...........................................................................161

Acoustic Noise .................................................................................161Mind Map Summary........................................................................162

Index...............................................................................................165

xiv MAGNETIC RESONANCE IMAGING

15

Magnetic resonance imaging (MRI) is a majormedical diagnostic tool. It makes it possible tovisualize and analyze a variety of tissue charac-teristics, blood flow and distribution, and sev-eral physiologic and metabolic functions.Much of this power comes from the ability toadjust the imaging process to be especially sen-sitive to each of the characteristics being eval-uated. Think of it as a multipurpose imagingand analytical procedure that can be config-ured and optimized to provide answers to awide range of clinical questions for virtually allparts and systems of the body.

Each imaging procedure is guided by a pro-tocol consisting of a selection from a choice ofimaging methods, selected values for a largenumber of imaging parameters or factors asso-ciated with the specific method, and the appli-cation of a variety of techniques to optimizeimage quality and acquire the images in theshortest time consistent with other procedurerequirements.

Even though many imaging protocols arepreprogrammed into modern MRI systems,maximum performance and benefit requires ahighly educated and trained staff to conductthe procedures and to interpret the results.

The physicians who are requesting, super-vising, and interpreting the MR examinationsrequire knowledge of MRI principles, methods,and techniques to select appropriate imagingmethods and techniques and to understandthe basis for the clinical information conveyedin the images.

The technologists who perform the exam-inations need a good knowledge and under-standing of the total process so that they canselect and modify protocols as necessary, mon-itor and optimize image quality, and providefor patient and staff safety.

Medical physicists who provide support tothe clinical activities with respect to imagequality and procedure optimization, and con-duct educational activities for other medicalprofessionals must also have a broad knowl-edge at the practical and applied levels.

This book is designed to meet the needs ofall who play a role in the MR imaging process.First, it develops the very important conceptsof the physical principles on which MR imag-ing is based. It then builds an understanding ofthe various methods and techniques that arethe heart of each imaging procedure. It givesspecial emphasis to image quality and the asso-ciated issues of optimizing protocols. Safety con-cerns are addressed in order to have an in-formed staff who can take a realistic approachto reducing risk and increasing patient comfortand acceptance.

The objective of this book is to help all ofus obtain maximum performance and benefitfrom the advanced and sophisticated MR tech-nology that is available today. Humans withthe knowledge of how to apply the variousimaging options to the wide range of clinicalneeds is, and will continue to be, a vital link inthe total MR imaging process.

Preface

xv

Tom Dixon, Ph.D., for years of stimulatingdiscussions on the physics and techniques of MRI,the review of this manuscript, and the many help-ful suggestions that have been incorporated.

Margaret Nix and Tammy Mann, for manuscriptpreparation.Jack Peterson, Ph.D., for technical and editorialcontributions.

xvi

Acknowledgments

xvii

MRI is not a difficult subject to learn. It is, how-ever, a somewhat complex topic because thereare many different parts of the total MRprocess. A useful knowledge of MRI does notconsist of memorized definitions and facts. Itconsists of a visual representation of the MRprocess in our minds. It is this type of knowl-edge that we can use to produce and under-stand images. This book is designed to help youdevelop the concepts of the various aspects ofMRI and to see how these concepts can beapplied to practical imaging situations.

Each chapter is summarized with a mindmap to help organize the many individual con-cepts. A mind map is an effective method forshowing relationships and developing a com-prehensive mental picture of a topic. It is agood way to provide some organization to thecomplexity of a topic such as MRI. Mind mapsare excellent tools for study and review. Thebest mind maps are often the ones that each ofus develops for our own use. You are encour-aged to add to and modify the printed mapsand to develop mind maps of your own.

Mind Maps

Mind Maps

The Physical Worldof

MRI

Imag

es

Equipment

Protocols

MindMaps Easy

Represent

He lpShow

“The Big Picture”

Illustrate

Relationships

ToRemember

TheReal World

BetterThan

WORDS

WORDS

1Magnetic ResonanceImage Characteristics

1

Introduction And OverviewMagnetic resonance imaging (MRI) is a med-ical imaging process that uses a magnetic fieldand radio frequency (RF) signals to produceimages of anatomical structures, of the pres-ence of disease, and of various biological func-tions within the human body. MRI producesimages that are distinctly different from theimages produced by other imaging modalities.A primary difference is that the MRI processcan selectively image several different tissuecharacteristics. A potential advantage of thisis that if a pathologic process does not alterone tissue characteristic and produce contrast,

it might be visible in an image because of itseffect on other characteristics. This causes theMRI process to be somewhat more complexthan most imaging methods. In order to opti-mize an MRI procedure for a specific clinicalexamination, the user must have a goodknowledge of the characteristics of the mag-netic resonance (MR) image and how thosecharacteristics can be controlled.

In this chapter we will develop a basicknowledge and overview of the MR image,how the image relates to specific tissue char-acteristics, and how image quality character-istics can be controlled.

The MR ImageThe MR image displays certain physical char-acteristics of tissue. Let us now use Figure 1-1to identify these characteristics and to seehow they are related.

The MR image is a display of RF signalsthat are emitted by the tissue during theimage acquisition process. The source of thesignals is a condition of magnetization that isproduced in the tissue when the patient isplaced in the strong magnetic field. The tissuemagnetization depends on the presence ofmagnetic nuclei. The specific physical charac-teristic of tissue or fluid that is visible in theimage depends on how the magnetic field is

being changed during the acquisition process.An image acquisition consists of an acquisi-tion cycle, like a heartbeat, that is repeatedmany times. During each cycle the tissue mag-netization is forced through a series ofchanges. As we will soon learn in much moredetail, all tissues and fluids do not progressthrough these changes at the same rate. It isthe level of magnetization that is present at aspecial “picture snapping time” at the end ofeach cycle that determines the intensity of theRF signal produced and the resulting tissuebrightness in the image.

MR images are generally identified with spe-cific tissue characteristics or blood conditions

2 MAGNETIC RESONANCE IMAGING

MAGNETIC RESONANCE IMAGE TYPES

Magnetized Tissue Image

Radio Frequency Signals

TISSUE CHARACTERISTICSPD - Proton DensityT1 - Longitudinal Relaxation TimeT2 - Transverse Relaxation Time

FLUID MOVEMENTVascular Flow (Angiography)PerfusionDiffusion

CHEMICAL SPECTROSCOPY

Figure 1-1. Physical characteristics of tissue and fluid movement that can be displayed in the magnetic resonance image. MRI can also provide certain chemical information

by applying spectroscopy analysis to the RF signals emitted by the tissue.

that are the predominant source of contrast.These characteristics determine the level of tis-sue magnetization and contrast present at thetime the “picture is snapped.” The equipmentoperator, who sets the imaging protocol, deter-mines the type of image that is to be producedby adjusting various imaging factors.

The characteristics that can be used as asource of image contrast fall into three ratherdistinct categories. The first, and most widelyused, category is the magnetic characteristicsof tissues. The second category is characteris-tics of fluid (usually blood) movement. Thethird category is the spectroscopic effectsrelated to molecular structure.

At this time we will briefly introduce each ofthese characteristics to set the stage for the muchmore detailed descriptions presented later.

Tissue Characteristics and Image Types

Proton Density (PD) Images

The most direct tissue characteristic that canbe imaged is the concentration or density ofprotons (hydrogen). In a proton density imagethe tissue magnetization, RF signal intensity,and image brightness are determined by theproton (hydrogen) content of the tissue. Tis-sues that are rich in protons will producestrong signals and have a bright appearance.

Magnetic Relaxation Times —T1 and T2 Images

During an MRI procedure the tissue magneti-zation is cycled by flipping it into an unstablecondition and then allowing it to recover. Thisrecovery process is known as relaxation. Thetime required for the magnetization to relaxvaries from one type of tissue to another. Therelaxation times can be used to distinguish

(i.e., produce contrast) among normal andpathologic tissues.

Each tissue is characterized by two relax-ation times: Tl and T2. Images can be createdin which either one of these two characteris-tics is the predominant source of contrast. It isusually not possible to create images in whichone of the tissue characteristics (e.g., PD, T1, orT2) is the only pure source of contrast. Typi-cally, there is a mixing or blending of the char-acteristics but an image will be more heavilyweighted by one of them. When an image isdescribed as a T1-weighted image, this meansthat T1 is the predominant source of contrastbut there is also some possible contaminationfrom the PD and T2 characteristics.

Fluid Movement and Image TypesVascular Flow

The MRI process is capable of producingimages of flowing blood without the use ofcontrast media. Although flow effects are oftenvisible in all types of images, it becomes thepredominant source of contrast in images pro-duced specifically for vascular or angiographicexaminations as described in Chapter 12.

Perfusion and Diffusion

It is possible to produce images that showboth perfusion and diffusion within tissue.These require specific imaging methods andare often characterized as functional imaging.

Spectroscopic and Chemical ShiftThe frequency of the RF signals emitted bytissue is affected to a small degree by the sizeand characteristics of the molecules contain-ing the magnetic nuclei. These differences infrequencies, the chemical shift, can be dis-played in images. It is also the basis of MRspectroscopy. Spectroscopy is the process of

MAGNETIC RESONANCE IMAGE CHARACTERISTICS 3

using magnetic resonance to analyze thechemical composition of tissue. Spectroscopymakes use of the fact that different molecularstructures have different resonant frequen-cies. Typically, the MR signals from a tissuespecimen are sorted and displayed on a fre-quency scale. The signals from differentchemical compounds will appear as peaksalong the frequency scale. This leads to theiridentity and measure of relative abundance.

What Do You See In An MR Image? We have discovered that an MR image can dis-play a variety of tissue and body fluid charac-teristics. However, there are several physicalcharacteristics that form the link between theimage and the tissue characteristics describedabove. Understanding this link gives us a bet-ter appreciation of how the tissue characteris-tics are made visible. We will use Figure 1-2 todevelop the link.

Radio Frequency Signal Intensity

The first thing we see in an image is RF signalintensity emitted by the tissues. Bright areasin the image correspond to tissues that emithigh signal intensity. There are also areas in animage that appear as dark voids because nosignals are produced. Between these twoextremes there will be a range of signal inten-sities and shades of gray that show contrast ordifferences among the various tissues.

Let us now move deeper into the imagingprocess and discover the relationship betweenRF signal intensity and other characteristics.

Tissue Magnetization

The condition within the tissue that producesthe RF signal is magnetization. At this point we will use an analogy to radioactive nuclide

imaging. In nuclear medicine procedures it isthe presence of radioactivity in the tissues thatproduces the radiation. In MRI it is the mag-netization within the tissues that produces theRF signal radiation displayed in the image.Therefore, when we look at an MR image, weare seeing a display of magnetized tissue.

We will soon discover that tissue becomesmagnetized when the patient is placed in astrong magnetic field. However, all tissues arenot magnetized to the same level. During theimaging process the tissue magnetization iscycled through a series of changes, but all tissuesdo not change at the same rate. It is this differ-ence in rates of change of the magnetization thatmakes the tissues different and produces muchof the useful contrast. This will be described inmuch more detail later when we will learn thatthese rates of change are described as magneticrelaxation times, T1 and T2.

It is the level of magnetization at specific“picture snapping” times during the imagingprocedure that determines the intensity of theresulting RF signal and image brightness. TheMR image is indeed an image of magnetizedtissue. Tissues or other materials that are notadequately magnetized during the imagingprocedure will not be visible in the image.

Protons (Magnetic Nuclei)The next thing we see is an image of protonsthat are the nuclei of hydrogen atoms. That iswhy an MRI procedure is often referred to asproton imaging.

The magnetization of tissue, which pro-duces the RF signals, comes from protons thatare actually small magnets (magnetic nuclei)present in the tissue. These small magnets areactually the nuclei of certain atoms that have aspecial magnetic property called a magneticmoment. Not all chemical substances have anadequate abundance of magnetic nuclei.


HydrogenThe only substance found in tissue that has anadequate concentration of magnetic nuclei toproduce good images is hydrogen. The nucleusof a hydrogen atom is a single proton. Therefore,the MR image is an image of hydrogen. Whentissue that contains hydrogen (small magneticnuclei), i.e., protons, is placed in a strong mag-netic field, some of the protons line up in thesame direction as the magnetic field. Thisalignment produces the magnetization in the

tissue, which then produces the RF signal. If atissue does not have an adequate concentra-tion of molecules containing hydrogen, it willnot be visible in an MR image.

Tissue Characteristics

As we have moved deeper into the imagingprocess we arrive again at the three tissuecharacteristics: PD, T1, and T2. It is thesecharacteristics that we want to see becausethey give us valuable information about the


WHAT DO YOU SEE IN AN MR IMAGE?

ImageBrightness

Tissue Voxels

Radio FrequencySignal Intensity

MagnetizedTissue Protons

MagneticMoment

MagnetizationVector

Proton Density, PDRelaxation Time, T1Relaxation Time, T2

LowLongShort

HighShortLong

Tissue Characteristics:

SPIN

Figure 1-2. The physical characteristics that form the link between the image and the three tissue characteristics.

tissues. These characteristics become visiblebecause each one has an effect on the level ofmagnetization that is present at the picturesnapping time in each imaging cycle. At thistime we will briefly describe the effect of eachand then develop the process in more detailin Chapters 4 and 5.

PD (Proton Density)

PD has a very direct effect on tissue magneti-zation and the resulting RF signal and imagebrightness. That is because the magnetizationis produced by the protons. Therefore, a tissuewith a high PD can reach a high level of mag-netization and produce an intense signal.

T1

When the imaging protocol is set to produce aT1-weighted image, it is the tissues with theshort T1 values that produce the highest mag-netization and are the brightness in the image.

T2

When the imaging protocol is set to producea T2-weighted image, it is the tissues with thelong T2 values that are the brightest. This isbecause they have a higher level of magneti-zation at the picture snapping time.

Spatial CharacteristicsFigure 1-3 illustrates the basic spatial charac-teristics of the MR image. MRI is basically a tomographic imaging process, althoughthere are some procedures, such as angiogra-phy, in which a complete anatomical volumewill be displayed in a single image. The pro-tocol for the acquisition process must be setup to produce the appropriate spatial charac-teristics for a specific clinical procedure. Thisincludes such factors as the number of slices,

slice orientation, and the structure withineach individual slice.

Slices

A typical examination will consist of at least oneset of contiguous slices. In most cases the entireset of slices is acquired simultaneously. How-ever, the number of slices in a set can be limitedby certain imaging factors and the amount oftime allocated to the acquisition process.

The slices can be oriented in virtually anyplane through the patient’s body. The majorrestriction is that images in the different planescannot generally be acquired simultaneously.For example, if both axial and sagittal imagesare required, the acquisition process must berepeated. However, there is the possibility ofacquiring 3-D data from a large volume of tis-sue and then reconstructing slices in the differ-ent planes, as will be described in Chapter 9.

Voxels

Each slice of tissue is subdivided into rows andcolumns of individual volume elements, orvoxels. The size of a voxel has a significanteffect on image quality. It is controlled by acombination of protocol factors as described inChapter 10 and should be adjusted to an opti-mum size for each type of clinical examina-tion. Each voxel is an independent source ofRF signals. That is why voxel size is a majorconsideration in each image acquisition.

Image Pixels

The image is also divided into rows andcolumns of picture elements, or pixels. In gen-eral, an image pixel represents a correspondingvoxel of tissue within the slice. The brightnessof an image pixel is determined by the intensityof the RF signal emitted by the tissue voxel.


Control Of Image CharacteristicsThe operator of an MRI system has tremendouscontrol over the characteristics and the qualityof the images that are produced. The five basicimage quality characteristics are represented inFigure 1-4. Each of these image characteristicsis affected by a combination of the imagingfactors that make up the acquisition protocol.

Not all types of clinical procedures requireimages with the same characteristics. There-fore, the primary objective is to use an imaging

protocol in which the acquisition process isoptimized for a specific clinical requirement.

Although each of the image characteristicswill be considered in detail in later chapters,we will introduce them here.

Contrast Sensitivity

Contrast sensitivity is the ability of an imag-ing process to produce an image of objects ortissues in the body that have relatively smallphysical differences or inherent contrast. Thecontrast that is to be imaged is in the form of


SPATIAL CHARACTERISTICS

Volume ofSlices

Image ofPixels

PixelSignal

Voxel

Slice ofVoxels

Figure 1-3. The spatial characteristics of MR images.

some specific physical characteristic. In x-rayimaging, including CT (computed tomogra-phy), difference in physical density is a prin-ciple source of contrast. One of the majoradvantages of MRI is that it has a high con-trast sensitivity for visualizing differencesamong the tissues in the body because thereare several sources of contrast; that is, it hasthe ability to image a variety of characteristics(PD, T1, T2) as described previously. Also,there is usually much greater variation amongthese characteristics than among the tissuedensity values that are the source of contrastfor x-ray imaging. If a certain pathologic con-dition does not produce a visible change inone characteristic, there is the possibility thatit will be visible by imaging some of the othercharacteristics.

Even though MRI has high contrast sensi-tivity relative to most of the other imagingmodalities, it must be optimized for each clin-ical procedure. This includes the selection ofthe characteristics, or sources of contrast, thatare to be imaged and then adjusting the proto-col factors so that the sensitivity to that specificcharacteristic is optimized. This is illustrated inFigure 1-5.

Detail A distinguishing characteristic of every imag-ing modality is its ability to image small objectsand structures within the body. Visibility ofanatomical detail (sometimes referred to as spa-tial resolution) is limited by the blurring thatoccurs during the imaging process. All medicalimaging methods produce images with some


SPATIAL

SPATIALARTIFA CTS

MatrixTR TE

MAGNETIC RESONANCE IMAGEQUALITY CHARACTERISTICS

Factors

Imaging Protocol

Figure 1-4. Image quality characteristics that can be controlled by the selection of protocol factors.

blurring but not to the same extent. The blur-ring in MRI is greater than in radiography.Therefore, MRI cannot image small structuresthat are visible in conventional radiographs.

In MRI, like all modalities, the amount ofblurring and the resulting visibility of detailcan be adjusted during the imaging process.Figure 1-6 shows images with different levels ofblurring and visibility of detail. The protocolfactors that are used to adjust detail and theassociated issues in their optimization will bediscussed in Chapter 10.

Noise

Visual noise is a major issue in MRI. The pres-ence of noise in an image reduces its quality,especially by limiting the visibility of low con-trast objects and differences among tissues.Figure 1-7 shows images with different levels of

visual noise. Most of the noise in MR imagesis the result of a form of random, unwantedRF energy picked up from the patient’s body.

The amount of noise can generally be con-trolled through a combination of factors asdescribed in Chapter 10. However, many ofthese factors involve compromises with othercharacteristics.

Artifacts

Artifacts are undesirable objects, such as streaksand spots, that appear in images which do notdirectly represent an anatomical structure.They are usually produced by certain interac-tions of the patient’s body or body functions(such as motion) with the imaging process.

There is a selection of techniques that canbe used to reduce the presence of artifacts. Thesewill be described in Chapter 14.


T1

ProtonDensity

(PD) T2

OPTIMIZED FOR:

Figure 1-5. The images produced when the contrast sensitivity is optimized for each of the three specific tissue characteristics.


Visibility of Detail

High

HighBlur

Low

Low

Figure 1-6. Images with different levels of blurring and visibility of anatomical detail.

High Noise Low Noise

Figure 1-7. Images with different levels of visual noise.

Spatial

The general spatial characteristics of the MRimage were described previously. However,when setting up an imaging protocol the spatialcharacteristics must be considered in the gen-eral context of image quality. As we will discoverlater, voxel size plays a major role in determin-ing both image detail and image noise.

Image Acquisition TimeWhen considering and adjusting MR imagequality, attention must also be given to thetime required for the acquisition process. Ingeneral, several aspects of image quality, suchas detail and noise, can be improved by usinglonger acquisition times.

Protocol OptimizationAn optimum imaging protocol is one inwhich there is a proper balance among theimage quality characteristics described aboveand also a balance between overall imagequality and acquisition time.

The imaging protocol that is used for aspecific clinical examination has a majorimpact on the quality of the image and thevisibility of anatomical structures and patho-logic conditions.

Therefore, the users of MRI must have agood knowledge of the imaging process andthe protocol factors and know how to setthem to optimize the image characteristics.

The overall process of optimizing proto-cols will be described in Chapter 11.


The magnetic resonance image is a display of radio frequency signal intensities that are emittedby magnetized tissue during the imaging process. The tissue becomes magnetized because it con-tains protons that are the magnetic nuclei of hydrogen atoms. When placed in the strong mag-netic field, some of the protons align with the field producing the tissue magnetization. The levelof magnetization at the time during the procedure when the “picture is snapped” is determined bya variety of tissue and fluid movement characteristics. By adjusting the imaging process it is pos-sible to produce images in which these various characteristics are the principal sources of contrast.

An advantage of MRI is the ability to selectively image a variety of tissue and fluid character-istics. If a specific pathologic condition is not visible when viewing one characteristic, there is thepossibility of seeing it by imaging some of the other characteristics.

During the imaging procedure a section of the patient’s body is divided first into slices, andthe slices are divided into a matrix of voxels. Each voxel is an independent RF signal source. Voxelsize can be adjusted and is what determines image detail and also affects image noise.

The five major image quality characteristics—contrast sensitivity, detail, noise, artifacts, andspatial—can be controlled to a great extent by the settings of the various protocol factors.

MRI is a powerful diagnostic tool because the process can be optimized to display a wide rangeof clinical conditions. However, maximum benefit requires a staff with the knowledge to controlthe process and interpret the variety of images.


SPATIALSPATIAL

SPATIAL

Slice of Voxels

Protons

Magnetic Moment Size

Determined By Affected By

Magnetized TissueDetermined

By

TissueVoxel

Display ofRadio FrequencySignal Intensity

ImageMatrix of Pixels

Can Be Adjusted to Visualize

TISSUE CHARACTERISTICSPD - Proton DensityT1 - Longitudinal Relaxation TimeT2 - Transverse Relaxation Time

FLUID MOVEMENTVascular Flow (Angiography)PerfusionDiffusion

CHEMICAL SPECTROSCOPY

TE MatrixTR

Control The Image Characteristics

Protocol Factors

ARTIFACTS

SPIN

Mind Map SummaryMagnetic Resonance Image Characteristics

2Magnetic Resonance

Imaging SystemComponents

13

Introduction And OverviewThe MRI system consists of several majorcomponents, as shown in Figure 2-1. At thistime we will introduce the components andindicate how they work together to create theMR image. The more specific details of theimage forming process will be explained inlater chapters.

The heart of the MRI system is a largemagnet that produces a very strong magneticfield. The patient’s body is placed in the mag-netic field during the imaging procedure. Themagnetic field produces two distinct effectsthat work together to create the image.

Tissue Magnetization

When the patient is placed in the magneticfield, the tissue becomes temporarily magnet-ized because of the alignment of the protons,as described previously. This is a very low-leveleffect that disappears when the patient isremoved from the magnetic field. The abilityof MRI to distinguish between different typesof tissue is based on the fact that different tis-sues, both normal and pathologic, willbecome magnetized to different levels or willchange their levels of magnetization (i.e.,relax) at different rates.

Tissue Resonance

The magnetic field also causes the tissue to“tune in” or resonate at a very specific radiofrequency. That is why the procedure is knownas magnetic resonance imaging. It is actually cer-tain nuclei, typically protons, within the tissuethat resonate. Therefore, the more compre-hensive name for the phenomenon that is thebasis of both imaging and spectroscopy isnuclear magnetic resonance (NMR).

In the presence of the strong magnetic fieldthe tissue resonates in the RF range. This causesthe tissue to function as a tuned radio receiverand transmitter during the imaging process.

The production of an MR image involves two-way radio communication between the tissuein the patient’s body and the equipment.

The Magnetic FieldFigure 2-2 shows the general characteristics ofa typical magnetic field. At any point within amagnetic field, the two primary characteris-tics are field direction and field strength.

Field Direction

It will be easier to visualize a magnetic field ifit is represented by a series of parallel lines, asshown in Figure 2-2. The arrow on each line


RadiofrequencyTransmitter

Radio FrequencyCoil

Magnetic CoilGradient Coils

GradientPowerSupply

Computer

Protocols

THE MRI SYSTEM

Image ReconstructionViewing Control

Processing

Operator Keyboard

Radio FrequencyReceiver

Figure 2-1. The major components of the Magnetic Resonance Imaging System.

indicates the direction of the field. On the sur-face of the earth, the direction of the earth’smagnetic field is specified with reference to thenorth and south poles. The north-south desig-nation is generally not applied to magneticfields used for imaging. Most of the electro-magnets used for imaging produce a magneticfield that runs through the bore of the magnetand parallel to the major patient axis. As themagnetic field leaves the bore, it spreads outand encircles the magnet, creating an externalfringe field. The external field can be a sourceof interference with other devices and is usu-ally contained by some form of shielding.

Field Strength

Each point within a magnetic field has a par-ticular intensity, or strength. Field strength isexpressed either in the units of tesla (T) or gauss(G). The relationship between the two units isthat 1.0 T is equal to 10,000 G or 10 kG. At theearth’s surface, the magnetic field is relativelyweak and has a strength of less than 1 G. Mag-netic field strengths in the range of 0.15 T to1.5 T are used for imaging. The significance of

field strength is considered as we explore thecharacteristics of MR images and image qualityin later chapters.

Homogeneity

MRI requires a magnetic field that is very uni-form, or homogeneous with respect to strength.Field homogeneity is affected by magnetdesign, adjustments, and environmental con-ditions. Imaging generally requires a homo-geneity (field uniformity) on the order of a few parts per million (ppm) within the imag-ing area.

High homogeneity is obtained by the pro-cess of shimming, as described later.

MagnetsThere are several different types of magnetsthat can be used to produce the magnetic field.Each has its advantages and disadvantages.

Superconducting

Most MRI systems use superconducting mag-nets. The primary advantage is that a super-conducting magnet is capable of producing amuch stronger and stable magnetic field thanthe other two types (resistive and permanent)considered below. A superconducting mag-netic is an electromagnet that operates in asuperconducting state. A superconductor is anelectrical conductor (wire) that has no resist-ance to the flow of an electrical current. Thismeans that very small superconducting wirescan carry very large currents without over-heating, which is typical of more conventionalconductors like copper. It is the combined abil-ity to construct a magnet with many loops orturns of small wire and then use large currentsthat makes the strong magnetic fields possible.

MAGNETIC RESONANCE IMAGING SYSTEM COMPONENTS 15

1.5 tesla (15,000 gauss)

5 gauss

Figure 2-2. The magnetic field produced bysuperconducting magnets.

THE MAGNETIC FIELD

There are two requirements for supercon-ductivity. The conductor or wire must be fabri-cated from a special alloy and then cooled to avery low temperature. The typical magnet con-sists of small niobium-titanium (Nb-Ti) wiresimbedded in copper. The copper has electricalresistance and actually functions as an insula-tor around the Nb-Ti superconductors.

During normal operation the electrical cur-rent flows through the superconductor with-out dissipating any energy or producing heat.If the temperature of the conductor shouldever rise above the critical superconductingtemperature, the current begins to produceheat and the current is rapidly reduced. Thisresults in the collapse of the magnetic field.This is an undesirable event known as a quench.More details are given in Chapter 15 on safety.Superconducting magnets are cooled with liq-uid helium. A disadvantage of this magnettechnology is that the coolant must be replen-ished periodically.

A characteristic of most superconductingmagnets is that they are in the form of cylin-drical or solenoid coils with the strong field inthe internal bore. A potential problem is thatthe relatively small diameter and the longbore produce claustrophobia in some patients.Superconducting magnetic design is evolvingto more open patient environments to reducethis concern.

Resistive

A resistive type magnet is made from a con-ventional electrical conductor such as copper.The name “resistive” refers to the inherentelectrical resistance that is present in all mate-rials except for superconductors. When a cur-rent is passed through a resistive conductor toproduce a magnetic field, heat is also pro-duced. This limits this type of magnet to rela-tively low field strengths.

Permanent

It is possible to do MRI with a non-electricalpermanent magnet. An obvious advantage isthat a permanent magnet does not requireeither electrical power or coolants for opera-tion. However, this type of magnet is also lim-ited to relatively low field strengths.

Both resistive and permanent magnets areusually designed to produce vertical magneticfields that run between the two magneticpoles, as shown in Figure 2-3. Possible advan-tages include a more open patient environ-ment and less external field than supercon-ducting magnets.

GradientsWhen the MRI system is in a resting state andnot actually producing an image, the mag-netic field is quite uniform or homogeneousover the region of the patient’s body. However,during the imaging process the field must bedistorted with gradients. A gradient is just a


PERMANENT OR RESISTIVE MAGNETS

The Magnetic Field

Arrows Indicate Field Direction

Figure 2-3. The magnetic field produced bytypical resistive or permanent magnets.

change in field strength from one point toanother in the patient’s body. The gradientsare produced by a set of gradient coils, whichare contained within the magnet assembly.During an imaging procedure the gradientsare turned on and off many times. This actionproduces the sound or noise that comes fromthe magnet.

The effect of a gradient is illustrated inFigure 2-4. When a magnet is in a “restingstate,” it produces a magnetic field that is uni-form or homogenous over most of thepatient’s body. In this condition there are nogradients in the field. However, when a gradi-ent coil is turned on by applying an electric cur-rent, a gradient or variation in field strength isproduced in the magnetic field.

Gradient Orientation

The typical imaging magnet contains threeseparate sets of gradient coils. These are ori-ented so that gradients can be produced in thethree orthogonal directions (often designatedas the x, y, and z directions). Also, two or moreof the gradient coils can be used together toproduce a gradient in any desired direction.

Gradient Functions

The gradients are used to perform many dif-ferent functions during the image acquisitionprocess. It is the gradients that create the spa-tial characteristics by producing the slices andvoxels that will be described in Chapter 9.The entire family of gradient echo imaging


A MAGNETIC FIELD GRADIENTGradient Coils

Field Strength

Uniform field strength

(Resting stage)

Gradient Coils

OFF

ON

Figure 2-4. A magnetic field gradient produced by a current in the gradient coil.

methods uses a gradient to produce the echoevent and signal which will be described inChapter 7. Gradients are also used to produceone type of image contrast (phase contrastangiography) for vascular imaging, as will bedescribed in Chapter 12, and in the func-tional imaging methods described in Chapter13. Gradients also are used as part of some ofthe techniques to reduce image artifacts, aswill be described in Chapter 14.

Gradient Strength

The strength of a gradient is expressed interms of the change in field strength per unitof distance. The typical units are millitesla permeter (mT/m). The maximum gradientstrength that can be produced is a design char-acteristic of a specific imaging system. Highgradient strengths of 20 mT/m or more arerequired for the optimum performance ofsome imaging methods.

Risetime and Slew-Rate

For certain functions it is necessary for the gra-dient to be capable of changing rapidly. Therisetime is the time required for a gradient toreach its maximum strength. The slew-rate isthe rate at which the gradient changes withtime. For example, a specific gradient systemmight have a risetime of 0.20 milliseconds(msec) and a slew-rate of 100 mT/m/msec.

Eddy Currents

Eddy currents are electrical currents that areinduced or generated in metal structures orconducting materials that are within a chang-ing magnetic field. Since gradients are strong,rapidly changing magnetic fields, they arecapable of producing undesirable eddy cur-rents in some of the metal components of themagnet assembly. This is undesirable because

the eddy currents create their own magneticfields that interfere with the imaging process.

Gradients are designed to minimize eddycurrents either with special gradient shieldingor electrical circuits that control the gradientcurrents in a way that compensates for theeddy-current effects.

ShimmingOne of the requirements for good imaging is ahomogeneous magnet field. This is a field inwhich there is a uniform field strength over theimage area. Shimming is the process of adjust-ing the magnetic field to make it more uniform.

Inhomogeneities are usually produced bymagnetically susceptible materials located inthe magnetic field. The presence of thesematerials produces distortions in the magneticfield that are in the form of inhomogeneities.This can occur in both the internal and exter-nal areas of the field. Each time a differentpatient is placed in the magnetic field, someinhomogeneities are produced. There aremany things in the external field, such asbuilding structures and equipment, that canproduce inhomogeneities. The problem is thatwhen the external field is distorted, these dis-tortions are also transferred to the internalfield where they interfere with the imagingprocess. Inhomogeneities produce a variety ofproblems that will be discussed later.

It is not possible to eliminate all of thesources of inhomogeneities. Therefore, shim-ming must be used to reduce the inhomo-geneities. This is done in several ways. Whena magnet is manufactured and installed, someshimming might be done by placing metalshims in appropriate locations. Magnets alsocontain a set of shim coils. Shimming is pro-duced by adjusting the electrical currents inthese coils. General shimming is done by the


engineers when a magnet is installed or serv-iced. Additional shimming is done for indi-vidual patients. This is often done automati-cally by the system.

Magnetic Field ShieldingThe external magnetic field surrounding themagnet is the possible source of two types ofproblems. One problem is that the field is sub-ject to distortions by metal objects (buildingstructures, vehicles, etc.) as described previ-ously. These distortions produce inhomo-geneities in the internal field. The secondproblem is that the field can interfere withmany types of electronic equipment such asimaging equipment and computers.

It is a common practice to reduce the sizeof the external field by installing shielding asshown in Figure 2-5. The principle of magneticfield shielding is to provide a more attractivereturn path for the external field as it passesfrom one end of the magnetic field to the other.This is possible because air is not a good mag-netic field conductor and can be replaced bymore conductive materials, such as iron. Thereare two types of shielding: passive and active.

Passive Shielding

Passive shielding is produced by surround-ing the magnet with a structure consisting ofrelatively large pieces of ferromagnetic mate-rials such as iron. The principle is that the


IRON

IRON

Figure 2-5. The principle of magnetic field shielding.

MAGNETIC FIELD SHIELDING

Unshielded Field

Unshielded Field

5 gauss

ferromagnetic materials are a more attractivepath for the magnetic field than the air. Ratherthan expanding out from the magnet, the mag-netic field is concentrated through the shieldingmaterial located near the magnet as shown inFigure 2-5. This reduces the size of the field.

Active ShieldingActive shielding is produced by additional coilsbuilt into the magnet assembly. They are de-signed and oriented so that the electrical cur-rents in the coils produce magnetic fields thatoppose and reduce the external magnetic field.

The Radio Frequency SystemThe radio frequency (RF) system provides thecommunications link with the patient’s bodyfor the purpose of producing an image. Allmedical imaging modalities use some form ofradiation (e.g., x-ray, gamma-ray, etc.) or energy(e.g., ultrasound) to transfer the image from thepatient’s body.

The MRI process uses RF signals to trans-mit the image from the patient’s body. The RFenergy used is a form of non-ionizing radia-tion. The RF pulses that are applied to thepatient’s body are absorbed by the tissue andconverted to heat. A small amount of theenergy is emitted by the body as signals usedto produce an image. Actually, the image itselfis not formed within and transmitted fromthe body. The RF signals provide information(data) from which the image is reconstructedby the computer. However, the resulting imageis a display of RF signal intensities produced bythe different tissues.

RF CoilsThe RF coils are located within the magnetassembly and relatively close to the patient’sbody. These coils function as the antennae forboth transmitting signals to and receiving

signals from the tissue. There are different coildesigns for different anatomical regions(shown in Figure 2-6). The three basic typesare body, head, and surface coils. The factorsleading to the selection of a specific coil willbe considered in Chapter 10. In some applica-tions the same coil is used for both transmit-ting and receiving; at other times, separatetransmitting and receiving coils are used.

Surface coils are used to receive signalsfrom a relatively small anatomical region toproduce better image quality than is possiblewith the body and head coils. Surface coils canbe in the form of single coils or an array of sev-eral coils, each with its own receiver circuitoperated in a phased array configuration. Thisconfiguration produces the high image qualityobtained from small coils but with the addedadvantage of covering a larger anatomicalregion and faster imaging.

Transmitter

The RF transmitter generates the RF energy,which is applied to the coils and then trans-mitted to the patient’s body. The energy isgenerated as a series of discrete RF pulses. As


THE RF COILS

Body

Signals SurfaceHead

Pulses

Figure 2-6. The three types of RF coils (body, head, and surface) that are theantennae for transmitting pulses and

receiving signals from the patient's body.

we will see in Chapters 6, 7, and 8, the char-acteristics of an image are determined by thespecific sequence of RF pulses.

The transmitter actually consists of severalcomponents, such as RF modulators and poweramplifiers, but for our purposes here we will con-sider it as a unit that produces pulses of RF energy.The transmitters must be capable of producingrelatively high power outputs on the order ofseveral thousand watts. The actual RF powerrequired is determined by the strength of themagnetic field. It is actually proportional to thesquare of the field strength. Therefore, a 1.5 T sys-tem might require about nine times more RFpower applied to the patient than a 0.5 T system.One important component of the transmitter isa power monitoring circuit. That is a safety fea-ture to prevent excessive power being applied tothe patient’s body, as described in Chapter 15.

ReceiverA short time after a sequence of RF pulses istransmitted to the patient’s body, the resonat-ing tissue will respond by returning an RF signal.These signals are picked up by the coils andprocessed by the receiver. The signals are con-verted into a digital form and transferred to thecomputer where they are temporarily stored.

RF PolarizationThe RF system can operate either in a linear or acircularly polarized mode. In the circularly polar-ized mode, quadrature coils are used. Quadraturecoils consist of two coils with a 90˚ separation.This produces both improved excitation effi-ciency by producing the same effect with half ofthe RF energy (heating) to the patient, and a bet-ter signal-to-noise ratio for the received signals.

RF ShieldingRF energy that might be in the environmentcould be picked up by the receiver and interfere

with the production of high quality images.There are many sources of stray RF energy,such as fluorescent lights, electric motors,medical equipment, and radio communica-tions devices. The area, or room, in which thepatient’s body is located must be shieldedagainst this interference.

An area can be shielded against external RFsignals by surrounding it with an electricallyconducted enclosure. Sheet metal and copperscreen wire are quite effective for this purpose.

The principle of RF shielding is that RF sig-nals cannot enter an electrically conductiveenclosure. The thickness of the shielding isnot a factor—even thin foil is a good shield.The important thing is that the room must becompletely enclosed by the shielding materialwithout any holes. The doors into imagingrooms are part of the shielding and should beclosed during image acquisition.

Computer FunctionsA digital computer is an integral part of an MRIsystem. The production and display of an MRimage is a sequence of several specific steps thatare controlled and performed by the computer.

Acquisition Control

The first step is the acquisition of the RF sig-nals from the patient’s body. This acquisitionprocess consists of many repetitions of animaging cycle. During each cycle a sequenceof RF pulses is transmitted to the body, thegradients are activated, and RF signals are col-lected. Unfortunately, one imaging cycle doesnot produce enough signal data to create animage. Therefore, the imaging cycle must berepeated many times to form an image. Thetime required to acquire images is determinedby the duration of the imaging cycle or cyclerepetition time—an adjustable factor known


as TR—and the number of cycles. The numberof cycles used is related to image quality. Morecycles generally produce better images. Thiswill be described in much more detail inChapters 10 and 11.

Protocols stored in the computer controlthe acquisition process. The operator canselect from many preset protocols for specificclinical procedures or change protocol factorsfor special applications.

Image ReconstructionThe RF signal data collected during the acqui-sition phase is not in the form of an image.However, the computer can use the collecteddata to create or “reconstruct” an image. Thisis a mathematical process known as a Fouriertransformation that is relatively fast and usu-ally does not have a significant effect on totalimaging time.

Image Storage and RetrievalThe reconstructed images are stored in the com-puter where they are available for additional

processing and viewing. The number of imagesthat can be stored—and available for immedi-ate display—depends on the capacity of thestorage media.

Viewing Control and Post Processing

The computer is the system component thatcontrols the display of the images. It makes itpossible for the user to select specific imagesand control viewing factors such as window-ing (contrast) and zooming (magnification).

In many applications it is desirable toprocess the reconstructed images to changetheir characteristics, to reformat an image or setof images, or to change the display of images toproduce specific views of anatomical regions.

These post-processing (after reconstruc-tion) functions are performed by a computer.In some MRI systems some of the post pro-cessing is performed on a work-station com-puter that is in addition to the computer con-tained in the MRI system.



Mind Map SummaryMagnetic Resonance Imaging System Components

MAGNETIC

FIELD

Strength HomogeneityDirection

Magnetic

Det

erm

ines

Redu

ces

Imp

rove

s

Shimming

Typeand

Design

MagneticSusceptibleMaterials

ShieldingReduces

EXTERNAL FIELD

Z Gradient

Z G

radi

ent

Imp

roves

X Gradient

Strength

Slew Rate

Rise Time

DIRECTIONS

Radio Frequency Coils

Shielding

Interf

erenc

e

Pulses Signals

GRADIENTS

COMPUTER

Controls Imaging Process

Reconstructs Images

The magnetic resonance imaging system consists of several major components that functiontogether to produce images. During the image acquisition process the patient’s body is placed ina strong magnetic field. At each point, the magnetic field has a specific direction. This direction isused as a reference for expressing the direction of tissue magnetization. The strength of a magneticfield is determined by the type and design of the magnet. Superconducting magnets can producestrong magnetic fields. Resistive and permanent magnets are limited to relatively weak fieldstrengths. The homogeneity, or uniformity of field strength is necessary for good imaging. Homo-geneity is reduced by magnetically susceptible materials that come into the field and produce dis-tortions. This can occur in both the external field and within a patient’s body. Shimming is theprocess of adjusting the magnetic field to make it more homogeneous. This can be achieved bypassive shims that are added when a magnet is installed and with active shimming produced byadjusting the currents in the shimming coils.

Shielding of the magnetic field reduces the size and strength of the external magnetic fieldand also improves homogeneity by protecting from interference caused by objects in the exter-nal field area.

A gradient is an intentional variation in magnetic field strength that is produced by the gra-dient coils. There are three basic gradient coils that are oriented to produce gradients in the three

orthogonal directions. Gradients perform several functions during the image acquisition process.An important characteristic of a gradient, especially for some advanced image procedures, is itsstrength and how fast it can be turned on and off.

The MRI process consists of an exchange of RF pulses and signals between the equipment andthe patient’s body. This is done through the RF coils that serve as the antenna for transmitting thepulses and receiving the signals. It is necessary to shield the imaging area by enclosing it in a con-ductive metal (copper) room to block external RF interference.

The imaging process is controlled by information stored in a computer. The protocols pro-grammed into the computer and selected by the operator guide the imaging process and determinethe characteristics of the images. The RF signals collected from the patient’s body during the acqui-sition process are used by the computer to reconstruct the image.


3Nuclear Magnetic

Resonance

25

Introduction And OverviewWhen certain materials, such as tissue, areplaced in a strong magnetic field, two thingshappen. The materials take on a resonant char-acteristic and they become magnetized. In thischapter we will consider the resonant charac-teristic. In Chapter 4 we will study the magne-tization effect. Resonance means the materialscan absorb and then re-radiate RF radiation ata specific frequency, like a radio receiver-trans-mitter, as illustrated in Figure 3-1. It is actuallythe nuclei of the atoms that resonate. The phe-nomenon is generally known as nuclear mag-netic resonance (NMR). The resonant fre-quency of material such as tissue is typically inthe RF range so that the emitted radiation is in

the form of radio signals. The specific resonantfrequency is determined by three factors asshown in the illustration and will be describedin detail later. The characteristics of the RF sig-nals emitted by the material are determined bycertain physical and chemical characteristicsof the material. The RF signals produced by theNMR process can be displayed either in theform of images (MRI) or as a graph depictingchemical composition (MR spectroscopy).

Magnetic NucleiMaterials that participate in the MR processmust contain nuclei with specific magneticproperties. In order to interact with a mag-netic field, the nuclei themselves must be

small magnets and have a magnetic propertyor magnetic moment, as shown in Figure 3-2.The magnetic characteristic of an individualnucleus is determined by its neutron-protoncomposition. Only certain nuclides with anodd number of neutrons and protons aremagnetic. Even though most chemical ele-ments have one or more isotopes with mag-netic nuclei, the number of magnetic iso-topes that might be useful for either imagingor in vivo spectroscopic analysis is somewhatlimited. Among the nuclides that are mag-netic and can participate in an NMR process,the amount of signal produced by eachnuclide varies considerably.

Spins

Protons and neutrons that make up a nucleushave an intrinsic angular momentum or spin.Pairs of protons and neutrons align in such away that their spins cancel. However, whenthere is an odd number of protons or neutrons

(odd mass numbers), some of the spins willnot be canceled and the total nucleus willhave a net spin characteristic. It is this spin-ning characteristic of a particle with an elec-tric charge (the nucleus) that produces a mag-netic property known as the magnetic moment.

It is for this reason that magnetic nuclei,such as protons, are often referred to as spins.

The magnetic property, or magneticmoment, of a nucleus has a specific direction.In Figure 3-2, the direction of the magneticmoment is indicated by an arrow drawnthrough the nucleus.

RF Signal Intensity

The intensity of the RF signal emitted by tis-sue is probably the most significant factor indetermining image quality and the timerequired to acquire an image. This importantissue is considered in Chapters 10 and 11. Wenow begin to introduce the factors that con-tribute to signal intensity.


MAGNETIC RESONANCE

Pulse

Nuclide(H-1, 42.58 MHz/T)

Field Strength

Determined By:

Molecular Structure(Chemical Shift)

Tissue

Resonant (Larmor) Frequency

Signal

Figure 3-1. The concept of Nuclear Magnetic Resonance (NMR).

During the imaging process, the body sec-tion is divided into an array of individual vol-ume elements, or voxels. It is the signal inten-sity from each voxel that determines imagequality. The signal is produced by the mag-netic nuclei within each voxel. Therefore, sig-nal intensity is, in general, proportional to thequantity of magnetic nuclei within an indi-vidual voxel. We now consider the factors thataffect the number of magnetic nuclei withinan individual voxel.

Relative Signal Strength

The relative signal strength from the variouschemical elements in tissue is determined bythree factors: (1) tissue concentration of theelement; (2) isotopic abundance; and (3) sen-sitivity of the specific nuclide.

In comparison to all other nuclides, hydro-gen produces an extremely strong signal. Thisresults from its high values for each of thethree contributing factors.

Of the three factors, only the concentra-tion, or density, of the nuclei varies frompoint to point within an imaged section oftissue. The quantity is often referred to as pro-ton density and is the most fundamental tissue

characteristic that determines the intensity of the RF signal from an individual voxel, andthe resulting pixel brightness. In most imagingsituations, pixel brightness is proportional tothe density (concentration) of nuclei (protons)in the corresponding voxel, although addi-tional factors, such as relaxation times, modifythis relationship.

Protons in solids, such as the tabletop andbone, do not produce signals. Signals comeonly from protons in molecules that are freeto move, as in a liquid state.

Tissue Concentration of Elements

The concentration of chemical elements in tissuecovers a considerable range, depending on tissuetype and such factors as metabolic or pathologicstate. The concentrations of elements in tissueare in two groups. Four elements—hydrogen, car-bon, nitrogen, and oxygen—typically make upat least 99% of tissue mass.

The most abundant isotopes of the four ele-ments are hydrogen-1, carbon-12, nitrogen-14,and oxygen-16. Note that the mass number ofhydrogen (1) is odd while the mass numbers ofthe other three (12, 14, 16) are even. Therefore,hydrogen is the only one of these four isotopesthat has a strong magnetic nucleus. Thenucleus of the hydrogen-1 atom is a single pro-ton. Among all the chemical elements, hydro-gen is unique in that it occurs in relatively highconcentrations in most tissues, and the mostabundant isotope (H-1) has a magnetic nucleus.

Other elements, such as sodium, phos-phorus, potassium, and magnesium, are pres-ent in very low concentrations. Calcium isconcentrated in bone or localized deposits.

Within this group of elements with low tis-sue concentrations are several with magneticnuclei. These include fluorine-19, sodium-23,phosphorus-31, and potassium-39.

NUCLEAR MAGNETIC RESONANCE 27

SPIN

MAGNETIC PROPERTIES OF NUCLEI

Magnetic Moment

Odd Mass NumbersHydrogen-1 Carbon-12

Oxygen-16

Even Mass Numbers

Non-Magnetic

Figure 3-2. Magnetic and non-magnetic nuclei.

Isotopic Abundance

Most chemical elements have several isotopes.When a chemical element is found in a natu-rally occurring substance, such as tissue, mostof the element is typically in the form of oneisotope, with very low concentrations of theother isotopic forms. For the three elements—carbon, nitrogen, and oxygen—that have ahigh concentration in tissue, the magneticisotopes are the ones with a low abundance inthe natural state. These include carbon-l3,nitrogen-15, and oxygen-17.

Relative Sensitivity and Signal Strength

The signal strength produced by an equal quan-tity of the various nuclei also varies over a con-siderable range. This inherent NMR sensitivityis typically expressed relative to hydrogen-1,which produces the strongest signal of all of thenuclides. The relative sensitivities of some mag-netic nuclides are shown in Table 3-1.

Table 3-1. Relative Sensitivities of Some Magnetic Nuclides

Nuclide SensitivityHydrogen-1 1.0Fluorine-19 0.83Sodium-23 0.093Phosphorus-1 0.066

In summary, hydrogen has a lot going for it:1) a high tissue concentration; 2) the mostabundant isotope (H-1) is magnetic; and 3) itproduces a relatively strong signal comparedto an equal concentration of other nuclei.That is why hydrogen is the only element thatis imaged with conventional MRI systems.

Radio Frequency EnergyDuring an imaging procedure, RF energy isexchanged between the imaging system andthe patient’s body. This exchange takes placethrough a set of coils located relatively closeto the patient’s body as we saw in Chapter 2.The RF coils are the antennae that transmitenergy to and receive signals from the tissue.

Pulses

RF energy is applied to the body in severalshort pulses during each imaging cycle. Thestrength of the pulses is described in terms ofthe angle through which they rotate or flipthe magnetic nuclei and the resulting tissuemagnetization, as described later. Many imag-ing methods use both 90˚ and 180˚ pulses ineach cycle.

Signals

At a specific time in each imaging cycle, thetissue is stimulated to emit an RF signal,which is picked up by the coils, analyzed, andused to form the image. The spin echo or gra-dient echo methods are generally used tostimulate signal emission. Therefore, the sig-nals from the patient’s body are commonlyreferred to as echoes.

Nuclear MagneticInteractionsThe NMR process is a series of interactionsinvolving the magnetic nuclei, a magneticfield, and RF energy pulses and signals.

Nuclear Alignment

Recall that a magnetic nucleus is characterized bya magnetic moment. The direction of the mag-netic moment is represented by a small arrowpassing through the nucleus. If we think of the


nucleus as a small conventional magnet, themagnetic moment arrow corresponds to thesouth pole-north pole direction of the magnet.

In the absence of a strong magnetic field,magnetic moments of nuclei are randomly ori-ented in space. Many nuclei in tissue are notin a rigid structure and are free to change direc-tion. In fact, nuclei are constantly tumbling, orchanging direction, because of thermal activ-ity within the material; in this case, tissue.

When a material containing magneticnuclei is placed in a magnetic field, the nucleiexperience a torque that encourages them toalign with the direction of the field. In thehuman body, however, thermal energy agi-tates the nuclei and keeps most of them fromaligning parallel to the magnetic field. Thenumber of nuclei that do align with the mag-netic field is proportional to the field strength.The magnetic fields used for imaging can alignonly a few of every million magnetic nucleipresent. However, this is sufficient to producea useful NMR effect.

Precession and Resonance

When a spinning magnetic nucleus alignswith a magnetic field, it is not fixed; thenuclear magnetic moment precesses, or oscil-lates, about the axis of the magnetic field, asshown in Figure 3-3. The precessing motion isa physical phenomenon that results from aninteraction between the magnetic field andthe spinning momentum of the nucleus.

Precession is often observed with a child’sspinning top. A spinning top does not standvertical for long, but begins to wobble, or pre-cess. In this case, the precession is caused by aninteraction between the earth’s gravitationalfield and the spinning momentum of the top.

The precession rate (cycles per second) isdirectly proportional to the strength of themagnetic field. It is this precessing motion

that makes a nucleus sensitive and receptiveto incoming RF energy when the RF frequencymatches the precession rate. This precessionrate corresponds to the resonant frequency. Itis the precessing nuclei, typically protons, thatare tuned to receive and transmit RF energy.

Excitation

If a pulse of RF energy with a frequency corre-sponding to the nuclear precession rate isapplied to the material, some of the energywill be absorbed by the individual nuclei. Theabsorption of energy by a nucleus flips itsalignment away from the direction of themagnetic field, as shown in Figure 3-4. Thisincreased energy places the nucleus in anunnatural, or excited, state.

In MRI an RF pulse is used that flips someof the nuclei into the transverse plane of themagnetic field. In this excited state the pre-cession is now transformed into a spinningmotion of the nucleus around the axis of themagnetic field. It should be noted that thisspinning motion is an enhanced precession


PROTON PRECESSION AND RESONANCE

Frequency = Precession Rate

RF Pulse

Magnetic FieldLongitudinal View

Figure 3-3. Magnetic nuclei precession and resonance in a magnetic field.

and is different from the intrinsic spin of anucleus about its own axis.

The significance of a magnetic nucleus spin-ning around the axis of the magnetic field isthat this motion now generates an RF signal asshown in Figure 3-5. It is this signal, from manynuclei, that is collected to form the MR image.

RelaxationWhen a nucleus is in an excited state, it expe-riences an increased torque from the magneticfield, urging it to realign. The nucleus canreturn to a position of alignment by transfer-ring its excess energy to other nuclei or thegeneral structure of the material. This processis known as relaxation.

Relaxation is not instantaneous followingan excitation. It cannot occur until thenucleus is able to transfer its excess energy.How quickly the energy transfer takes placedepends on the physical characteristics of thetissue. In fact, the nuclear relaxation rate (ortime) is, in many cases, the most significant

factor in producing contrast among differenttypes of tissue in an image.

We are more interested in the collectiverelaxation of many nuclei that produce themagnetization of tissue and will return to thispoint in the next chapter.

ResonanceThe significance of the nuclear precession isthat it causes the nucleus to be extremely sen-sitive, or tuned, to RF energy that has a fre-quency identical with the precession fre-quency (rate). This condition is known asresonance and is the basis for all MR proce-dures. NMR is the process in which a nucleusresonates, or “tunes in,” when it is in a mag-netic field.

Resonance is fundamental to the absorp-tion and emission of energy by many objectsand devices. Objects are most effective inexchanging energy at their own resonant fre-quency. The resonance of an object or device


EXCITATION

Magnetic Field Views

RF Pulse

90˚

90˚ Flip

Longitudinal Transverse

Figure 3-4. The excitation of a magnetic nucleus by the application of a pulse of RF energy.

is determined by certain physical characteris-tics. Let us consider two common examples.

Radio receivers operate on the principleof resonant frequency. A receiver can select aspecific broadcast station because each sta-tion transmits a different frequency. Tuning aradio is actually adjusting its resonant fre-quency. Its receiver is very sensitive to radiosignals at its resonant frequency and insensi-tive to all other frequencies.

The strings of a musical instrument alsohave specific resonant frequencies. This is thefrequency at which the string vibrates to pro-duce a specific audio frequency, or musical note.The resonant frequency of a string dependson the amount of tension. It can be changed,or tuned, by changing the tension. This issomewhat analogous to the resonant fre-quency of a magnetic nucleus being depend-ent on the strength of the magnetic field inwhich it is located.

Larmor Frequency

The resonant frequency of a nucleus is deter-mined by a combination of nuclear charac-teristics and the strength of the magneticfield. The resonant frequency is also knownas the Larmor frequency. The specific relation-ship between resonant frequency and fieldstrength is an inherent characteristic of eachnuclide and is generally designated the gyro-magnetic ratio. The Larmor frequencies [inmegahertz (MHz)] for selected nuclides in amagnetic field of 1 T are shown in Table 3-2.

Table 3-2. Larmor Frequencies for SelectedNuclides in a Magnetic Field of 1 T

Larmor Frequency Nuclide (MHz)Hydrogen-1 42.58Fluorine-19 40.05Phosphorus-31 17.24Sodium-23 11.26

The fact that different nuclides have differentresonant frequencies means that most MRprocedures can “look at” only one chemicalelement (nuclide) at a time.

Field Strength

For all nuclides, the resonant frequency is pro-portional to the strength of the magneticfield. In a very general sense, increasing themagnetic field strength increases the tensionon the nuclei (as with the strings of a musicalinstrument) and increases the resonant fre-quency. The fact that a specific nuclide can betuned to different radio frequencies by vary-ing the field strength (i.e., applying gradients)is used in the imaging process.


Magnetic FieldTransverse View

RF Signal

Spin Rate = Frequency

RF SIGNAL PRODUCTION

Figure 3-5. RF signal production by magnetic nuclei spinning in the transverse

plane of a magnetic field.

Chemical Shift

The resonant frequency of magnetic nuclei,such as protons, is also affected by the struc-ture of the molecule in which they are located.

When a proton, or other magnetic nucleus,is part of a molecule, it is slightly shieldedfrom the large magnetic field. The amount ofshielding depends on the chemical composi-tion of the molecule. This means that protonsin different chemical compounds will be inslightly different field strengths and will there-fore resonate at different frequencies. Thischange in resonant frequency from one com-pound to another is known as chemical shift. Itcan be used to perform chemical analysis in thetechnique of MR spectroscopy and to produceimages based on chemical composition. How-ever, in conventional MRI the chemical-shifteffect can be the source of an unwanted artifact.

In tissue the chemical shift in resonantfrequency between the fat and water isapproximately 3.3 ppm, as shown in Figure 3-6. At a field strength of 1.5 T the protons havea basic resonant frequency of approximately64 MHz. Multiplying this by 3.3 gives a water-fat chemical shift of approximately 210 Hz. Ata field strength of 0.5 T the chemical shiftwould be only 70 Hz.

There are several imaging techniques thatcan be used to selectively image either thewater or fat tissue components. One approachis to suppress either the fat or water signalwith specially designed RF pulses. This tech-nique is known as spectral presaturation andwill be described in Chapter 8. Another tech-nique makes use of the fact that the signalsfrom water and fat are not always in step, orin phase, with each other and can be sepa-rated to create either water or fat images.


Water

Fat

Relative Resonant Frequency

Sign

al In

tens

ity

0 1

210 Hzat

1.5T

2 3 4 5 ppm

CHEMICAL SHIFT

Figure 3-6. The chemical shift effect on the relative resonant frequency of protons in fat and in water.

Mind Map SummaryNuclear Magnetic Resonance


MAGNETIC FIELD

FIELD

STRENGTH

H-1 is42.58 MHz/T

Longitudinal

RF Pulse

RF Signal

Transverse Spinning

MoleculeStructure

Magnetic Nuclei

H-1 has HighRelative Sensitivity

ImageOf

Protons(Hydrogen)

Chemical Shift

*Used for Spectroscopy*Used to Suppress Fat Signals*Can Produce Artifacts

Resonant FrequencyDetermined By

Strength

IsotopicAbundance

TissueConcentration

Nuclide

Determined By

H IGHH

-1

SPIN

When a magnetic nucleus is located in a strong magnetic field, it resonates. In effect, it becomesa tuned radio receiver and transmitter. The resonance occurs because the spinning nucleus pre-cesses at a rate that is in the radio frequency range. The resonant frequency is determined by threefactors. Each specific nuclide has a unique resonant frequency. The resonant frequency is affectedto a small degree by the structure of the molecule containing the magnetic nucleus. This, thechemical shift effect, is useful for spectroscopy and to suppress fat signals in images. It can alsolead to a certain type of image artifact. The resonant frequency is directly proportional to thestrength of the magnetic field. This is useful because it makes it possible to tune the various partsof a body to different frequencies by applying magnetic field gradients.

When an RF pulse is applied to a magnetic nucleus oriented in the longitudinal direction, itcan be flipped into the transverse plane. There the nucleus spins around the axis of the magneticfield and generates an RF signal. It is the signals from many spinning nuclei that are collectedand used to form the image. It is necessary to have strong signals to produce good images. Signalstrength depends on three factors. Each magnetic nuclide has a unique sensitivity or relative sig-nal strength. All chemical elements have several different isotopes, but all isotopes of an elementare usually not in the form of magnetic nuclei. Therefore, the abundance of the magnetic isotope

for a specific element has a major effect on signal strength. To produce strong signals a tissue musthave a relatively high concentration of a chemical element and the most abundant isotope ofthat element must be magnetic.

Hydrogen is the only chemical element with a high concentration in tissue and body fluids inthe form of an isotope that has a magnetic nucleus. Therefore, MR imaging is essentially limitedto visualizing only one chemical element, hydrogen.


4Tissue Magnetization

And Relaxation

35

Introduction And Overview

We have considered the behavior of individualnuclei when placed in a magnetic field. MRIdepends on the collective, or net, magneticeffect of a large number of nuclei within a spe-cific voxel of tissue. If a voxel of tissue containsmore nuclei aligned in one direction than inother directions, the tissue will be temporarilymagnetized in that particular direction. Thisprocess is illustrated in Figure 4-1. In theabsence of a magnetic field, the nuclei are ran-domly oriented and produce no net magneticeffect. This is the normal state of tissue beforebeing placed in a magnetic field. When thetissue is placed in a magnetic field, and some

of the nuclei align with the field, their com-bined effect is to magnetize the tissue in thedirection of the magnetic field. A large arrow,the magnetization vector, is used to indicate theamount and direction of the magnetization.When tissue is placed in a magnetic field, themaximum magnetization that can be pro-duced depends on three factors: (1) the con-centration (density) of magnetic nuclei, typi-cally protons, in the tissue voxel; (2) themagnetic sensitivity of the nuclide; and (3) thestrength of the magnetic field. Since an imag-ing magnetic field aligns a very small fractionof the magnetic nuclei, the tissues are neverfully magnetized. The amount of tissue mag-netization determines the strength of the RF

signals emitted by the tissue during an imagingor analytical procedure. This, in turn, affectsimage quality and imaging time, as explainedin Chapter 10.

Let us recall that an MR image is an imageof magnetized tissue and that the contrast wesee is produced by different levels of magneti-zation that exist in the different tissues at thetime when “the picture is snapped.” As we willsee in this chapter the level of magnetizationat specific times during the imaging process isdetermined by the three tissue characteristics:proton density (PD), T1, and T2.

We will now see how these characteristicsproduce image contrast.

Tissue MagnetizationWhen tissue is placed in a magnetic field, itreaches its maximum magnetization within afew seconds and remains at that level unless it isdisturbed by a change in the magnetic field or bypulses of RF energy applied at the resonant fre-quency. The MRI procedure is a dynamic processin which tissue is cycled through changes in itsmagnetization during each imaging cycle.

Magnetic DirectionThe direction of tissue magnetization is speci-fied in reference to the direction of the appliedmagnetic field, as shown in Figure 4-2. Thereare two principle directions that tissue is mag-netized during the imaging process. Longitu-dinal magnetization is when the tissue is mag-netized in a direction parallel to the directionof the field. Transverse magnetization is whenthe direction of tissue magnetization is at a 90°angle with respect to the direction of the mag-netic field and is in the transverse plane.

Magnetic FlippingThe direction of tissue magnetization can bechanged or flipped by applying a pulse of RFenergy. This is done many times throughoutthe imaging process.

Flip Angle

The angle the magnetization is flipped is deter-mined by the duration and strength of the RFpulse. Pulses are characterized by their flip angles.

Pulses with 90° and 180° flip angles are themost common but smaller flip angle pulses arealso used in some imaging methods, such asgradient echo imaging.

The 90˚ Pulse, Saturation and Excitation

When a 90˚ pulse is applied to longitudinalmagnetization, it flips it into the transverseplane as shown in Figure 4-3. This has twoeffects. First, it reduces the longitudinal mag-netization to zero, a condition called satura-tion. It also produces transverse magnetization.As we will soon learn, transverse magnetiza-tion is an unstable or excited condition. There-fore, when a 90˚ pulse is applied to longitudi-nal magnetization, it produces both saturationof the longitudinal magnetization and a con-dition of excitation (transverse magnetization).


MAGNETIZATION

Longitudinal View

MAGNETIZATION VECTOR

Figure 4-1. The magnetization of tissueproduced by the alignment of magnetic

nuclei (protons) in a magnetic field.

The actual direction of magnetization isnot limited to longitudinal or transverse. Itcan exist in any direction. In principle, mag-netization can have both longitudinal andtransverse components. Since the two com-ponents have distinctly different characteris-tics, we consider them independently.

Longitudinal MagnetizationAnd Relaxation As we have seen, when tissue is placed in amagnetic field, it becomes magnetized in thelongitudinal direction. It will remain in thisstate until the magnetic field is changed oruntil the magnetization is redirected by theapplication of an RF pulse. If the magnetiza-tion is temporarily redirected by an RF pulse,it will then, over a period of time, return to

its original longitudinal position. If we con-sider only the longitudinal magnetization, itregrows after it has been reduced to zero, orsaturated. This regrowth, or recovery, of lon-gitudinal magnetization is the relaxationprocess, which occurs after saturation. Thetime required for the longitudinal magneti-zation to regrow, or relax, depends on char-acteristics of the material and the strength ofthe magnetic field.

Longitudinal magnetization does notgrow at a constant rate, but at an exponentialrate, as shown in Figure 4-4. An importantconcept to remember is that the MR image isan image of magnetized tissue with brightnessindicating the level of magnetization. Duringthe relaxation process, the level of magnetiza-tion is changing. Therefore, the brightness oftissue (if we could see it) is also changing as

TISSUE MAGNETIZATION AND RELAXATION 37

MAGNETIC DIRECTION

Longitudinal

Views of the Magnetic Field in a Voxel

Transverse

Figure 4-2. Longitudinal and transverse magnetization.

indicated by the scale on the right of the illus-tration. Saturation turns the tissue dark andthen it recovers brightness during the relax-ation period.

The characteristic that varies from onetype of tissue to another, and can be used toproduce image contrast, is the time requiredfor the magnetization to re-grow, or the relax-ation time. Because of its exponential nature, itis difficult to determine exactly when the mag-netization has reached its maximum. The con-vention is to specify the relaxation time interms of the time required for the magnetizationto reach 63% of its maximum. This time, thelongitudinal relaxation time, is designated T1.The 63% value is used because of mathemati-cal, rather than clinical, considerations. Lon-gitudinal magnetization continues to growwith time, and reaches 87% of its maximum


LONGITUDINAL MAGNETIZATIONRELAXATION (GROWTH)

Imag

e Br

ight

ness

Mag

netiz

atio

n (%

)

T1

63%

100

00

200 400 600 800 1000

Time (msec)

Figure 4-4. The growth of longitudinalmagnetization (and tissue brightness) duringthe relaxation process following saturation.

MAGNETICSATURATION AND EXCITATION

Saturation Excitation

RF Pulse

90˚

TransverseExcitationProduced

LongitudinalMagnetization

Reduced to Zero

Figure 4-3. The application of a 90˚ RF pulse to longitudinal magnetization produces saturation ofthe longitudinal magnetization and creates transverse magnetization, an excited condition.

after two T1 intervals, and 95% after three T1intervals. For practical purposes, the magneti-zation can be considered fully recovered afterapproximately three times the T1 value of thespecific tissue. We will see later that this mustbe taken into consideration when setting upan imaging procedure.

T1 Contrast

The time required for a specific level of longi-tudinal magnetization regrowth varies fromtissue to tissue. Figure 4-5 shows the regrowthof two tissues with different T1 values. In thisillustration we watch the intensity of bright-ness of a voxel of tissue during the relaxationprocess. Let us recall that the brightness of atissue (RF signal intensity) is determined by thelevel of magnetization existing in a voxel of tis-sue at any instant in time. What we see in animage depends on when we “snap the picture”during the relaxation process. The importantthing to notice is that the tissue with the short-est T1 has the highest level of magnetizationat any particular time. The clinical significanceof this is that tissues with short T1 values willbe bright in T1-weighted images.

Table 4-1 lists typical T1 values for varioustissues. Two materials establish the lower andupper values for the T1 range: fat has a shortT1, and fluid falls at the other extreme (longT1). Therefore, in T1-weighted images, fat isgenerally bright, and fluid [cerebrospinal fluid(CSF), cyst, etc.] is dark. Most other body tissuesare within the range between fat and fluid.

The longitudinal relaxation process involvesan interaction between the protons and theirimmediate molecular environment. The rate ofrelaxation (T1 value) is related to the naturallyoccurring molecular motion. The molecularmotion is determined by the physical state ofthe material and the size of the molecules. Therelatively rigid structure of solids does not pro-vide an environment for rapid relaxation,which results in long T1 values. Molecularmotion in fluids, and fluid-like substances, ismore inducible to the relaxation process. Inthis environment molecular size becomes animportant characteristic.

Relaxation is enhanced by a general match-ing of the proton resonant frequency and thefrequency associated with the molecularmotions. Therefore, factors that change either ofthese two frequencies will generally have aneffect on T1 values.

Molecular Size

Small molecules, such as water, have fastermolecular motions than large molecules, suchas lipids. The frequencies associated with themolecular motion of water molecules are bothhigher and more dispersed over a larger rangefor the larger molecules. This reduces thematch between the frequencies of the protonsand the frequencies of the molecular environ-ment. This is why water and similar fluidshave relatively long T1 values. Larger mole-cules, which have slower and less dispersedmolecular movement, have a better frequency


T1 CONTRAST

Imag

e Br

ight

nessShort T

1

Long T1

Mag

netiz

atio

n (%

)

100

00

200 400 600 800 1000

Contrast

Time (msec)

Figure 4-5. The formation of contrastbetween two tissues with different T1 values.

match with the proton resonant frequencies.This enhances the relaxation process and pro-duces short T1 values. Fat is an excellentexample of a large molecular structure thatexhibits this characteristic.

Tissues generally contain a combinationof water and a variety of larger molecules.Some of the water can be in a relatively freestate while other water is bound to some ofthe larger molecules. In general, the T1 valueof the tissue is probably affected by theexchange of water between the free and thebound states. When the water is bound tolarger molecular structures, it takes on themotion characteristics of the larger molecule.Factors such as a pathologic process, whichalters the water composition of tissue, willgenerally alter the T1 values.

Magnetic Field Strength EffectT1 values depend on the strength of the mag-netic field. This is because the field strengthaffects the resonant frequency of the protons.As field strength is increased, the resonant fre-quency also increases and becomes lessmatched to the molecular motion frequen-cies. This results in an increase in T1 values,as indicated in Table 4-1.

Let us now combine two factors to createa T1 image as illustrated in Figure 4-6. Onefactor is that different tissues have different

T1 values and rates of regrowth of longitudinalmagnetization. This then causes the differenttissues to be at different levels of magnetiza-tion (brightness) when the picture is snappedduring the relaxation period. Here we see theorder of tissue brightness is inversely related toT1 values. In principle, the tissues with shortT1 values get brighter faster and are at a higherlevel when the picture is snapped.

Transverse MagnetizationAnd Relaxation Transverse magnetization is produced by apply-ing a pulse of RF energy to the magnetized tissue.This is typically done with a 90˚ pulse, whichconverts longitudinal magnetization into trans-verse magnetization. Transverse magnetizationis an unstable, or excited, condition and quicklydecays after the termination of the excitationpulse. The decay of transverse magnetization isalso a relaxation process, which can be charac-terized by specific relaxation times, or T2 values.Different types of tissue have different T2 valuesthat can be used to discriminate among tissuesand contribute to image contrast.

Transverse magnetization is used duringthe image formation process for two reasons:(1) to develop image contrast based on differ-ences in T2 values; and (2) to generate the RFsignals emitted by the tissue. Longitudinal


Table 4-1. T2 and T1 Values for Various Tissues

T2 T1 (0.5 T) T1 (1.5 T)Tissue (msec) (msec) (msec)Adipose (Fat) 80 210 260Liver 42 350 500Muscle 45 550 870White Matter 90 500 780Gray Matter 100 650 920CSF 160 1800 2400

magnetization is an RF silent condition anddoes not produce any signal. However, trans-verse magnetization is a spinning magneticcondition within each tissue voxel, and thatgenerates an RF signal. As we will see in thenext chapter, each imaging cycle must con-clude with transverse magnetization to pro-duce the RF signal used to form the image.

The characteristics of transverse magneti-zation and relaxation are quite different fromthose for the longitudinal direction. A majordifference is that transverse magnetizationis an unstable condition and the relaxation

process results in the decay, or decrease, inmagnetization, as shown in Figure 4-7. The T2value is the time required for 63% of the initialmagnetization to dissipate. After one T2, 37%of the initial magnetization is present.

T2 ContrastThe difference in T2 values of tissues is thesource of contrast in T2-weighted images. Thisis illustrated in Figure 4-8. Here we watch twotissues, with different T2 values, during therelaxation process. We see that they are bothgetting darker with time as the magnetization


T1-WEIGHTED IMAGE

Tissue T1 Values (msec)

Fat White Gray Fluid

Short 200 400 600 800 1000 Long

Figure 4-6. A T1 image showing the relationship of tissue brightness (signal intensity) to T1 values and level of magnetization during the longitudinal relaxation process.

decays. However, they are not getting darkerat the same rate. The tissue with the shorterT2 becomes darker faster leaving the tissuewith the longer T2 to be bright at times dur-ing the relaxation time.

What we will actually see in a T2-weightedimage, as shown in Figure 4-9, depends on thelevel of magnetization at the time when wesnap the picture. The important thing toobserve here is that the tissues with long T2values are bright in T2 images.

In general, a T2-weighted image appears tobe a reversal of a T1-weighted image. Tissuesthat are bright in one image are dark in theother image. This is because of a combinationof two factors. One factor is that T1 and T2 val-ues are generally related. Even though T2 val-ues are much shorter than T1 values, as shownin Table 4-1, they are somewhat proportional.Tissues with long T1 values usually have longT2 values. The other factor is that the order ofbrightness in a T2 image is in the same directionas the T2 values. Remember, it was a reversedrelationship for T1 images.

The decay of transverse magnetization(i.e., relaxation) occurs because of a dephasingamong individual nuclei (protons) within theindividual voxels, as shown in Figure 4-10.

Two basic conditions are required fortransverse magnetization: (1) the magneticmoments of the nuclei must be oriented in thetransverse direction, or plane; and (2) a major-ity of the magnetic moments must be in thesame direction, or in phase, within the trans-verse plane. When a nucleus has a transverseorientation, it is actually spinning around anaxis that is parallel to the magnetic field.

After the application of a 90˚ pulse, thenuclei have a transverse orientation and arerotating together, or in phase, around the mag-netic field axis. This rotation or spin is a resultof the normal precession discussed earlier. Theprecession rate, or resonant frequency, dependson the strength of the magnetic field where thenuclei are located. Nuclei located in field areaswith different strengths spin (precess) at differ-ent rates. Even within a very small volume oftissue, nuclei are in slightly different magneticfield strengths. As a result, some nuclei spinfaster than others. Also, there are interactions


TRANSVERSE MAGNETIZATIONRELAXATION (DECAY)

Imag

e Br

ight

ness

Mag

netiz

atio

n (%

)

T2

37%

100

00

20 40 60 80 100

Time (msec)

Figure 4-7. The decay of transversemagnetization during the relaxation process

and the associated tissue brightness.

Contrast

Long T2

Short T2

T2 CONTRAST

Imag

e Br

ight

ness

Mag

netiz

atio

n (%

)

100

00

20 40 60 80 100

Time (msec)

Figure 4-8. The formation of T2 contrastduring the decay of transverse magnetization.

(spin-spin interactions) among the spinningnuclei. After a short period of time, the nucleiare not spinning in phase. As the directions ofthe nuclei begin to spread and they dephase,the magnetization of the tissue decreases. Ashort time later, the nuclei are randomly ori-ented in the transverse plane, and there is notransverse magnetization.

Proton Dephasing

Two major effects contribute to the dephasingof the nuclei and the resulting transverse relax-ation. In the imaging process the spin echotechnique is used to separate the two sourcesof dephasing, as we will see in Chapter 6.

T2 Tissue Characteristics

One effect is the exchange of energy amongthe spinning nuclei (spin-spin interactions),which results in relatively slow dephasingand loss of magnetization. The rate at whichthis occurs is determined by characteristics ofthe tissue. It is this dephasing activity that ischaracterized by the T2 values as shown inTable 4-1.

T2* Magnetic Field Effects

A second effect, which produces relativelyrapid dephasing of the nuclei and loss oftransverse magnetization, is the inherentinhomogeneity of the magnetic field within


Figure 4-9. A T2 image showing the relationship of tissue brightness (signal intensity) to T2 values.

each individual voxel. The field inhomogenei-ties are sufficient to produce rapid dephasing.This effect, which is different from the basicT2 characteristics of the tissue, tends to maskthe true relaxation characteristics of the tissue.In other words, the actual transverse magne-tization relaxes much faster than the tissuecharacteristics would indicate. This real relax-ation time is designated as T2*. The value ofT2* is usually much less than the tissue T2value, as illustrated in Figure 4-11. Several fac-tors can contribute to field inhomogeneitiesand to T2* decay. One is the general conditionof the magnetic field. Some fields are morehomogeneous than others. Another factor isthat different tissues or materials in the bodymight have different magnetic susceptibilities.Susceptibility is a characteristic of a materialthat determines its ability to become magnet-ized when it is in a magnetic field. If a regionof tissue contains materials with different sus-ceptibilities, this results in a reduction of fieldhomogeneity.

Magnetic SusceptibilityThe magnetization of tissue that we have beendiscussing is a nuclear magnetic effect pro-duced by the alignment of magnetic nuclei ina magnetic field. Other materials can becomemagnetized by other, non-nuclear effects.

Many materials are susceptible to magneticfields and become magnetized when locatedin fields. The susceptibility of a material isdetermined by the orbital electrons in theatom rather than the magnetic properties ofthe nucleus. Significant susceptibility is pres-ent only when there are unpaired electrons inthe outer orbit.

There are three general types of materialswith respect to magnetic susceptibility: diamag-netic, paramagnetic, and ferromagnetic. The pri-mary characteristic of each type is the amountand direction of magnetization that the mate-rial develops when placed in a magnetic field.There are situations when each type plays a rolein the MR imaging process.


TRANSVERSE MAGNETIZATION DECAY

Dephasing

Tissue Voxels

Figure 4-10. The dephasing of protons that produces transverse magnetization decay.

Contrast Agents

The inherent tissue characteristics (PD, T1, andT2) do not always produce adequate contrast forsome clinical objectives. It is possible to admin-ister materials (i.e., contrast agents) that willalter the magnetic characteristics within specifictissues or anatomical regions. There are severaldifferent types of contrast agents, which willnow be considered. Contrast agents used in MRIare generally based on relaxation effects.

Diamagnetic Materials

Diamagnetic materials have negative and rela-tively low magnetic susceptibility. This meansthat they develop only low levels of magneti-zation and it is in a direction opposite to the

direction of the magnetic field. Although manybiological molecules are diamagnetic, this isnot a significant factor in MR imaging.

Paramagnetic MaterialsParamagnetic materials play an important rolein contrast enhancement. They are materialswith unpaired electrons that give each atom apermanent magnetic property. In paramag-netic materials each atom is magneticallyindependent, which distinguishes it fromother materials to be discussed later.

Paramagnetic substances include metalions such as gadolinium, manganese, iron, andchromium. Other substances such as nitroxidefree radicals and molecular oxygen also haveparamagnetic properties.


Time (msec)

100

0

FACTORS AFFECTING TRANSVERSEMAGNETIZATION RELAXATION

Tissue Characteristics (T2)

Field Inhomogeneity

Susceptibility Effects

Imag

e Br

ight

ness

Mag

netiz

atio

n (%

)

37%T2* T2

Figure 4-11. Comparison of relaxation produced by the T2 characteristics of tissue and the T2* effects associated with magnetic field inhomogeneities.

Gadolinium has seven unpaired electronsin its orbit, which give it a very strong magneticproperty. It must be chelated to reduce its toxi-city. An example is gadolinium chelated todiethylene triamine penta-acetic acid (GaDTPA).

When a paramagnetic substance, such asgadolinium, enters an aqueous solution, itaffects the relaxation rate of the existing pro-tons. It does not produce a signal itself. In rela-tively low concentrations, the primary effect isto increase the rate of longitudinal relaxationand shorten the value of T1. In principle, thefluctuating magnetic field from the individualparamagnetic molecules enhances the relax-ation rate. The primary result is an increase insignal intensity with T1-weighted images. It isclassified as a positive contrast agent.

Signal intensity will generally increasewith the concentration of the paramagneticagents until a maximum intensity is reached.This intensity is very dependent on the imag-ing parameters. Higher concentrations willgenerally produce a reduction of signal inten-sity. This occurs because the transverse relax-ation rate is also increased, which results in ashortening of the T2 value.

Superparamagnetic MaterialsWhen materials with unpaired electrons are con-tained in a crystalline structure, they produce astronger magnetic effect (susceptibility) in com-parison with the independent molecules of aparamagnetic substance. The susceptibility ofsuperparamagnetic materials is several orders ofmagnitude greater than that of paramagneticmaterials. These materials are in the form of smallparticles. Iron oxide particles are an example.

The particles produce inhomogeneities inthe magnetic field, which results in rapid de-phasing of the protons in the transverse planeand a shortening of T2.

Superparamagnetic materials in the form oflarge particles generally reduce signal intensityand are classified as negative contrast agents.When in the form of very small particles, theyreduce T1 and increase signal intensity.

Ferromagnetic MaterialsFerromagnetic is the name applied to iron andonly a few other materials that have magneticproperties like iron. These materials have a veryhigh susceptibility and develop a high level ofmagnetism when placed in a magnetic field.



Mind Map SummaryTissue Magnetization And Relaxation

Longitudinal Transverse

Produces

Tiss

ue B

right

ness

Tiss

ue B

right

ness

Mag

netiz

atio

n

Mag

netiz

atio

n

Short Long Short Long

RF Pulse

Excitation

SaturationTime Time

67%

T1 T1

Bright TissueDark Tissue or Fluid

Dark TissueBright Tissue or Fluid

T2 T2

TISSUE MAGNETIZATION

RECOVERY or RELAXATION

T1-Weighted Image T2-Weighted Image

When tissue containing magnetic nuclei, i.e., protons, is placed in a strong magnetic field, the tis-sue becomes magnetized. It is initially magnetized in the longitudinal direction. However, byapplying a pulse of RF energy the magnetization can be flipped into the transverse plane. Bothlongitudinal and transverse magnetization have characteristics that can be used to develop imagecontrast. An imaging procedure can be adjusted to display the different types of contrasts.

When a 90˚ RF pulse is applied to longitudinal magnetization, it produces two effects. First, ittemporarily destroys the longitudinal magnetization, a condition known as saturation. It also pro-duces transverse magnetization, a condition known as excitation because transverse magnetizationis an unstable excited state.

After a saturation pulse is applied, the longitudinal magnetization will recover or regrow, aprocess known as relaxation. The rate of regrowth is a characteristic of each specific tissue and isdescribed by its T1 value, the longitudinal relaxation time. A tissue with a short T1 will recover its


magnetization fast and will appear bright in a T1-weighted image. Tissues with longer T1 values willrecover magnetization somewhat slower and will be relatively dark in T1-weighted images.

Following the production of transverse magnetization by the RF pulse the magnetization beginsto decay or relax. The rate of relaxation is a characteristic of each specific tissue and is expressedby the T2 values, the transverse relaxation time. A tissue with a short T2 will lose its transversemagnetization rapidly and will appear relatively dark in T2-weighted images. Tissues and body flu-ids with long T2 values will retain their transverse magnetization longer and will appear bright inT2-weighted images.

5The Imaging Process

49

Introduction And OverviewThe MR imaging process consists of two majorfunctions as shown in Figure 5-1. The first is theacquisition of RF signals from the patient’s bodyand the second is the mathematical reconstruc-tion of an image from the acquired signals.

In this chapter we will develop a generaloverview of the imaging process and set thestage for considering the different methodsand techniques that are used to produce opti-mum images for various clinical needs.

k SpaceDuring the acquisition process the signals arecollected, digitized, and stored in computermemory in a configuration known as k space.

The k space is divided into lines of data that arefilled one at a time. One of the general require-ments is that the k space must be completelyfilled before the image reconstruction can becompleted. The size of k space (number of lines)is determined by the requirements for imagedetail and will be discussed in Chapters 9 thru 11.

Acquisition

The acquisition process consists of an imagingcycle that is repeated many times. The timerequired for a complete acquisition is deter-mined by the duration of the cycle multipliedby the number of cycles. The duration of a cycleis TR (Time of Repetition), the adjustable proto-col factor that is used to select the different

types of image contrast. Also, the number ofcycles used in an acquisition is adjustable. Thenumber of cycles depends on the quality of theimage that is required. The complete relation-ship between number of imaging cycles andimage quality characteristics is described inChapter 10.

ReconstructionThe image reconstruction process is usuallyfast compared to the acquisition process andgenerally does not require any decisions oradjustments by the operator.

Imaging ProtocolEach imaging procedure is controlled by a pro-tocol that has been entered into the computer.Issues that must be considered in selecting,modifying, or developing a protocol for a spe-cific clinical procedure include:

∑The imaging method to be used∑The image types (PD, T1, T2, etc.)∑Spatial characteristics (slice thickness,

number, etc.)∑Detail and visual noise requirements∑Use of selective signal suppression

techniques∑Use of artifact reduction techniques

In the following chapters we will addresseach of these issues and the specific protocolfactors that are used to produce the desiredimage characteristics.

Imaging MethodsThere are several different imaging methodsthat can be used to create MR images. The prin-cipal difference among these methods is thesequence in which the RF pulses and gradientsare applied during the acquisition process.


TR TR TR TR TR TR TR TR TR TR TR TR

THE MR IMAGING PROCESS

Acquisition Reconstruction

Image

Row

s

“k” Space

Imaging CyclesFinishStart

Figure 5-1. The two functions, acquisition and reconstruction, that make up the MR image production process.

Therefore, the different methods are often re-ferred to as the different pulse sequences. Anoverview of the most common methods isshown in Figure 5-2. As we see, the differentmethods are organized in a hierarchy structure.For each imaging method there is a set of fac-tors that must be adjusted by the user to pro-duce specific image characteristics.

The selection of a specific imaging methodand factor values is generally based on require-ments for contrast sensitivity to a specific tis-sue characteristic (PD, T1, T2) and acquisitionspeed. However, other characteristics such asvisual noise and the sensitivity to specific arti-facts might vary from method to method.

All of the imaging methods belong to oneor both of the two major families, spin echo orgradient echo. The difference between the two

families of methods is the process that is usedto create the echo event at the end of eachimaging cycle. For the spin echo methods, theecho event is produced by the application of a180° RF pulse, as will be described in Chapter 6.For the gradient echo methods the event is pro-duced by applying a magnetic field gradient, asdescribed in Chapter 7. Each method has veryspecific characteristics and applications.

The Imaging Cycle A common characteristic of all methods is thatthere are two distinct phases of the image acqui-sition cycle, as shown in Figure 5-3. One phase is associated with longitudinal magnetizationand the other with transverse magnetization. Ingeneral, T1 contrast is developed during the lon-gitudinal magnetization phase and T2 contrast is

THE IMAGING PROCESS 51

TE

TE

TI

TI

TR

TR

TS

TE

TR

FLIPANGLE

IMAGING METHODS

Spin Echo

Spin Echo

Inversion Recovery

GRandSE

Echo Planar

MagneticPreparation

Pulses

Small AngleGradient Echo

Gradient Echo

Figure 5-2. The principal spin echo and gradient echo imaging methods. GRandSE, or GRASE, is a combination of the two methods.

developed during the transverse magnetizationphase. PD contrast is always present, but becomesmost visible when it is not overshadowed byeither T1 or T2 contrast. The predominant typeof contrast that ultimately appears in the imageis determined by the duration of the two phasesand the transfer of contrast from the longitudi-nal phase to the transverse phase.

The duration of the two phases (longitudi-nal and transverse) is determined by the selectedvalues of the protocol factors, TR (Time of Repe-tition) and TE (Time to Echo).

TRTR is the time interval between the beginningof the longitudinal relaxation, following satu-ration, and the time at which the longitudinal

magnetization is converted to transverse mag-netization by the excitation pulse. This iswhen the picture is snapped relative to the lon-gitudinal magnetization.

Because the longitudinal relaxation takes arelatively long time, TR is also the duration ofthe image acquisition cycle or the cycle repe-tition time (Time of Repetition).

TETE is the time interval between the beginningof transverse relaxation following excitationand when the magnetization is measured toproduce image contrast. This happens at theecho event and is when the picture is snappedrelative to the transverse magnetization.Therefore, TE is the Time to Echo event.


Time (msec)

100

0

T1 C

ON

TRAS

T

T2CONTRAST

PD CONTRAST

Shor

t T1

Short T2

Long T2

Long

T1

Low PD

High PD

TR TE

Imag

e Br

ight

ness

Mag

netiz

atio

n (%

)

EchoEvent

RF RF

ONE IMAGING CYCLE


TransverseMagnetization

0 800 1600 100 200

Figure 5-3. The longitudinal and transverse magnetization phases of an imaging cycle. T1 and PD contrast are produced during the longitudinal phase and

T2 contrast is produced during the transverse phase.

Excitation

The transition from the longitudinal magneti-zation phase to the transverse magnetizationphase is produced by applying an RF pulse.This is generally known as the excitationprocess because the transverse magnetizationrepresents a more unstable or “excited” statethan longitudinal magnetization.

The excitation pulse is characterized by aflip angle. A 90˚ excitation pulse converts all ofthe existing longitudinal magnetization intotransverse magnetization. This type of pulse isused in the spin echo methods. However, thereare methods that use excitation pulses withflip angles that are less than 90 .̊ Small flipangles (<90˚) convert only a fraction of theexisting longitudinal magnetization into trans-verse magnetization and are used primarily toreduce acquisition time with the gradient echomethods described in Chapter 7.

The Echo Event and Signals

The transverse magnetization phase termi-nates with the echo event, which produces theRF signal. This is the signal that is emitted by thetissue and used to form the image. The echoevent is produced by applying either an RFpulse or a gradient pulse to the tissue, as willbe described in Chapters 6 and 7.

Contrast Sensitivity In MRI the usual procedure is to select one ofthe tissue characteristics (PD, T1, T2) andthen adjust the imaging process so that it hasmaximum, or at least adequate, contrast sen-sitivity for that specific characteristic. Thisproduces an image that is heavily weighted bythat characteristic. The contrast sensitivity ofthe imaging process and the resulting imagecontrast is determined by the specific imag-ing method and the combination of imaging

protocol factor values, which we will considerin much more detail in later chapters. Thediscussion in this chapter will be based on theconventional spin echo method that usesonly two factors, TR and TE, to control con-trast sensitivity. However, it establishes someprinciples that apply to all methods.

T1 Contrast

During the relaxation (regrowth) of longitudinalmagnetization, different tissues will have differ-ent levels of magnetization because of their dif-ferent growth rates, or T1 values. Figure 5-4 com-pares two tissues with different T1 values.

The tissue with the shorter T1 value expe-riences a faster regrowth of longitudinal mag-netization. Therefore, during this period oftime it will have a higher level of magnetiza-tion, produce a more intense signal, andappear brighter in the image. In T1-weightedimages brightness or high signal intensity isassociated with short T1 values.

At the beginning of each imaging cycle,the longitudinal magnetization is reduced tozero (saturation) by an RF pulse, and thenallowed to regrow, or relax. This is what hap-pens in the spin echo method. In some otherimaging methods, as we will see in the nexttwo chapters, the cycle might begin with eitherpartially saturated or inverted longitudinalmagnetization. In all cases, T1 contrast isformed during the regrowth process. At a timedetermined by the selected TR value, the cycleis terminated and the magnetization value isconverted to transverse, measured and dis-played as a pixel intensity, or brightness, anda T1-weighted image is produced.

In principle, at the beginning of eachimaging cycle all tissues are dark. As the tissuesregain longitudinal magnetization, they becomebrighter. The brightness, or intensity, withwhich they appear in the image depends on


when during the regrowth process the cycle isterminated and the picture is snapped. This isdetermined by the selected TR value. When ashort TR is used, the regrowth of the longitu-dinal magnetization is interrupted before itreaches its maximum. This reduces signalintensity and tissue brightness within theimage but produces T1 contrast.

Increasing TR increases signal intensity andbrightness up to the point at which magneti-zation is fully recovered, which is determinedby the PD of each tissue. For practical purposes,this occurs when the TR exceeds approximatelythree times the T1 value for the specific tissues.Although it takes many cycles to form a com-plete image, the longitudinal magnetization isalways measured at the same time in each cycleas determined by the setting of TR.

To produce a T1-weighted image, a value forTR must be selected to correspond with the timeat which T1 contrast is significant between thetwo tissues. Several factors must be considered inselecting TR. If T1 contrast is represented by theratio of the tissue magnetization levels, it is at itsmaximum very early in the relaxation process.However, the low magnetization levels presentat that time do not generally produce adequateRF signal levels for many clinical applications.The selection of a longer TR produces greater sig-nal strength but less T1 contrast.

The selection of TR must be appropriate forthe T1 values of the tissues being imaged. If aTR value is selected that is equal to the T1 valueof a tissue, the picture will be snapped whenthe tissue has regained 63% of its magnetiza-tion. This represents the time when there is


TR

T1Contrast

RFPULSE


100

0

RFPULSE

67%

Imag

e Br

ight

ness

Shor

t T1

Long T1

T1 CONTRASTLongitudinal Phase

Mag

netiz

atio

n (%

)

Figure 5-4. The amount of T1 contrast captured during the longitudinal magnetization phase isdetermined by the value of TR that is selected by the operator.

maximum contrast between tissues with smalldifferences in T1 values.

Proton Density (PD) ContrastThe density, or concentration, of protons ineach tissue voxel determines the maximumlevel of magnetization that can be obtained.Differences in PD among tissues can be used to produce image contrast, as illustrated in Fig-ure 5-5. Here we see the growth of longitudi-nal magnetization for two tissues with thesame T1 values but different relative PDs. Thetissue with the lowest PD (80) reaches a maxi-mum magnetization level that is only 80%that of the other tissue. The difference in mag-netization levels at any point in time is because

of the difference in PD and is therefore thesource of PD contrast.

Although there is some PD contrast earlyin the cycle, it is generally quite small in com-parison to the T1 contrast.

The basic difference between T1 contrastand PD contrast is that T1 contrast is producedby the rate of growth (relaxation), and PD con-trast is produced by the maximum level towhich the magnetization grows. In general, T1contrast predominates in the early part of therelaxation phase, and PD contrast predom-inates in the later portion. T1 contrast gradu-ally gives way to PD contrast as magnetizationapproaches the maximum value. A PD-weightedimage is produced by selecting a relatively long


TR

PDContrast

Low PD

High PD

RFPULSE

RFPULSE


T1Contrast

Imag

e Br

ight

ness

PROTON DENSITY CONTRAST

Longitudinal Phase

Figure 5-5. Proton density (PD) contrast is captured by setting TR to relatively long values. At thattime the magnetization is determined by PD and is not T1 as in the earlier part of the cycle.

TR value so that the image is created or “thepicture is snapped” in the later portion of therelaxation phase, where tissue magnetizationsapproach their maximum values. The TR valuesat which this occurs depend on the T1 valuesof the tissues being imaged.

It was shown earlier that tissue reaches95% of its magnetization in three T1s. There-fore, a TR value that is at least three times theT1 values for the tissues being imaged pro-duces almost pure PD contrast.

T2 ContrastNow let us turn our attention to the transversephase. During the decay of transverse magne-tization, different tissues will have different

levels of magnetization because of differentdecay rates, or T2 values. As shown in Figure 5-6, tissue with a relatively long T2 value willhave a higher level of magnetization, producea more intense signal, and appear brighter inthe image than a tissue with a shorter T2 value.

Figure 5-6 shows the decay of transversemagnetization for tissues with different T2 val-ues. The tissue with the shortest T2 value losesits magnetization faster than the other tissues.

The difference in T2 values of tissue can betranslated into image contrast. For the purposeof this illustration we assume that the two tis-sues begin their transverse relaxation with thelevels of magnetization determined by the PD.This is the usual case where the PD contrast


EchoEvent

TE

T2Contrast

RFPULSE

Excitation

PDContrast

FromLongitudinal

PhaseIm

age

Brig

htne

ss

T2 CONTRAST

TransversePhase

Short T2

Long T2

Figure 5-6. The formation of T2 contrast during the transverse magnetization phase. The amount of T2 contrast captured depends on the selected value of TE, the Time to Echo event.

present at the end of the longitudinal phasecarries over to the beginning of the transversephase. In effect, the transverse phase beginswith PD contrast but adds T2 contrast as timeelapses. The decay of the magnetization pro-ceeds at different rates because of the differentT2 values. The tissue with the longer T2 valuemaintains a higher level of magnetization thanthe other tissue and will remain bright longer.The difference in the tissue magnetizations atany point in time represents contrast.

At the beginning of the cycle there is no T2contrast, but it develops and increases through-out the relaxation process. At the echo eventthe magnetization levels are converted into RFsignals that are displayed as image pixel bright-ness; this is the time to echo event (TE) and isselected by the operator. Maximum T2 contrastis generally obtained by using a relatively longTE. However, when a very long TE value is used,

the magnetization and the RF signals might betoo low to form a useful image. In selecting TEvalues, a compromise must often be madebetween T2 contrast and good signal intensity.

The transverse magnetization characteris-tics of tissue (T2 values) are, in principle,added to the longitudinal characteristics car-ried over from the longitudinal phase (e.g., T1and PD) to form the MR image. Usually we donot want to add T2 contrast to T1 contrast.That is because these two types of contrastoppose each other. Remember in Chapter 1 wesaw that tissues that are bright in T1 imagesare dark in T2 images. This means that if wewere to mix T1 and T2 contrast in the sameimage, one would cancel the other. When set-ting up a protocol for a T2 image it is neces-sary to use a long TR (in addition to a long TE)so that no, or very little, T1 contrast carriesover to the transverse phase.


Mind Map SummaryThe Imaging Process


“k”Space Image

Reconstruction

Longitudinal Magnetization Phase Transverse Magnetization Phase

Ro

ws

TR TR TR TR

T1 CONTRAST PD CONTRAST T2 CONTRAST

180˚RF

GR A DIE NT

TR TE

Brig

htne

ss

Uses

Spin Echo

EchoEvent

Gradient Echo

Imaging Methods

To CreateUses To Create

The MR imaging process is one of creating contrast among tissues based on their magneticcharacteristics. The primary characteristics are proton density (PD), T1, and T2. It is a dynamicactivity in which the magnetization levels of the various tissues are undergoing almost constantchange. During each imaging cycle there are two distinct magnetization phases: longitudinal andtransverse. Different types of contrast are developed in each of these phases.

After application of a saturation pulse, which reduces the longitudinal magnetization to zero,the magnetization begins to regrow, a process known as relaxation. The rate of regrowth for a spe-cific tissue is determined by that tissue’s T1 value. Tissues with short T1 values grow faster thantissues with long T1 values. During this regrowth, T1 contrast in the form of different levels ofmagnetization is created among the tissues. This is the contrast that will be displayed in an imageif the protocol parameters are set to produce a T1-weighted image. In a T1-weighted image, tissueswith short T1 values will be bright. Tissues and fluid with long T1 values will be darker.

When an RF pulse is applied to longitudinal magnetization, it converts (flips) it to transversemagnetization, an unstable excited magnetic condition that decays with time. This decay processis the transverse magnetization relaxation process. The rate of decay of a specific tissue dependson that tissue’s T2 value. Tissues with short T2 values decay faster than tissues with longer T2 val-ues. When the imaging protocol factors are set to produce a T2-weighted image, tissues with shortT2 values will be dark and tissues and fluids with longer T2 values will be bright.

6Spin Echo

Imaging Methods

59

Introduction And OverviewSpin echo is the name of the process that usesan RF pulse to produce the echo event. It is alsothe name for one of the specific imaging meth-ods within the spin echo family of imagingmethods; all of which use the spin echoprocess. We will first discuss the spin echoprocess and see how an RF pulse can producean echo event and signal and then consider thespin echo methods.

The Spin Echo ProcessThe decay of transverse magnetization (i.e.,relaxation) occurs because of dephasing amongindividual nuclei, as described in Chapter 4.

Let us recall that two basic conditions arerequired for transverse magnetization: (1) themagnetic moments of the nuclei must be ori-ented in the transverse direction, or plane, and(2) a majority of the moments must be in thesame direction within the transverse plane.When a nucleus has a transverse orientation, itis actually precessing or rotating around an axisthat is parallel to the magnetic field.

After the application of a 90˚ excitationpulse, the nuclei have a transverse orientationand are precessing together, or in-phase, aroundthe magnetic field axis. This is the normal pre-cession discussed earlier but flipped into thetransverse plane. However, within an individualvoxel some nuclei precess or spin faster thanothers. After a short period of time, the nuclei

are not spinning in-phase. As the directions ofthe nuclei begin to spread, the magnetizationof the tissue decreases. A short time later, thenuclei are randomly oriented in the transverseplane; there is no transverse magnetization.

The two factors that contribute to the de-phasing of the nuclei and the resulting trans-verse relaxation will now be reviewed againhere. One is an exchange among the spinningnuclei (spin-spin interactions), which results inrelatively slow dephasing and loss of magneti-zation. The rate at which this occurs is deter-mined by characteristics of the tissue. It is thisdephasing activity that is characterized by theT2 values and the source of contrast that wewant to capture in T2 images. A second factor,which produces relatively rapid dephasing ofthe nuclei and loss of transverse magnetization,is the inhomogeneity of the magnetic field.Even within a small volume of tissue, the fieldinhomogeneities are sufficient to produce rapiddephasing. This effect, which is generally unre-lated to the T2 characteristics of the tissue,tends to mask the true relaxation characteris-tics of the tissue. In other words, the actualtransverse magnetization relaxes much fasterthan the tissue characteristics would indicate.We remember that this real relaxation time isdesignated as T2*. The value of T2* is alwaysmuch less than the tissue T2 value. As a result,the transverse magnetization disappears beforeT2 contrast can be formed.

We are about to discover that spin echo is aprocess for recovering the lost transverse mag-netization and making it possible to produceimages of the three tissue characteristics,including T2.

An RF signal is produced whenever there is transverse magnetization. Immediately afteran excitation pulse, a so-called free inductiondecay (FID) signal is produced. The intensityof this signal is proportional to the level of

transverse magnetization. Both decay ratherrapidly because of the magnetic field inho-mogeneities just described. The FID signal isnot used in the spin echo methods. It is usedin the gradient echo methods to be describedin Chapter 7.

The spin echo process is used to compen-sate for the dephasing and rapid relaxationcaused by the field inhomogeneities and torestore the magnetization to the level thatdepends only on the tissue T2 characteristics.The sequence of events in the spin echo processis illustrated in Figure 6-1.

Transverse magnetization is produced witha 90˚ RF excitation pulse that flips the longitu-dinal magnetization into the transverse plane.Immediately following the RF pulse, each voxelis magnetized in the transverse direction. How-ever, because of the local magnetic field in-homogeneities within each voxel, the protonsprecess at different rates and quickly slip out ofphase. This produces the rapid decay charac-terized by T2* and the associated FID signal. Atthis time the protons are still rotating in thetransverse plane, but they are out of phase.

If a 180˚ pulse is applied to the tissue con-taining these protons, it flips the protonsaround an axis in the transverse plane; thisreverses their direction of rotation as illustratedin Figure 6-2. This causes the fast protons to belocated behind the slower ones. As the fasterprotons begin to catch up with the slower ones,they regain a common alignment, or comeback into phase. This, in turn, causes the trans-verse magnetization to reappear and form theecho event. However, the magnetization doesnot grow to the initial value because the relax-ation (dephasing) produced by the tissue isnot reversible. The rephasing of the protonscauses the magnetization to build up to a leveldetermined by the T2 characteristics of the tis-sue. As soon as the magnetization reaches this


maximum, the protons begin to move out ofphase again, and the transverse magnetizationdissipates. Another 180˚ pulse can be used toproduce another rephasing. In fact, this is whatis done in multi-echo imaging and will bedescribed later in this chapter.

RF Pulse Sequence

The different imaging methods are produced bythe type (flip angle) and time intervals betweenthe applied RF pulses. The basic pulse sequencefor the spin echo method is shown in Figure 6-3. Each cycle begins with a 90˚ excitation pulsethat produces the initial transverse magnetiza-tion and a later 180˚ pulse that rephases theprotons to produce the echo event.

The time between the initial excitation and the echo signal is TE. This is controlled by

adjusting the time interval between the 90˚and the 180˚ pulses, which is 1/2 TE.

The Spin Echo MethodThis method can be used to produce images of thethree basic tissue characteristics: PD, T1, and T2.The sensitivity to a specific characteristic isdetermined by the values selected for the twotime intervals or imaging factors, TR and TE.

The process of creating images with thethree types of contrast (PD, T1, and T2)described in the last chapter was a descriptionof the spin echo method. There we saw that thetype of image that was produced depended onthe values selected for the two protocol factors,TR and TE. We will now review that processwith a few more details specifically as it appliesto the spin echo method.

SPIN ECHO IMAGING METHODS 61

100

0RF

PULSE

180°RF

PULSE

90°

Signal

EchoEvent

T2*

Tissue T2 Relaxation

Dephasing Rephasing

Rephasing

TE

THE SPIN ECHO PROCESS

Excitation

Imag

e Br

ight

ness

Mag

netiz

atio

n (%

)

Figure 6-1. The spin echo process showing the use of a 180˚ pulse to rephase the protons and to produce an echo event.

Proton Density (PD) ContrastPD contrast develops as the longitudinal mag-netization approaches its maximum, which isdetermined by the PD of each specific tissue.Therefore, relatively long TR values arerequired to produce a PD-weighted image.Short TE values are generally used to reduce T2contrast contamination and to maintain a rel-atively high signal intensity.

T1 ContrastTo produce image contrast based on T1 dif-ferences between tissues, two factors must be considered. Since T1 contrast develops dur-ing the early growth phase of longitudinal

magnetization, relatively short TR values mustbe used to capture the contrast. The secondfactor is to preserve the T1 contrast during thetime of transverse relaxation. The basic problemis that if T2 contrast is allowed to develop, itgenerally counteracts T1 contrast. This isbecause tissues with short T1 values usuallyhave short T2 values. The problem arisesbecause tissues with short T1s are generallybright, whereas tissues with short T2s havereduced brightness when T2 contrast is pres-ent. T2 contrast develops during the TE timeinterval. Therefore, a T1-weighted image isproduced by using short TR values and shortTE values.


RFPULSE

180˚SpinDirection

SpinDirection

Dephasing Rephasing

THE SPIN ECHO PROCESS

Figure 6-2. The 180˚ pulse sets up the protons so that they rephase.

T2 Contrast

The first step in producing an image with sig-nificant T2 contrast is to select a relatively longTR value. This minimizes T1 contrast contam-ination and the transverse relaxation processbegins at a relatively high level of magnetiza-tion. Long TE values are then used to allow T2contrast time to develop.

The spin echo method is the only methodthat produces true T2 contrast. That is becauseit is able to rephase the protons and remove theT2* effect.

Multiple Spin Echo

It is possible to produce a series of echo eventswithin one cycle as illustrated in Figure 6-4. Thisis done by applying several 180˚ pulses aftereach 90˚ excitation pulse. The advantage is thatecho events with different TE values are pro-duced in one acquisition cycle. Separate imagesare formed for each TE value. This makes it pos-sible to create both a PD image (short TE) and aT2 image (long T2) in the same acquisition.

Table 6-1 summarizes the combination ofTR and TE values used to produce the three


1/2 TE

TE

One Cycle Next Cycle

TR

SPIN ECHO TIME INTERVALS

FID Signal Echo Signal

180˚RF

Pulse

90˚RF

Pulse

Figure 6-3. The RF pulses and time intervals in a spin echo imaging cycle.

basic image types with the spin echo method.Optimum values of TR and TE for a specificprotocol might vary because of considerationsfor other factors such as image acquisitiontime, number of slices, etc.

Inversion RecoveryInversion recovery is a spin echo imagingmethod used for several specific purposes. Oneapplication is to produce a high level of T1 con-trast and a second application is to suppress thesignals and resulting brightness of fat and flu-ids. The inversion recovery pulse sequence isobtained by adding an additional 180˚ pulse to

the conventional spin echo sequence, as shownin Figure 6-5. The pulse is added at the begin-ning of each cycle where it is applied to thelongitudinal magnetization carried over fromthe previous cycle. Each cycle begins as the180˚ pulse inverts the direction of the lon-gitudinal magnetization. The regrowth (recov-ery) of the magnetization starts from a negative(inverted) value, rather than from zero, as inthe spin echo method.

The inversion recovery method, like thespin echo method, uses a 90˚ excitation pulse toproduce transverse magnetization and a final180˚ pulse to produce a spin echo signal. That is


TE

TE 1 2 0

2 0

180˚90˚ 180˚EchoEvent

EchoEvent

MULTIPLE ECHO IMAGING

PD Image

1st Echo Signal 2nd Echo Signal

T2-Weighted Image

Figure 6-4. A multiple spin echo imaging that produces both a PD and T2 image in the same acquisition.

why it is classified as one of the spin echo, ratherthan gradient echo, methods. An additionaltime interval is associated with the inversionrecovery pulse sequence. The time between theinitial 180˚ pulse and the 90˚ pulse is designatedthe Time after Inversion (TI). It can be varied bythe operator and used as a contrast control.

T1 ContrastThe principal characteristic of many inversionrecovery images is high T1 contrast. This occursbecause the total longitudinal relaxation timeis increased because it starts from the invertedstate. There is more time for the T1 contrast to


Table 6-1. Selection of TR and TE values to produce the three image types with spin echo method.Values shown are typical but can be varied to some extent to accommodate specificimaging conditions.

T1 Image PD Image T2 ImageTR Short Long Long

(500 msec) (2000 msec) (2000 msec)

TE Short Short Long(15-20 msec) (15-20 msec) (120 msec)

TI

INVERSION RECOVERY

Spin EchoLong T1

90˚RF

Pulse

180˚RF

Pulse

180˚RF

Pulse

Short T1

Long

itudi

nal M

agne

tizat

ion

Imag

e Br

ight

ness

T1 Contrast

Figure 6-5. The inversion recovery method with TI set to produce an image with high T1 contrast.

develop. A T1 image produced by the inversionrecovery method is compared to one producedby the spin echo method in Figure 6-6. Notice

the significant difference in contrast. The useof the inversion method for other applicationswill be discussed in Chapter 8.


TR TR

TI

TE TE

560

14

1800

25

400

T1 IMAGESSpin Echo Inversion Recovery

Figure 6-6. Comparison of T1 images produced by spin echo and inversion recovery methods.

Mind Map SummarySpin Echo Imaging Methods


DephasingDephasing

ExcitationPulse

RephasingRules

EchoEvent

EchoEvent

90˚ 180˚ 180˚

Tran

sver

se M

agne

tizat

ion

Brig

htne

ses

RephasingRephasing

T 2 T IS S U E CHA R ACT E R IS T IC

1st

2nd

Field Inhomogeneitiesand

Susceptibility Effects

Produce

TE

T2*

Spin echo is a technique used to produce an echo event by applying a 180˚ RF pulse to thedephased transverse magnetization. This compensates for the dephasing produced by field inho-mogeneities and makes it possible to produce images that show the T2 characteristics of tissue. Thetime to the echo event, TE, is a protocol factor that can be adjusted to produce different weight-ings to the T2 contrast. When a short TE value is selected, the T2 effect is reduced, and the result-ing image will be either a PD or T1-weighted image, depending on the selected TR value.

It is possible to use a series of 180˚ RF pulses within one cycle to produce multiple echoevents, each with a different TE value. Both PD and T2-weighted images can be acquired in thesame acquisition.

There is actually a family of spin echo methods that all use the spin echo process to create theecho event. These include the spin echo and inversion recovery methods as well as the GRASEmethod that uses both spin echoes and gradient echoes in the same acquisition.

7Gradient Echo

Imaging Methods

69

Introduction And OverviewIt is possible to produce an echo event byapplying a magnetic field gradient without a180˚ RF pulse to the tissue as in the spin echomethods. There are several imaging methodsthat use the gradient echo technique to pro-duce the RF signals and these make up the gra-dient echo family of methods.

The primary advantage of the gradientecho methods over the spin echo methods isthat gradient echo methods perform fasterimage acquisitions. Gradient echo methods aregenerally considered to be among the fasterimaging methods. They are also used in someof the angiographic applications because gra-dient echo generally produces bright blood, as

we will see in Chapter 12, as well as for func-tional imaging, as described in Chapter 13. Onelimitation of the gradient echo methods is theydo not produce good T2-weighted images, aswill be described later in this chapter. However,by combining the gradient and spin echomethods, this limitation can be overcome.

At this time we will develop the concept ofgradient echo and then consider the specificgradient echo imaging methods and theircharacteristics.

The Gradient Echo ProcessTransverse magnetization is present only whena sufficient quantity of protons are spinningin-phase in the transverse plane. As we have

seen, the decay (relaxation) of transverse mag-netization is the result of proton dephasing.We also recall that an RF signal is being pro-duced any time there is transverse magnetiza-tion and the intensity of the signal is propor-tional to the level of magnetization.

With the spin echo technique we use an RFpulse to rephase the protons after they havebeen dephased by inherent magnetic field inho-mogeneities and susceptibility effects withinthe tissue voxel. With the gradient echo tech-nique the protons are first dephased, on pur-pose, by turning on a gradient and thenrephased by reversing the direction of the gra-dient, as shown in Figure 7-1. A gradient echocan only be created when transverse magneti-zation is present. This can be either during the

free induction decay (FID) period or during aspin echo event. In Figure 7-1 the gradient echois being created during the FID. Let us now con-sider the process in more detail.

First, transverse magnetization is producedby the excitation pulse. It immediately beginsto decay (the FID process) because of the mag-netic field inhomogeneities within each indi-vidual voxel. The rate of decay is related to thevalue of T2*. A short time after the excitationpulse a gradient is applied, which produces avery rapid dephasing of the protons and reduc-tion in the transverse magnetization. Thisoccurs because a gradient is a forced inhomo-geneity in the magnetic field. The next step isto reverse the direction of the applied gradient.Even though this is still an inhomogeneity in


TE

GRADIENT ECHO

Gradient

Signal

Dephasing

DephasingRephasing

T2* Decay (Field Effect)

Tran

sver

se M

agne

tizat

ion

Excitation<90˚

RFPULSE

Figure 7-1. The gradient echo process using a magnetic field gradient to produce an echo event during the FID.

the magnetic field, it is in the opposite direc-tion. This then causes the protons to rephaseand produce an echo event. As the protonsrephase, the transverse magnetization will reap-pear and rise to a value determined by the FIDprocess. The gradient echo event is a rather well-defined peak in the transverse magnetizationand this, in turn, produces a discrete RF signal.

The TE is determined by adjusting the timeinterval between the excitation pulse and thegradients that produce the echo event. TE val-ues for gradient echo are typically muchshorter than for spin echo, especially when thegradient echo is produced during the FID.

Small Angle Gradient Echo MethodsThe gradient echo technique is generally usedin combination with an RF excitation pulsethat has a small flip angle of less than 90˚. Wewill discover that the advantage of this is thatit permits the use of shorter TR values and this,in turn, produces faster image acquisition.

One source of confusion is that each man-ufacturer of MRI equipment has given his gra-dient echo imaging methods different tradenames. In this text we will use the generic nameof small angle gradient echo (SAGE) method.

The SAGE method generally requires ashorter acquisition time than the spin echomethods. It is also a more complex methodwith respect to adjusting contrast sensitivitybecause the flip angle of the excitation pulsebecomes one of the adjustable protocol imag-ing factors.

Excitation/Saturation-Pulse Flip Angle

We recall that the purpose of the excitation/saturation pulse applied at the beginning of animaging cycle is to convert or flip longitudinal

magnetization into transverse magnetization.When a 90˚ pulse is used, all of the existing lon-gitudinal magnetization is converted into trans-verse magnetization, as we have seen with thespin echo methods. The 90˚ pulse reduces thelongitudinal magnetization to zero (i.e., com-plete saturation) at the beginning of each imag-ing cycle. This then means that a relatively longTR interval must be used to allow the longitudi-nal magnetization to recover to a useful value.The time required for the longitudinal magneti-zation to relax or to recover is one of the majorfactors in determining acquisition time. Theeffect of reducing TR when 90˚ pulses are used isshown in Figure 7-2. As the TR value is decreased,the longitudinal magnetization grows to a lowervalue and the amount of transverse magnetiza-tion and RF signal intensity produced by eachpulse is decreased. The reduced signal intensityresults in an increase in image noise as describedin Chapter 10. Also, the use of short TR intervalswith a 90˚ pulse (as in spin echo) cannot producegood PD or T2-weighted images.

One approach to reducing TR and increas-ing acquisition speed without incurring the dis-advantages that have just been described is touse a pulse that has a flip angle of less than 90˚.A small flip-angle (<90˚) pulse converts only a

GRADIENT ECHO IMAGING METHODS 71

TR

TRShort

LongHigh

Low

Long

itudi

nal M

agne

tizat

ion

Sign

al In

tens

ity

Figure 7-2. The effect of reducing TR on therecovery of longitudinal magnetization within

a cycle and the resulting signal intensity when using 90˚ pulses.

fraction of the longitudinal magnetization intotransverse magnetization. This means that thelongitudinal magnetization is not completelydestroyed or reduced to zero (saturated) by thepulse, as shown in Figure 7-3.

Reducing the flip angle has two effects thatmust be considered together. The effect that wehave just observed is that the longitudinalmagnetization is not completely destroyed andremains at a relatively high level from cycle tocycle, even for short TR intervals. This willincrease RF signal intensity compared to theuse of 90˚ pulses. However, as the flip angle is

reduced, a smaller fraction of the longitudinalmagnetization is converted into transversemagnetization. This has the effect of reducingsignal intensity. The result is a combination ofthese two effects. This is illustrated in Figure 7-4. Here we see that as the flip angle is increasedover the range from 0–90˚, the level of longitu-dinal magnetization at the beginning of a cycledecreases. On the other hand, as the angle isincreased, the fraction of this longitudinal mag-netization that is converted into transversemagnetization increases and RF signal inten-sity increases. The combination of these two


TRTR TR

FLIPANGLE

LONGITUDINAL MAGNETIZATION

Small Flip Angle Pulses

Figure 7-3. The effect of using small flip-angle pulses on longitudinal magnetization.

effects is shown in Figure 7-5. Here we see howchanging flip angle affects signal intensity. Theexact shape of this curve depends on the spe-cific T1 value of the tissue and the TR interval.For each T1/TR combination there is a differ-ent curve and a specific flip angle that pro-duces maximum signal intensity.

Let us now use Figure 7-6 to compare themagnetization of two tissues with differentT1 values as we change flip angle. Contrastbetween the two tissues is represented by thedifference in magnetization levels. At this pointwe are assuming a short TE and considering thecontrast associated with only the longitudinalmagnetization. The flip-angle range is dividedinto several specific segments as shown.


With the SAGE method the contrast sensitivityfor a specific tissue characteristic is controlledby three protocol factors. As with spin echo, TRand TE have an effect. However, the flip anglebecomes the factor with the greatest effect oncontrast. We will now see how changing flipangle can be used to select specific types of con-trast with a basic gradient echo method.

T1 Contrast

Relatively large flip angles (45˚–90˚) produce T1contrast. This is what we would expect becauselarge flip angles (close to 90˚) and short TR andTE values are similar to the factors used to


FLIPANGLE

EFFECT OF FLIP ANGLEON MAGNETIZATION

(SHORT TR)

Longitudinal

Rela

tive

Mag

netiz

atio

n

Transverse

0 15 30 45 60 75 90

Figure 7-4. The effect of pulse flip angle on the level of both longitudinal and transverse magnetization after the pulse is applied.


FLIPANGLE

EFFECT OF FLIP ANGLEON CONTRAST

Rela

tive

Mag

netiz

atio

n

Sign

al In

tens

ity

0 15 30 45 60 75 90

PDConstrast

T1Constrast

Long T1

Short T1

FLIPANGLE

EFFECT OF FLIP ANGLEON SIGNAL INTENSITY

Longitudinal

Transverse

Rela

tive

Mag

netiz

atio

n

Sign

al In

tens

ity

0 15 30 45 60 75 90

Figure 7-5. The relationship of signal intensity to flip angle.

Figure 7-6. The effect of flip angle on contrast.

produce T1 contrast with the spin echo method.Here, with gradient echo, we observe a loss ofT1 contrast as the flip angle is decreased signif-icantly from 90˚.

Low Contrast

There is an intermediate range of flip-anglevalues that produces very little, if any, contrast.This is the region in which the PD and T1 con-trast cancel each other for many tissues, such asgray and white matter.

Proton Density (PD) Contrast

Relatively low flip-angle values produce PDcontrast. As the flip angle is reduced within thisregion, there is a significant decrease in mag-netization and the resulting signal intensity.

Up to this point we have observed gener-ally how changing the flip angle of the excita-tion pulse affects signal intensity and contrast.In the SAGE imaging method the flip angle isone of the imaging factors that must beadjusted by the user. However, it becomessomewhat complex because the specific effectof flip angle is modified by the other imagingfactors and techniques used to enhance a spe-cific type of contrast.

T2 and T2* Contrast

We recall that T2 contrast is produced by thedecay characteristics of transverse magnetiza-tion and that there are two different decayrates, T2 and T2*. The slower decay rate is deter-mined by the T2 characteristics of the tissue.The faster decay is produced by small inhomo-geneities within the magnetic field oftenrelated to variations in tissue susceptibility dif-ferences. This decay rate is determined by theT2* of the tissue environment. When a spinecho technique is used, the spinning protonsare rephased, and the T2* effect is essentially

eliminated. However, when a spin echo tech-nique is not used, the transverse magnetizationdepends on the T2* characteristics. The gradi-ent echo technique does not compensate forthe inhomogeneity and susceptibility effectdephasing as the spin echo technique does.Also, without using a spin echo process thelong TE values necessary to produce T2 contrastcannot be achieved. Therefore, a basic gradientecho imaging method is not capable of pro-ducing true T2 contrast. The contrast will bedetermined primarily by the T2* characteristics.The amount of T2* contrast in an image isdetermined by the selected TE value. In general,longer TE values (but short compared to thoseused in spin echo) produce more T2* contrast.

Contrast Enhancement

In addition to using combinations of TR, TE,and flip angle to control the contrast character-istics, some gradient echo methods have otherfeatures for enhancing certain types of contrast.

When SAGE methods are used with rela-tively short TR values, there is the possibilitythat some of the transverse magnetization createdin one imaging cycle will carry over into thenext cycle. This happens when the TR values arein the same general range as the T2 values of thetissue. SAGE methods differ in how they use thecarry-over transverse magnetization.

A typical SAGE sequence is limited to oneRF pulse per cycle. If additional pulses wereused, as in the spin echo techniques, theywould affect the longitudinal magnetizationand upset its condition of equilibrium. How-ever, because of the relatively short TR values itis possible for the repeating small-angle excita-tion pulses to produce a spin echo effect. Thiscan occur only when the TR interval is notmuch longer than the T2 value of the tissue.


Associated with each excitation pulse,there are actually two components of thetransverse magnetization. There is the FID pro-duced by the immediate pulse and a spin echocomponent produced by the preceding pulses.The spin echo component is related to the T1characteristics of the tissue. The FID compo-nent is related to the T2* characteristics. Thecontrast characteristics of the imaging methodare determined by how these two componentsare combined. Different combinations areobtained by altering the location of the gradi-ent echo event relative to the transverse mag-netization and by turning the spin echo com-ponent on or off as described below.

Mixed Contrast

When both the FID and spin echo componentsare used, an image with mixed contrast char-acteristics will be obtained. This method pro-duces a relatively high signal intensity com-pared to the methods described below.

Spoiling and T1 Contrast Enhancement

An image with increased Tl contrast is obtainedby suppressing the spin echo component. Thisis known as spoiling. The spin echo component,which is a carryover of transverse magnetiza-tion from previous cycles, can be destroyed orspoiled by either altering the phase relationshipof the RF pulses or by applying gradient pulsesto dephase the spinning protons.

The basic SAGE method discussed up tothis point permits faster (than spin echo)image acquisition because the TR can be set toshorter values. However, the gradient echoprocess can be used in methods that providefast acquisition based on an entirely differentprinciple. We will now consider methods thatachieve their speed by filling many rows of kspace during one acquisition cycle.

In Chapter 5 we saw that in the acquisitionphase the signal data is being directed into k

space from which the image will be recon-structed. The k space is filled one row at a time.The number of rows in the k space for a spe-cific image depends on the required imagedetail. The process that directs the signals intoa specific row of k space is the spatial encodingfunction performed by one of the gradients.This will be described in Chapter 9. In con-ventional spin echo and SAGE imaging onlyone row of k space is filled with each imagingcycle. This is because there is only one echosignal produced per cycle that can be encodedto go to a specific row of k space. This meansthat the size of k space determines the mini-mum number of cycles that an acquisitionmust have. We are about to see some gradientecho methods that can fill many rows of kspace in one imaging cycle. This is achieved byusing the gradient echo process to producemany echo events from the transverse magne-tization that is present during one cycle.

Echo Planar Imaging (EPI) Method Echo planar is the fast gradient echo imagingmethod that is capable of acquiring a completeimage in a very short time. However, it requiresan MRI system equipped with strong gradientsthat can be turned on and off very rapidly. Allsystems do not have this capability. The EPImethod consists of rapid, multiple gradientecho acquisitions executed during a single spinecho event. The unique characteristic of thismethod is that each gradient echo signalreceives a different spatial encoding and isdirected into a different row of k space. Theactual spatial encoding process will bedescribed in Chapter 9. Here we are consideringonly the general concept of EPI and how itachieves rapid acquisition.

The basic EPI method is illustrated in Fig-ure 7-7. Here we see the actions that occur


within one imaging cycle. The cycle usuallybegins with a spin echo pulse sequence thatproduces a spin echo event consisting of aperiod of transverse magnetization as describedin Chapter 6. In conventional spin echo imag-ing we obtain only one signal (and fill one rowof k space) from this period of magnetization.What we are about to do with EPI is to chopthis one spin echo event into many shorter gra-dient echo events. The signals from each gradi-ent echo event will receive different spatialencodings and fill different rows of k space.

It is possible to fill all of the k space andacquire a complete image in one cycle. This isdescribed as single shot EPI. This is not alwayspractical because it might place some limita-tions on the image quality that can beachieved and is also very demanding on the

gradients. A more practical approach is todivide the acquisition into multiple shots(cycles) with each filling some fraction of thetotal k space.

The important factor is the number of gra-dient echo events created in each cycle. This isan adjustable protocol factor and is generallyknown as the EPI speed factor. This is the factorby which the acquisition time is reduced com-pared to a conventional method using thesame TR value.

GRADIENT AND SPIN ECHO(GRASE) METHOD

The GRASE method is, as the name indicates,a combination of gradient and spin echomethods. It provides the fast acquisition capa-bility of gradient echo (EPI) with the superior


Ro

ws

ECHO PLANAR IMAGING

SpinEchoEvent

ManyGradient

Echo Events“k“ Space

EPI Speed Factor = 6

Gradient Pulses

Figure 7-7. The production of many gradient echo events within one imaging cycle with the echo planar imaging (EPI) method.

contrast characteristics of spin echo, includingthe ability to produce T2 images.

The GRASE method is illustrated in Figure7-8 where we see the actions occurring withinone imaging cycle. The basic cycle is a multiplespin echo as described in Chapter 5. The differ-ence is that in conventional multiple echo,each of the echo events have different TE val-ues and are used to form several images; typi-cally, a PD and a T2 image with the same acqui-sition. Here the multiple spin echo is used for adifferent purpose. The multiple spin echoes areused to cover more of k space. As we see, eachof the spin echo events is cut into many gradi-ent echoes by the EPI process. This reduces theacquisition time by two factors: the total speed

factor is the number of multiple spin echoesmultiplied by the EPI speed factor.

Magnetization Preparation Both SAGE with short TR values and EPI canproduce very rapid acquisitions. However, theshort time intervals between the gradient echoevents do not provide sufficient time for goodlongitudinal magnetization contrast (T1 or PD)to be formed. This problem is solved by“preparing” the magnetization and formingthe contrast just one time at the beginning ofthe acquisition cycle, as shown in Figure 7-9.Two options are shown.

The longitudinal magnetization is preparedby applying either a saturation pulse, as in the


“k”Space

Ro

ws

GRADIENT AND SPIN ECHO

Spin Echo Speed Factor = 2EPI Speed Factor = 6

Total = 12

ManyGradient

Echo Events

Multiple SpinEcho Events

Gradient Pulses

Figure 7-8. The use of the GRASE method to fill many rows of k space and produce a fast acquisition.

spin echo method, or an inversion pulse, as inthe inversion-recovery method. As the longitu-dinal magnetization relaxes, contrast is formedbetween tissues with different T1 and PD val-ues. After a time interval [TI or TS (Time after

Saturation)] selected by the operator, a rapidgradient echo acquisition begins.

The total acquisition time for this methodis the time required by the acquisition cyclesplus the TI or TS time interval.


Gradient Echo Acquisition

Contrast

Contrast

SaturationPulse

Inversion Pulse

Long

itudi

nal

Mag

netiz

atio

nLo

ngitu

dina

l M

agne

tizat

ion

TS

TI

180˚

90˚

Figure 7-9. Using preparation pulses to produce longitudinal magnetization contrast prior to a rapid gradient echo acquisition.

The common characteristic of the gradient echo imaging methods is that a magnetic field gra-dient is used to produce the echo event rather than a 180˚ RF pulse, as is used in the spin echomethods. One of the principal advantages of the gradient echo process is that it is a relatively fastimaging method.

By using a gradient, and not an RF pulse, to produce the echo event, it is possible to use satu-ration/excitation pulses with flip angles less than 90˚; thereby all the longitudinal magnetizationis not destroyed (saturated) at the beginning of each cycle. Because some longitudinal magnetiza-tion carries over from cycle to cycle, it is possible to reduce the TR value and still produce usefulsignal levels. The reduced TR values result in faster imaging. The flip angle of the RF pulse is anadjustable protocol factor that controls the type of contrast produced.

Echo planar imaging is a gradient echo method in which many echo events, each with a dif-ferent phase encoding step, are created during each imaging cycle. This makes it possible to fill mul-tiple rows of k space, which results in very fast imaging. GRASE is an imaging method that com-bines the principles of echo planar and fast (turbo) spin echo to produce rapid imaging acquisitions.

When very fast gradient echo methods are used, there is not sufficient time between the echoevents for significant tissue relaxation and contrast to develop. Therefore, the desired contrast isdeveloped at the beginning of the acquisition by applying either inversion or saturation “magne-tization preparation” pulses. Then, when the desired contrast has developed, a rapid acquisitionis performed.


GradientEcho


Preparation Pulses

Small ShortTR Values

“k” Space

Rapid Phase Encoding

Rapid Acquisition

FastImaging

Flip Angle

FLIPANGLE

TRMakes

Possible

Makes

Determined By

Possible

Makes

Possible

180˚

ImageContrast

andType

Echo Planarand GRASE

Mind Map SummaryGradient Echo Imaging Methods

8Selective

Signal Suppression

81

Introduction And OverviewThere are many times when it is desirable toselectively suppress the signals from specifictissues or anatomical regions. This is done fora variety of reasons including the enhance-ment of contrast between certain tissues andthe reduction of artifacts. During the acquisi-tion process signals can be suppressed based onseveral properties of a tissue or fluid that makeit different from other surrounding tissues.These include differences in T1 values, reso-nant frequencies, and molecular binding prop-erties. Also, signals from specific anatomicalregions can be suppressed or “turned off,” usu-ally to prevent interference with imaging inother areas. We will now see how these tech-niques are used.

Fat and fluid are two materials in the bodythat can produce very intense signals andbrightness in images. This occurs with fat in T1images and with fluid in T2 images. A possibleproblem is that these bright regions can reducethe visibility of other tissues and pathologicconditions in the area.

T1-Based Fat And Fluid SuppressionLet us recall that fat has very short T1 values (260msec) and fluids have very long T1 values (2000msec). These values are outside of the range ofthe T1 values of other tissues in the body and areseparate and not mixed in with the others. Thismakes it possible to use T1 as a characteristic forthe selective suppression of both fat and fluid.

STIR Fat SuppressionSTIR is an inversion recovery method with theTI adjusted to selectively suppress the signalsfrom fat. This uses the fact that fat has a rela-tively short T1 value and recovers its longitu-dinal magnetization faster than the other tis-sues after the inversion pulse. The importantpoint here is that the magnetization of fatpasses through the zero level before the othertissues, as shown in Figure 8-1. The TI intervalis selected so that the “picture is snapped” byapplying the excitation pulse at that time.Because the fat has no magnetization at thattime, it will not produce a signal. Since this isachieved with relatively short values for TI,this method of fat suppression is often referredto as Short Time Inversion Recovery (STIR).

STIR is just the inversion recovery (IR)method with the TI set to a relatively lowvalue. The description of the basic IR methodin Chapter 6 shows how the factor TI is usedto select the time at which the longitudinalmagnetization “picture is snapped” and themagnetization is converted into image con-trast. The ability to use this method to sup-press the signals from fat is based on the factthat the longitudinal magnetization of fatpasses through zero at a time before and sepa-rated from the other tissues. Setting the TI tomeasure the longitudinal magnetization at thetime when fat is at zero produces no signal andfat will be dark in the image.

The best TI value to suppress the signalsfrom fat depends on the T1 value of fat, which


TI

Fat

Other Tissues

SHORT TIME INVERSION RECOVERY

(Short)

Long

itudi

nal M

agne

tizat

ion

Spin Echo

180˚RF

Pulse

180˚RF

Pulse

90˚RF

Pulse

Image Contrast

Zero Magnetization

Figure 8-1. The use of STIR to suppress signals from fat by setting TI to a value (short) that willimage the longitudinal magnetization at the time when fat is relaxing through the zero level.

SHORT TIME INVERSION RECOVERY

depends on the strength of the magnetic field. Itwill generally be in the range of 120 to 150 msecfor field strengths in the 0.5 T to 1.5 T range.

Another consideration with STIR is thatthe TR must be set relatively long (1500–2000msec), compared to a T1 image acquisitionwith spin echo using a TR value of approxi-mately 500 msec. This additional time isrequired for the longitudinal magnetization tomore fully recover after the excitation pulseand before the next cycle can begin.

Fluid SuppressionThe suppression of signals from fluids can beachieved by using the IR Method with the TIset to relatively long values as shown in Figure

8-2. This works because the long T1 values offluids are well separated from the T1 values ofother tissues. By setting the TI to a long valueas shown, the longitudinal magnetization isconverted to transverse and the “picture issnapped” when the fluid is at a zero value. Flu-ids appear as dark regions in the image. Whenfluid suppression is used with a T2 imageacquisition (long TE), the usually bright fluidis suppressed but other tissues with long T2values, such as pathologic tissue, remain bright.

Acquisition time is a special concern withthis method. That is because when long TI val-ues are used, the TR values must also be long(5000–6000 msec) and that increases theacquisition time. For this reason, the practical

SELECTIVE SIGNAL SUPPRESSION 83

FLUID SUPPRESSION

Zero Magnetization

(Long)

T I

Image Contrast

Spin EchoFluid

Other Tissues

180˚RF

Pulse

180˚RF

Pulse

90˚RF

Pulse

Long

itudi

nal M

agne

tizat

ion

Figure 8-2. The suppression of fluid by selecting a long TI that will image the longitudinalmagnetization at the time when fluid is relaxing through zero.

FLUID SUPPRESSION

thing is to use this method with one of the fastacquisition techniques.

SPIR Fat SuppressionSpectral Presaturation with Inversion Recovery(SPIR) is a fat suppression technique whichmakes use of the fact that fat and the water con-tent of tissues resonate at different frequencies(on the RF frequency spectrum) as described inChapter 3. We must be careful not to confusethe two fat suppression methods, STIR and SPIR.As we have just seen, STIR uses the difference inT1 values to selectively suppress the signals fromfat. Now with SPIR, we will use the differencesin resonant frequency to suppress the fat signals.This technique is illustrated in Figure 8-3. The

unique feature of this method is that the imag-ing cycle begins with an inversion pulse that isapplied at the fat resonant frequency. This selec-tively inverts the longitudinal magnetization ofthe fat without affecting the other tissues. TheTI is set so that the spin echo excitation pulse isapplied at the time when the fat longitudinalmagnetization is passing through zero. Thisresults in T1 and T2 images with the signalsfrom fat removed.

The advantage of the SPIR method is thatthe contrast of tissues with relatively short T1values is not diminished as it might be withthe STIR method. For example, the use ofgadolinium contrast media reduces the T1value of the water component of tissue. These


SPECTRAL PRESATURATIONINVERSION RECOVERY

Applied at theResonant Frequency

of Fat

Zero Magnetization

Spin Echo

Image Contrast

Fat

Fat

OtherTissues

Long

itudi

nal M

agne

tizat

ion

180˚RF

Pulse

RFPulse

90˚RF

Pulse

Figure 8-3. Suppressing the signals from fat by applying an inversion pulse tuned to the resonantfrequency of fat so that it does not affect the other tissues.

SPECTRAL PRESATURATIONINVERSION RECOVERY

short T1 value signals would be suppressed bySTIR, but not by SPIR.

There are some precautions that must beobserved when using SPIR. They relate to hav-ing very good magnetic field homogeneity.Recall that the resonant frequency is con-trolled by the field strength in each location.Therefore, for the RF suppression pulse to accu-rately suppress the fat magnetization over theimage area, the fat must be resonating at pre-cisely the same frequency. This requires a veryhomogeneous (within just a few parts per mil-lion) magnetic field. This is achieved by shim-ming the field before the acquisition, remov-ing metal objects that might distort the field,and by using a relative small field of view.

An alternative to the SPIR method is toapply a saturation rather than an inversion

pulse tuned to the fat resonant frequency. Thisis sometimes referred to as chemical saturation.

Magnetization Transfer Contrast (MTC)Magnetization Transfer Contrast (MTC) is atechnique that enhances image contrast byselectively suppressing the signals from specifictissues. The amount of suppression dependson a specific tissue’s magnetization transfer char-acteristics. Maximum suppression is obtainedfor tissues that have a high level of magnetiza-tion transfer.

The MTC technique is illustrated in Figure8-4. It is based on the principle that the pro-tons in tissue are in different states of mobil-ity, which we will designate as the “free” pooland the “bound” pool.


MAGNETIZATION TRANSFERBound Water Pool

(No Signals)Free Water Pool(Signal Source)

Fluids

MostTissues

MagnetizationTransfer

SaturationPulse

90̊RF

Pulse

Resonant Frequency

Figure 8-4. The use of magnetization transfer between different types of tissue to suppress selective signals.

Free Proton Pool

The protons that produce signals and are visiblein MRI are not rigidly bound and might be con-sidered to be “free” and in a general “semi-solid”structure. This environment produces relativelylong T2 values (in comparison to the boundstate) and a relatively narrow resonant frequency.

Bound Proton Pool

Most tissues also contain protons that are morerigidly bound and associated with more “solid”structures such as large macromolecules andmembranes. These structures have very shortT2 values. This means that the transverse mag-netization decays before it can be imaged withthe usual methods. Therefore, these protonsdo not contribute to the image. An importantcharacteristic of these protons is that they havea much broader resonant frequency spectrumthan the “free” protons.

Magnetization Transfer

Magnetization transfer is a process in which thelongitudinal magnetization of one pool influ-ences the longitudinal magnetization in theother pool. In other words, the longitudinalmagnetizations of the two pools are coupledtogether but not to the same degree in all tis-sues. The MTC process makes use of this differ-ence in coupling to selectively suppress the sig-nals from certain tissues. This is how it is done.

Selective Saturation

The objective of this technique is to saturate andsuppress selective signals from specific tissues toincrease the contrast.

Prior to the beginning of the imaging acqui-sition cycle a saturation pulse is applied at a fre-quency that is different from the resonant fre-quency of the “free” protons. Therefore, it doesnot have a direct effect on the protons that are

producing the signals. However, the saturationpulse is within the broader resonant frequencyof the “bound” protons. It produces saturationof the longitudinal magnetization in the“bound” pool.

The effect of the saturation is now trans-ferred to the longitudinal magnetization of the“free” pool by the magnetization transferprocess. The key is that the transfer is not thesame for all tissues. Only the tissues with a rel-atively high magnetization transfer couplingand a significant bound pool concentrationwill experience the saturation and have theirsignals reduced in intensity.

Fluids, fat, and bone marrow have very little,if any, magnetization transfer. Therefore, theywill not experience the transferred saturation,and will remain relatively bright in the images.

Most other tissues have some, but varyingdegrees of, magnetization transfer. When theMTC technique is used, the saturation producedby the RF pulse applied to the “bound” protonswill be transferred to the “free” protons, but onlyin those tissues that have a significant magneti-zation transfer capability. The result is that thesetissues will be saturated to some degree and theirsignal intensities will be reduced.

Therefore, MTC is a way of enhancing con-trast in an image by suppressing the signalsfrom tissues that have a relatively high mag-netization transfer. One example is to use MTCto reduce the brightness (signal intensity) ofbrain tissue so that the vascular structures willbe brighter in angiography.

Regional SaturationThere are procedures in which it is desirable to suppress signals from specific anatomicalregions. The two major applications of this areto reduce motion-induced artifacts, as describedin Chapter 14, and to suppress the signals fromblood that is flowing in a specific direction, as


discussed in Chapter 12. At this time we willconsider the general technique, which is illus-trated in Figure 8-5.

Let us recall that gradients are used to varythe magnetic field strength across a patient’sbody. In the presence of a gradient one regionof the body is in a different field strength fromanother and is therefore tuned to a differentresonant frequency. This makes it possible to

apply RF pulses selectively to specific regionswithout affecting adjacent regions.

In Chapter 14 we will see that a majorsource of artifacts in MRI is the motion ormovement of tissues and fluids. The motionproduces errors in the spatial encoding of thesignals that causes them to be displayed in thewrong location in the image. Signals frommoving tissues and fluids are displayed asstreaks, which are undesirable artifacts.

With the regional saturation technique theobjective is to suppress selective signals origi-nating from one region, usually the moving tis-sue or fluid, without affecting these signals in theregion that is being imaged. The specific appli-cations of this will be described in Chapter 14.

Prior to the imaging cycle pulse sequence,a saturation pulse is selectively applied to theregion that is to be suppressed. The saturationpulse is given a frequency that is different fromthe frequency of the other imaging pulses. Thisis so that it will be tuned to the resonant fre-quency of the region that is to be suppressed.This region will have a resonant frequency dif-ferent from the imaged area because of thepresence of the gradient as described above.

The region that is saturated is a three-dimensional (3-D) volume or slab of tissue. Itis important that the slab be properly posi-tioned in relationship to the imaged area forbest results.

The application of regional saturation tosuppress artifacts will be discussed in moredetail when we consider artifacts in Chapter 14.


SaturatedRegion

Saturation EchoExcitation

RF Pulses

ImageArea

90̊ 90̊ 180̊

Figure 8-5. The application of a saturationpulse can be directed to a specific anatomical

region to suppress undesirable signals from moving tissues.

REGIONAL SATURATION

It is often desirable to suppress the signals and resulting brightness of selected tissues oranatomical regions to improve visibility of other tissues or general image quality. It is possible toselectively suppress signals from specific tissues if the tissues are significantly different from theother tissues in terms of some MR characteristic.

Signals from fat, generally very bright in T1 images, can be suppressed with two techniques.Because fat has a very short T1 value compared to other tissues, it can be suppressed with the STIRmethod, an inversion recovery method in which the TI is set to snap the picture when the mag-netization of fat is passing through the zero level. The resonant frequency of fat molecules isslightly different from water molecules because of the chemical shift effect. The SPIR method makesuse of this by applying an RF pulse at the fat frequency to reduce the fat magnetization to the zerolevel at the beginning of each imaging cycle.

Signals from fluid can be suppressed by using an inversion recovery method with the TI set toa long value. This works because fluids have long T1 values and the fluid’s magnetization passesthrough the zero level significantly later and separate from that of tissues. The MTC technique canbe used to reduce signal intensity from tissues that have a relatively high magnetization transfercharacteristic. This can be used to enhance image contrast.

Saturation pulses can be selectively applied to specific anatomical regions to suppress any sig-nals that could occur from tissues or fluids in that region. This is useful for reducing motion arti-facts and also for reducing the signals from flowing blood in specific anatomical regions.


Fluid

Long

Long Different

BoundH2OPool

Different

Short

Short

SaturationPulse

InversionRecovery

STIRSPIR

Fat

MagnetizationTransfer

Moving Tissue

RESONANTFREQUENCY

90˚TI TI

T1

Imag

e

Mind Map SummarySelective Signal Suppression

9Spatial Characteristics

of the Magnetic Resonance Image

89

Introduction And OverviewThe MR image formation process subdivides asection of the patient’s body into a set of slicesand then each slice is cut into rows andcolumns to form a matrix of individual tissuevoxels. This was introduced first in Chapter 1and illustrated in Figure 1-3. The RF signalfrom each individual voxel must be separatedfrom all of the other voxels and its intensitydisplayed in the corresponding image pixel, asshown in Figure 9-1. This is achieved byencoding or addressing the signals during theacquisition phase and then, in effect, deliver-ing the signal intensities to the appropriatepixels which have addresses within the imageduring the reconstruction phase. Because there

are two dimensions, or directions, in an image,two different methods of encoding must beused. This is analogous to mail that must haveboth a street name and a house number in theaddress. We are about to see that the twomethods of addressing the signals are calledfrequency-encoding and phase-encoding. Onemethod is applied to one direction in theimage and the other method is used to addressin the other direction.

This two-step process consisting of the sig-nal acquisition phase followed by the imagereconstruction phase is illustrated in Figure 9-2.Different actions happen in these two phasesthat must be considered when setting up animaging procedure.


Frequency Address


Tissue Voxels

RF Signal

Frequency Encoding

Phas

e En

codi

ng

Phas

e A

ddre

ss

Image Pixels

Figure 9-1. The relationship of tissue voxels to image pixels.

MRI IMAGE PRODUCTION

Signal Acquisition

Encoding Direct ions

Tissue Voxels “k” Space

Image Reconstruction

Image Pixels

Fourier Transform

Frequency

Phas

e

Phas

e

RF SignalData

Figure 9-2. The two phases—signal acquisition and image reconstruction—that are required to produce an MR image.


MRI IMAGE PRODUCTION

Signal AcquisitionDuring the acquisition phase the RF signals areemitted by the tissue and received by the RFcoils of the equipment. During this process thesignals from the different slices and voxels aregiven distinctive frequency and phase charac-teristics so that they can be separated from theother signals during image reconstruction. Theacquisition phase consists of an imaging cyclethat is repeated many times. The time requiredfor image acquisition is determined by the timeTR, which is the duration of one cycle or its rep-etition time, and the number of cycle repeti-tions. The number of cycles is determined bythe image quality requirements. In general, thequality of an image can be improved by increas-ing the number of acquisition cycles. This isconsidered in much more detail in Chapter 10.

The result of the image acquisition processis a large amount of data collected and stored incomputer memory. At this point the data repre-sent RF signal intensities characterized by thetwo characteristics, frequency and phase. The con-cept of frequency and phase will be developedlater. At this point in the process the data are notyet in the form of an image but are located in kspace. The data will later be transformed intoimage space by the reconstruction process.

Image ReconstructionImage reconstruction is a mathematical processperformed by the computer. It transforms thedata collected during the acquisition phase intoan image. We can think of reconstruction as theprocess of sorting the signals collected duringthe acquisition and then delivering them to theappropriate image pixels. The mathematicalprocess used is known as Fourier transformation.Image reconstruction is typically much fasterthan image acquisition and requires very little,if any, control by the user.

Image CharacteristicsThe most significant spatial characteristic of animage is the size of the individual tissue vox-els. Voxel size has a major effect on both thedetail and noise characteristics of the image.The user can select the desired voxel size byadjusting a combination of imaging factors, asdescribed in Chapter 10.

GradientsThe spatial characteristics of an MR image areproduced by actions of the gradients appliedduring the acquisition phase. Magnetic fieldgradients are used first to select slices and thengive the RF signals the frequency and phasecharacteristics that create the individual voxels.

As we will see later, a gradient in one direc-tion is used to create the slices, and then gra-dients in the other directions are used to cutthe slices into rows and columns to create theindividual voxels. However, these functionscan be interchanged or shared among the dif-ferent gradient coils to permit imaging in anyplane through the patient’s body.

The functions performed by the variousgradients usually occur in a specific sequence.During each individual image acquisition cyclethe various gradients will be turned on and offat specific times. As we will see later, the gradi-ents are synchronized with other events such asthe application of the RF pulses and the acqui-sition of the RF signals.

Slice SelectionThere are two distinct methods used to createthe individual slices. The method of selectiveexcitation actually creates the slice during theacquisition phase. An alternative method is toacquire signals from a large volume of tissue(like an organ) and then create the slices during

SPATIAL CHARACTERISTICS OF THE MR IMAGE 91

the reconstruction process. These are oftenreferred to as 2-D (volume) and 3-D (volume)acquisitions. However, each produces data thatare reconstructed into slice images. Both meth-ods have advantages and disadvantages, whichwill be described later.

Selective Excitation

The first gradient action in a cycle defines thelocation and thickness of the tissue slice to beimaged. We will illustrate the procedure for aconventional transaxial slice orientation.Other orientations, such as sagittal, coronal,and angled combinations, are created by inter-changing and combining gradient directions.

Slice selection using the principle of selec-tive excitation is illustrated in Figure 9-3.When a magnetic field gradient is orientedalong the patient axis, each slice of tissue is ina different field strength and is tuned to a dif-ferent resonant frequency. Remember, this isbecause the resonant frequency of protons isdirectly proportional to the strength of themagnetic field at the point where they arelocated. This slice selection gradient is present

whenever RF pulses are applied to the body.Since RF pulses contain frequencies within alimited range (or bandwidth), they can excitetissue only in a specific slice. The location ofthe slice can be changed or moved along thegradient by using a slightly different RF pulsefrequency. The thickness of the slice is deter-mined by a combination of two factors: (1) thestrength, or steepness, of the gradient, and (2)the range of frequencies, or bandwidth, in theRF pulse.

Multi-Slice Imaging

In most clinical applications, it is desirable tohave a series of images (slices) covering a spe-cific anatomical region. By using the multi-slice mode, an entire set of images can beacquired simultaneously. The basic principle isillustrated in Figure 9-4. The slices are sepa-rated by applying the RF pulses and detectingthe signals from the different slices at differenttimes, in sequence, during each imaging cycle.

When the slice selection gradient is turnedon, each slice is tuned to a different resonantfrequency. A specific slice can be selected for


Magnetic Field Strength

Selected Slice

Gradient

Tissue Resonancehighlow

Frequency

RF Pulse Frequency

Figure 9-3. The use of a gradient to tune a specific slice so that it can be selectively excited by an RF pulse.

SELECTIVE EXCITATION

excitation by adjusting the RF pulse frequencyto correspond to the resonant frequency ofthat slice. The process begins by applying anexcitation pulse to one slice and collecting theecho signal. Then, while that slice undergoeslongitudinal relaxation before the next cyclecan begin, the excitation pulse frequency isshifted to excite another slice. This process isrepeated to excite and collect signals from theentire set of slices at slightly different timeswithin one TR interval.

The advantage of multi-slice imaging is thata set of slices can be imaged in the same time asa single slice. The principal factor that limits thenumber of slices is the value of TR. It takes a cer-tain amount of time to excite and then collect thesignals from each slice. The maximum numberof slices is the TR value divided by the timerequired for each slice. This limitation is espe-cially significant for T1-weighted images thatuse relatively short TR values.

A factor to consider when selecting the slic-ing mode is that multiple slice selective excita-

tion cannot produce the contiguous slices thatthe volume acquisition technique can. Withselective excitation there is the possibility thatwhen an RF excitation/saturation pulse is appliedto one slice of tissue, it will also produce someeffect in an adjacent slice. This is a reason for leav-ing gaps between slices during the acquisition.

Volume Acquisition Volume (3-D) image acquisition has the advan-tage of being able to produce thinner and morecontiguous slices. This is because of the processused to slice the tissue. Rather than producingeach slice during the acquisition phase, the slic-ing is done during the reconstruction phaseusing the process of phase-encoding. The actualprocess of phase-encoding will be describedlater in this chapter. At this time we only con-sider how it is used for slicing. With this method,no gradient is present when the RF pulse isapplied to the tissue. Since all tissue within ananatomical region, such as the head, is tuned tothe same resonant frequency, all tissues are


0 200 400 600 800 1000 mSec

Excitation Sequence1 Cycle Time (TR)

Repeat for each cycle

Figure 9-4. Multiple slice imaging applies pulses to and produces signals from different slices within one imaging cycle.

MULTIPLE SLICE IMAGING

excited simultaneously. The next step, as illus-trated in Figure 9-5, is to apply a phase-encod-ing gradient in the slice selection direction. Involume imaging, phase-encoding is used tocreate the slices in addition to creating thevoxel rows as described below. The phase-encoding gradient used to define the slicesmust be stepped through different values, cor-responding to the number of slices to be cre-ated. At each gradient setting, a complete setof imaging cycles must be executed. Therefore,the total number of cycles required in oneacquisition is multiplied by the number ofslices to be produced. This has the disadvan-tage of causing 3-D volume acquisitions tohave a relatively long acquisition time com-pared to 2-D multiple slice acquisitions. Thatis why this type of acquisition is often usedwith one of the faster imaging methods.

The primary advantage of volume imagingis that the phase-encoding process can gener-ally produce thinner and more contiguousslices than the selective excitation process usedin 2-D slice acquisition. The primary disad-vantage is longer acquisition times.

Frequency EncodingA fundamental characteristic of an RF signal isits frequency. Frequency is the number ofcycles per second of the oscillating signal. Thefrequency unit of Hertz (Hz) corresponds toone cycle per second. Radio broadcast stationstransmit signals on their assigned frequency.By tuning our radio receiver to a specific fre-quency we can select and separate from allother signals the specific broadcast we want toreceive. In other words, the radio broadcastsfrom all of the stations in a city are frequencyencoded. The same process (frequency-encod-ing) is used to cause voxels to produce signalsthat are different and can be used to create onedimension of the image.

Let us review the concept of RF signal pro-duction by voxels of tissue, as shown in Figure9-6. RF signals are produced only when trans-verse magnetization is present. The uniquecharacteristic of transverse magnetization thatproduces the signal is a spinning magneticeffect, as shown. The transverse magnetiza-tion spins around the axis of the magnetic


Slices Producedby

Phase EncodingExcited Volume

LastCycle

FirstCycle

Slice

Phase EncodingGradient

Figure 9-5. The 3-D volume acquisition process uses the phase-encoding process to produce thin slices.

3-D VOLUME ACQUISITION

field. A spinning magnet or magnetization inthe vicinity of a coil forms a very simple elec-tric generator. It generates one cycle for eachrevolution of the magnetization. When themagnetization is spinning at the rate of mil-lions of revolutions per second, the result is anRF signal with a frequency in the range ofMegahertz (MHz).

Resonant FrequencyThe frequency of the RF signal is determined bythe spinning rate of the transverse magnetiza-tion. This, in turn, is determined by two factors,as was described in Chapter 3. One factor is thespecific magnetic nuclei (usually protons) andthe other is the strength of the magnetic field

in which the voxel is located. When imagingprotons, the strength of the magnetic field isthe factor used to vary the resonant frequencyand the corresponding frequency of the RF sig-nals. In Figure 9-6 we see two voxels located indifferent strength fields. The result is that theyproduce different frequency signals.

Figure 9-7 shows the process of frequencyencoding the signals for a row of voxels. In thisexample, a gradient is applied along the row.The magnetic field strength is increased fromleft to right. This means that each voxel islocated in a different field strength and is res-onating at a frequency different from all of theothers. The resonant and RF signal frequenciesincrease from the left to right as shown.


MagneticField

MagnetizationSpin

RF SignalFrequency

Fast

Strong

Weak

Slow

Low

High

Figure 9-6. The effect of field strength on the frequency of RF signals produced by transverse magnetization.

The frequency-encoding gradient is on atthe time of the echo event when the signals areactually being produced. The signals from allof the voxels in a slice are produced simulta-neously and are emitted from the body mixedtogether to form a composite signal at the timeof the echo event. The individual signals willbe separated later by the reconstruction processto form the voxels.

Phase-EncodingPhase is a relationship between one signal andanother, as illustrated in Figure 9-8. Here we seetwo voxels producing RF signals. The transversemagnetization is spinning at the same rate andproducing signals that have the same fre-quency. However, we notice that one signal ismore advanced in time or is out of step withthe other. In other words, the two signals areout of phase. The significance of voxel-to-voxelphase in MRI is that it can be used to separatesignals and create one dimension in the image.

A phase difference is created by temporarilychanging the spinning rate of the magnetization

of one voxel with respect to another. This hap-pens when the two voxels are located in mag-netic fields of different strengths. This can beachieved by turning on a gradient, as shown inFigure 9-9.

Let us begin the process of phase encodingby considering the column of voxels shown inthe illustration. We are assuming that all vox-els have the same amount of transverse mag-netization and that the magnetization is spin-ning in-phase at the time just prior to thephase-encoding process.

When the phase-encoding gradient isturned on, we have the condition illustratedwith the center column of voxels. The strengthof the magnetic field is increasing from bottomto top. Therefore, the magnetization in eachvoxel is spinning at a different rate with thespeed increasing from bottom to top. Thiscauses the magnetization from voxel to voxelto get out of step or produce a phase difference.The phase-encoding gradient remains on for ashort period of time and then is turned off.This leaves the condition represented by the


FREQUENCY ENCODING

LowRadio Frequency

Magnetic Field GradientWeak

Strong

High

Figure 9-7. The frequency encoding of a row of voxels within a slice.

FREQUENCY ENCODING


PHASE ENCODING

Strong

WeakOff On Off

Magnetic Field Gradient

Figure 9-9. Phase-encoding produced by turning on a gradient for a short time, and then turning it off. The phase difference remains.

MagneticField

Same Same DifferentPhase

SameFrequency

MagnetizationSpin

RF SignalPhase

Figure 9-8. The concept of phase between the signals from two voxels.

column of voxels on the right. This is the con-dition that exists at the time of the echo eventwhen the signals are actually produced. As wesee, the signals from the individual voxels aredifferent in terms of their phase relationship.In other words, the signals are phase-encoded.All of the signals are emitted at the same timeand mixed together as a composite echo sig-nal. Later, the reconstruction process will sortthe individual signal components.

Phase-encoding is the second function per-formed by a gradient during each cycle, asshown in Figure 9-10. During each passthrough an imaging cycle, the phase-encodinggradient is stepped to a slightly different value.

The signals acquired with each phase-encoding gradient strength fills one row of kspace. This is a very important point thatshould be emphasized: Each row of k space isreserved for signals with a specific degree of phase-encoding. The degree of phase-encoding is deter-mined by the strength and duration of thephasing gradient applied during each cycle.Therefore, the phase-encoding process must be

repeated depending on the size of k space andthat is determined by the image matrix size inthe phase-encoded direction.

One MRI phase-encoding step produces acomposite signal from all voxels within a slice.The difference from one step to another is thatindividual voxel signals have a different phaserelationship within the composite signal.

To reconstruct an image by the conventional2-D Fourier transformation method, one com-posite signal, or phase-encoded step, must be col-lected for each voxel to be created in the phase-encoding direction. Therefore, the minimumnumber of steps required to produce an image isdetermined by the size of the image matrix andk space. It takes 256 phase-encoding steps to pro-duce an image with a 256 ¥ 256 matrix.

The Gradient CycleWe have seen that various gradients are turnedon and off at specific times within each imag-ing cycle. The relationship of each gradient tothe other events during an imaging cycle isshown in Figure 9-10. The three gradient activ-ities are:1. The slice selection gradient is on when RF

pulses are applied to the tissue. This limitsmagnetic excitation, inversion, and echoformation to the tissue located within thespecific slice.

2. The phase-encoding gradient is turned onfor a short period in each cycle to produce aphase difference in one dimension of theimage. The strength of this gradient ischanged slightly from one cycle to anotherto fill the different rows of k space neededto form the image.

3. The frequency-encoding gradient is turnedon during the echo event when the signalsare actually emitted by the tissue. Thiscauses the different voxels to emit signalswith different frequencies.


GRADIENTS

SliceSelection

PhaseEncoding

FrequencyEncoding

RF Pulses Echo Event

Last

First

One Imaging Cycle

Figure 9-10. The relationship of the threegradient actions—slice selection, phase-encoding, and frequency-encoding—

to each other and to the RF pulses and signals.They are applied in different directions.

Because of the combined action of thethree gradients, the individual voxels withineach slice emit signals that are different in tworespects—they have a phase difference in onedirection and a frequency difference in theother. Although these signals are emitted at thesame time, and picked up by the imaging sys-tem as one composite signal at the time of theecho event in each cycle, the reconstruction

process can sort the signals into the respectivecomponents and display them in the correctimage pixel locations.

Image ReconstructionThe next major step in the creation of an MRimage is the reconstruction process. Recon-struction is the mathematical process performedby the computer that converts the collected


Figure 9-11. The concept of signal encoding (addressing) and image reconstruction (sorting and delivery).

Voxels(Magnetization)

CompositeEcho Signal

(Many Frequencies)

Signal Encoding(Addressing)

Phase Address

Freq

uenc

y A

ddre

ss

Image Pixels

Composite Echo Signal(From All Voxels)

Image Reconstruction(Sorting and Delivery)

Tissue Voxels

Acquisition

Reconstruction

Phase

FourierTransform

Computer

Frequency

Pixel(Brightness)

Inte

nsity

High

Low

Time

Low

High

Freq

uenc

y

Freq

uenc

y-En

code

d

signals in k space into an actual image. There areseveral reconstruction methods, but the oneused for most clinical applications is the 2-DFourier transformation.

It is a mathematical procedure that cansort a composite signal into individual fre-quency and phase components. Since eachvoxel in a row emits a different signal fre-quency and each voxel in a column a differentphase, the Fourier transformation can deter-mine the location of each signal componentand direct it to the corresponding pixel.

Let us now use the concept illustrated inFigure 9-11 to summarize the spatial charac-teristics of the MR image. We will use a postalanalogy for this purpose.

In the image each column of pixels has aphase address corresponding to different streetnames. Each row of pixels has a frequencyaddress corresponding to house numbers. There-fore, each individual pixel has a unique addressconsisting of a combination of frequency and

phase values analogous to a street name andhouse number.

The frequency- and phase-encoding processduring acquisition “writes” an address on thesignal from each voxel. These signals are mixedtogether and collected in a “post-office” called kspace. The signals (“mail”) are then sorted by theFourier transform process and hopefully deliv-ered to the correct pixel address in the image.

In Chapter 14 we will see that if a voxel oftissue moves during the acquisition process, itmight not receive the correct phase addressand the signal will be delivered to the wrongpixel. This creates ghost images and streak arti-facts in the phase-encoded direction.

The chemical-shift artifact is caused by thedifference in signal frequency between tissuescontaining water and fat. When it is present inan image, signals from the water componentsand fat will be offset by a few pixels. We willsee how this is controlled in Chapter 14.



During an MRI procedure a section of a patient’s body is first divided into a set of slices, andthen each slice is divided into a matrix of voxels. These actions are produced by the gradients.

Two methods can be selected to produce the slices. The most common method, 2-D multipleslice acquisition, applies a gradient so that an individual slice is tuned to a resonant frequency dif-ferent from the other slice positions. This gradient is turned on when the RF pulses are applied.Therefore, only the tissue in a specific slice is excited and goes through the process to produce sig-nals. An alternate method, 3-D volume acquisition, uses phase-encoding to produce slices. It isgenerally capable of producing thinner, more contiguous slices.

Two different methods are used to cut a slice into voxels. Phase-encoding is used in one direc-tion, and frequency-encoding in the other. Phase-encoding is produced by applying a gradient tothe transverse magnetization during each imaging cycle. To produce sufficient phase-encodinginformation to permit image reconstruction, many different phase-encoding gradient strengthsmust be used. In the typical imaging procedure the phase-encoding gradient strength is changed

Sliceof Voxels

Sign

als

“k” Space

SignalsPhase

Encoding

EchoEvent

SliceSelection

RF Pulses

Gradients

Uses

Selective Excitation

Uses

PhaseEncoding

ThinSlices

ToProduce

3-D VolumeAcquisition

Volumeof Slices

2-D Multiple SliceAcquisition

Mind Map SummarySpatial Characteristics of the Magnetic Resonance Image

from cycle to cycle. The strength of the phase-encoding gradient, in effect, directs the signal datainto a specific row of k space. All the rows of k space must be filled with data before the imagereconstruction can be performed. The number of rows of k space is one of the factors that deter-mine how many imaging cycles must be used, which, in turn, affects image acquisition time.

Frequency-encoding is produced by applying a gradient at the time of the echo event duringeach cycle.


10Image Detail and Noise

103

Introduction And Overview

Two characteristics of the MR image thatreduce the visibility of anatomical structuresand objects within the body are blurring andvisual noise. These were introduced in Chapter1 as image quality characteristics. Both imageblurring and visual noise are undesirable char-acteristics that collectively reduce the overallquality of an image and the objects in theimage as illustrated in Figures 1-6 and 1-7. Inan image, the combined effects of blur andnoise produce a “curtain of invisibility” thatextends over some objects based on objectcharacteristics. This is shown in Figure 10-1,where we see objects arranged according totwo characteristics. In the horizontal direction,

the objects are arranged according to size.Decreasing object size corresponds to increas-ing detail. In the vertical direction, the objectsare arranged according to their contrast. Theobject in the lower left is both large and has ahigh level of contrast. This is the object thatwould be most visible under a variety of imag-ing conditions. The object that is always themost difficult to see is the small, low contrastobject, which in Figure 10-1 would be locatedin the upper right corner.

In every imaging procedure we can assumethat some potential objects within the body willnot be visible because of the blurring and noisein the image. This loss of visibility is representedby the “curtain” or area of invisibility indicatedin Figure 10-1. The location of the boundary

between the visible and invisible objects, oftenreferred to as a contrast-detail curve, is deter-mined by the amount of blurring and noiseassociated with a specific imaging procedure. Ingeneral, blurring reduces the visibility ofanatomical detail or other small objects that arelocated in the lower right region. Visual noisereduces the visibility of low contrast objectslocated in the upper left region.

The imaging protocol determines theboundary of visibility by altering the amountof blurring and noise. These two characteristicsare determined by the combination of manyadjustable imaging factors. It is a complexprocess because the factors that affect visibilityof detail (blurring) also affect noise, but in theopposite direction. As we will see when a pro-tocol is changed to improve visibility of detail,the noise is increased. Another point to con-sider is that several of the factors that have an

effect on both image detail and noise also affectimage acquisition time, which will be discussedin Chapter 11. Therefore, when formulating animaging protocol one must consider the mul-tiple effects of the imaging factors and thenselect factor values that provide an appropriatecompromise and an optimized acquisition fora specific clinical study with respect to detail(blurring), noise, and acquisition speed.

We will now consider the many factorsthat have an effect on the characteristics ofimage detail and noise.

Image DetailThe ability of a magnetic resonance image toshow detail is determined primarily by the sizeof the tissue voxels and corresponding imagepixels. Pixel size can be changed without majortradeoffs. However, as we are about to observe,there are significant effects of changing voxel


.

.

EFFECT OF NOISE AND BLUR

Low

HighLarge

Object Size (Detail)

Visible

Invisible

Small

Ob

ject

Co

ntr

ast

Figure 10-1. The impact of image noise and blurring on object visibility. Noise reduces visibility of low contrast objects. Blur reduces visibility of small objects.

size that must be considered. The real challengeis selecting a voxel size that is optimum for aspecific clinical procedure.

In principle, all structures within an indi-vidual voxel are blurred together and repre-sented by the signal intensity representing thatvoxel. It is not possible to see details within avoxel, just the voxel itself. When we view anMR image, we are actually looking at an imageof a matrix, or array, of the voxels. We usuallydo not see the individual voxels because theyare so small and they might be interpolated intoeven smaller image pixels. However, even if wedo not see the individual voxels, their size deter-mines the anatomical detail that we can see.The amount of image blurring is determined bythe dimensions of the individual voxels.

Three basic imaging factors determine thedimensions of a tissue voxel, as illustrated inFigure 10-2. The dimension of a voxel in the

plan of the image is determined by the ratio ofthe field of view (FOV) and the size of thematrix. Both of these factors can be used toadjust image detail. The thickness of the slice isa factor in voxel signal intensity.

The selection of the FOV is determined pri-marily by the size of the body part being imaged.One problem that can occur is the appearance offoldover artifacts when the FOV is smaller thanthe actual body section. However, there are arti-fact suppression techniques that can be used toreduce this foldover problem, as described inChapter 14. The maximum useful FOV is usuallylimited by the dimensions and characteristics ofthe RF coil. The important thing to remember isthat smaller image FOVs and smaller voxels pro-duce better visibility of detail.

Matrix size refers to the number of voxelsin the rows or columns of the matrix. Thematrix size is a protocol factor selected by the

IMAGE DETAIL AND NOISE 105

ThicknessFOVMatrix

IMAGE DETAIL

Voxel Size

Figure 10-2. Voxel size and detail in MR images is determined by the values selected for the three protocol factors: FOV, matrix size, and slice thickness.

operator before the imaging procedure. Typi-cal matrix dimensions are in the range of 128to 512 mm.

Noise SourcesRandom RF energy can be generated by ther-mal activity within electrical conductors andcircuit components of the receiving system. Inprinciple, the patient’s body is a component ofthe RF receiving system. Because of its mass, itbecomes the most significant source of imagenoise in most imaging procedures. The specificnoise source is the tissue contained within thesensitive FOV of the RF receiver coils. Somenoise might be generated within the receivercoils or other electronics, but it is usually muchless than the noise from the patient’s body.

Many devices in the environment produceRF noise or signals that can interfere with MRI.These include radio and TV transmitters, elec-trosurgery units, fluorescent lights, and com-puting equipment. All MR units are installedwith an RF shield, as described in Chapter 2, toreduce the interference from these externalsources. External interference is not usually aproblem with a properly shielded unit. Whenit does occur, it generally appears as an imageartifact rather than the conventional randomnoise pattern.

Signal-To-Noise Considerations Image quality is not dependent on theabsolute intensity of the noise energy butrather the amount of noise energy in relationto the image signal intensity. Image qualityincreases in proportion to the signal-to-noiseratio. When the intensity of the RF noise is lowin proportion to the intensity of the image sig-nal, the noise has a low visibility. In situations

where the signals are relatively weak, the noisebecomes much more visible. The principle isessentially the same as with conventional TVreception. When a strong signal is received,image noise (snow) is generally not visible;when one attempts to tune in to a weak TV sig-nal from a distant station, the noise (noise)becomes significant.

In MRI, the loss of visibility resulting fromthe noise can be reduced by either reducing thenoise intensity or increasing the intensity of the sig-nals. This is illustrated in Figure 10-3. Let usnow see how this can be achieved.

Voxel SizeOne of the major factors that affects signalstrength is the volume of the individual voxels.The signal intensity is proportional to the totalnumber of protons contained within a voxel.Large voxels, that contain more protons, emitstronger signals and result in less image noise.Unfortunately, as we have just discovered, largevoxels reduce image detail. Therefore, when thefactors for an imaging procedure are beingselected, this compromise between signal-to-noise ratio and image detail must be consid-ered. The major reason for imaging relativelythick slices is to increase the voxel signal inten-sity and it also allows shorter TE values.

Field StrengthThe strength of the RF signal from an individ-ual voxel generally increases in proportion tothe square of the magnetic field strength. How-ever, the amount of noise picked up from thepatient’s body often increases with fieldstrength because of adjustments in the band-width factor for the higher fields. This isdescribed in Chapter 14. Because of differencesin system design, no one precise relationshipbetween signal-to-noise ratio and magnetic


field strength applies to all systems. In general,MRI systems operating at relatively high fieldstrengths produce images with higher signal-to-noise ratios than images produced at lowerfield strengths, when all other factors are equal.

Tissue CharacteristicsSignal intensity, and the signal-to-noise ratio,depend to some extent on the magnetic char-acteristics of the tissue being imaged. For a spe-cific set of imaging factors, the tissue charac-teristics that enhance the signal-to-noiserelationship are high magnetic nuclei (proton)concentration, short T1, and long T2. The pri-mary limitation in imaging nuclei other thanhydrogen (protons) is the low tissue concen-tration and the resulting low signal intensity.

TR and TE

Repetition time (TR) and echo time (TE) are thefactors used to control contrast in most imag-ing methods. We have observed that these twofactors also control signal intensity. This mustbe taken into consideration when selecting thefactors for a specific imaging procedure.

When a short TR is used to obtain a T1-weighted image, the longitudinal magnetiza-tion does not have the opportunity to approachits maximum and produce high signal intensity.In this case, some signal strength must be sacri-ficed to gain a specific type of image contrast.Also, when TR is reduced to decrease imageacquisition time, image noise can become thelimiting factor.


FOV

Matrix

TE

TR

T1

T2

SIGNAL STRENGTH

Coil Design

Distance

Proton Density

Voxel Size

Thickness

Field Strength

Figure 10-3. Factors that affect the signal-to-noise ratios in MR images.

When long TE values are used, the trans-verse magnetization and the resulting signal itproduces can decay to very low values. Thiscauses the images to display more noise.

RF CoilsThe most direct control over the amount ofnoise energy picked from the patient’s body isachieved by selecting appropriate characteristicsof the RF receiver coil. In principle, noise isreduced by decreasing the amount of tissuewithin the sensitive region of the coil. Mostimaging systems are equipped with inter-changeable coils. These include a body coil, ahead coil, and a set of surface coils as shown inFigure 10-4. The body coil is the largest coil andusually contains a major part of the patient’s tis-sue within its sensitive region. Therefore, body

coils pick up the greatest amount of noise. Also,the distance between the coil and the tissuevoxels is greater than in other types of coils. Thisreduces the intensity of the signals actuallyreceived by the coil. Because of this combina-tion of low signal intensity and higher noisepickup, body coils generally produce a lower sig-nal-to-noise ratio than the other coil types.

In comparison to body coils, head coils areboth closer to the imaged tissue and generallycontain a smaller total volume of tissue withintheir sensitive region. Because of the increasedsignal-to-noise characteristic of head coils, rel-atively small voxels can be used to obtain abetter image detail.

The surface coil provides the highest signal-to-noise ratio of the three coil types. Because ofits small size, it has a limited sensitive region


RADIO FREQUENCY COILS

Head

Body

Medium

Noise Received from Body

High Low

Surface

Figure 10-4. Both the amount of noise and the intensity of the signal received depend on the RF receiving coils. The body coil picks up the most noise and the weakest signal,

resulting in the highest noise level in the image.

and picks up less noise from the tissue. Whenit is placed on or near the surface of the patient,it is usually quite close to the voxels and picksup a stronger signal than the other coil types.The compromise with surface coils is that theirlimited sensitive region restricts the useful FOV,and the sensitivity of the coil is not uniformwithin the imaged area. This non-uniformityresults in very intense signals from tissue nearthe surface and a significant decrease in signalintensity with increasing depth. The relativelyhigh signal-to-noise ratio obtained with surfacecoils can be traded for increased image detail byusing smaller voxels.

Receiver Bandwidth

Bandwidth is the range of frequencies (RF) thatthe receiver is set to receive and is an adjustable

protocol factor. It has a significant effect on theamount of noise picked up. This is because thenoise is distributed over a wide range of fre-quencies, whereas the signal is confined to arelatively narrow frequency range. Therefore,when the bandwidth is increased, more noiseenters the receiver. The obvious question is:Why increase bandwidth? One reason is that awider bandwidth reduces the chemical shiftartifact that will be described in Chapter 14.Also, wider bandwidths are the result of shortsignal sampling, or “picture snapping” timesthat are useful in some applications.

Averaging

One of the most direct methods used to con-trol the signal-to-noise characteristics of MRimages is the process of averaging two or more


NSA 4

THE EFFECT OF AVERAGING

Averaging

Images from Four Acquisitions

Figure 10-5. An image with reduced noise is created by averaging the signals from four acquisitions.

signal acquisitions. In principle, each basicimaging cycle (phase-encoding step) is repeatedseveral times and the resulting signals are aver-aged to form the final image as illustrated inFigure 10-6. The averaging process tends toreduce the noise level because of its statisticalfluctuation nature, from one cycle to another.You can think of it as acquiring four images byrepeating the basic acquisition four times. Thenthe signal intensities in each pixel position inthe four images are averaged to produce anintensity value for the new averaged image.

The disadvantage of averaging is that itincreases the total image acquisition time inproportion to the number of cycle repetitions

or number of signals averaged (NSA). The NSAis one of the protocol factors set by the opera-tor. Typical values are 1 (no averaging), 2, or 4,depending on the amount of noise reductionrequired. The general relationship is that theNSA must be increased by a factor of 4 toimprove the signal-to-noise ratio by a factor of2. The signal-to-noise ratio is proportional tothe square root of the NSA. Sometimes thenoise contribution from independent acquisi-tions adds; sometimes it cancels. Since the sig-nals always add, adding or averaging inde-pendently acquired images improves thesignal-to-noise ratio.



Two important image quality characteristics are blurring, which reduces visibility of smallobjects or detail, and image noise, which reduces visibility of low contrast objects. Both of thesecharacteristics depend on design characteristics of the imaging system and the combination ofselected protocol factors. The principal source of blurring in an MR image is the voxel size. This isbecause all tissues within an individual voxel are blurred together and represented by one signal.An image does not display any detail within the individual voxels. It is just a display of a matrixof voxels. Voxel size, and the resulting blurring, can be adjusted with the three protocol factors:FOV, matrix, and slice thickness.

The level of noise that appears in an image depends on the relationship (ratio) of the signalstrength from the individual voxels and the noise strength coming from a region of the patient’sbody. The visible noise is reduced by increasing signal strength. This can be done by increasing themagnetic field strength, increasing voxel size, increasing TR, and decreasing TE. The field strengthis a design characteristic and cannot be changed by the operator. Increasing voxel size to decreasenoise has the adverse effect of also increasing blurring. Voxel sizes must be chosen to provide anappropriate balance between blurring and noise.

FOV

TR

TE Bandwidth

ReducesVisibility

ReducesVisibility

Noise Strength

FromBody

SignalStrength

FromVoxels

Reduces Increases

IncreasesIncreases

Increases

Increases

Decreases

Increases

Produces

VoxelSize

of Small Objects(Detail)

of LowContrast Objects

Sets

Field Strength

Ratio

NSAMatrix

Thickness

Mind Map SummaryImage Detail and Noise

The noise strength picked up from the patient’s body is determined by the mass of tissue con-tained within the sensitive pickup region of the RF coils. Surface coils that cover a relatively smallanatomical region and are also close to the signal source (voxels) produce a high signal-to-noiserelationship that results in lower image noise. The RF receiver bandwidth can be adjusted to blocksome of the noise energy from being received. However, decreasing the bandwidth to reduce noisehas the adverse effect of increasing the chemical shift artifact.

Signal averaging is a useful technique for reducing noise but has the adverse effect of increas-ing acquisition time.


11Acquisition Time And

Procedure Optimization

113

Introduction And OverviewA significant time is usually required for theacquisition of an MR image or a set of multi-slice images. It is generally desirable to keepacquisition time as short as possible for a vari-ety of reasons including: increasing patientthroughput, reducing problems from patientmotion, and performing dynamic imagingprocedures. It is technically possible to reduceacquisition time by changing several protocolfactors such as TR and matrix size. However,the basic protocol factors that affect acquisi-tion time also have an effect on the character-istics and quality of the image. In general,changing factors to reduce acquisition timealso decreases image quality.

When setting up an imaging protocol theacquisition time must be considered withrespect to the necessary requirements for imagequality. An optimized protocol is one in whichthere is a good balance among the image qual-ity requirements and acquisition time.

In this chapter we will consider the vari-ous factors that determine acquisition time,how they relate to image characteristics andquality, and how the various requirements canbe balanced.

Acquisition TimeFigure 11-1 illustrates a basic MR image acqui-sition process and identifies the factors thatdetermine acquisition time. We recall that an

image acquisition consists of an imaging cyclethat is repeated many times. The duration ofthe cycle is TR, the protocol factor that is usedto control certain image characteristics such ascontrast. The primary factor that determinesthe number of times the cycle must be repeatedis the number of rows of k space that must befilled. This, in turn, is determined by the imagematrix size in the phase-encoded direction. Aswe observed in Chapter 9, it is the strength ofthe phase-encoding gradient during each cyclethat directs the signals into a specific row of kspace. Therefore, the phase-encoding gradientmust be stepped through a range of differentstrengths, with the number of steps correspon-ding to the number of rows of k space thatmust be filled. The number of rows of k space

that must be filled is determined by the imagematrix size in the phase-encoded direction. InChapter 10 we saw that matrix size is one of theprotocol factors that determines image detailand also has an effect on image noise. Reduc-ing matrix size in an effort to reduce acquisi-tion time without changing the FOV willdecrease image detail.

In a basic acquisition the time required isTR multiplied by the matrix size in the phase-encoded direction.

Time = TR ¥ matrix size

In an acquisition, the time can be adjusted bychanging either of these two factors. However,we will soon learn that there are some addi-tional factors that are related to certain imaging



Phase-encoding Gradient Steps

Cycles

Acquisition Time

ImageMatrix

“k” SpaceRows

Figure 11-1. The acquisition time is determined by the cycle duration, TR, and number of cycles required to fill the k space.

ACQUISITION TIME

techniques and methods that also have an effecton acquisition time. When applied, some willincrease while others will decrease the time.

Cycle Repetition Time, TRUntil this point, we have used TR as anadjustable protocol factor to control the type ofimage (T1, T2, PD) that is being acquired. Werecall from Chapter 6 that with spin echo imag-ing TR values of approximately 500 msec areused to acquire T1 images, but values as long as2000 msec are required for PD and T2 images. TRis the time that is required, within an individualcycle, for the longitudinal magnetization to relaxand recover to the appropriate level for the typeof image that is being acquired. When the inver-sion recovery method is used, especially for fluidsuppression (Chapter 8), even longer TR valuesare required. The point is that certain TR valuesare required to produce specific image types andthey cannot be arbitrarily shortened to reduceacquisition time. As we learned in Chapter 7,much shorter TR values can be used with thegradient echo methods. This is why gradient

echo methods are generally much faster thanthe spin echo methods but are not appropriatefor many imaging procedures.

Matrix SizeWe recall from Chapter 10 that the matrix sizeand the FOV are the two factors that determinevoxel size in the plane of the image. Voxel sizedetermines the amount of blurring and imagedetail. Small voxels and minimum blurring arerequired for good detail. If the matrix size isreduced without changing the FOV, voxel sizewill be increased and there will be a reductionin image detail. Let us now use Figure 11-2 tosee how matrix size can be adjusted to obtainan optimum balance between acquisition timeand image detail.

It is only the matrix size in the phase-encoded direction that has an effect on acqui-sition time. Recall that this is because thismatrix dimension determines the number ofrows of k space that must be filled. Matrix sizein the frequency-encoded direction does notaffect acquisition time.

ACQUISITION TIME AND PROCEDURE OPTIMIZATION 115

00

00

REDUCED ACQUISITION

“k” Space

Phas

e

Phas

e

Frequency Frequency

Reduced Matrix Size

RectangularFOV

Figure 11-2. Reducing matrix size in the phase-encoded direction reduces the lines of k space to befilled. Then, reducing the FOV to a rectangle reduces the voxel size and restores image detail.

Reduced Matrix in Phase-Encoded Direction

One approach to optimizing an acquisition isto reduce the matrix size in the phase-encodeddirection to a value that is less than the matrixsize in the frequency-encoded direction. It is acommon practice to set the matrix size in thephase-encoded direction to some percentage(such as 80%) of the size in the frequency-encoded direction. This has the effect of reduc-ing acquisition time by 20% with minimaleffect on image detail.

The selectable basic matrix sizes are binarymultiplies such as 128, 256, 512, and 1024.This is necessary in order for the k space data tobe in a form that can be easily processed in theimage reconstruction phase. When basicmatrix size is selected, it is one of these values,with 256 being the most common for mostprocedures. When the matrix size is reducedby some percentage in the phase-encodeddirection, the computer fills in the unused rowof k space with zeros to make the reconstruc-tion process work.

Rectangular Field of View

Decreasing matrix size without changing theFOV does produce an increased voxel size.However, if the FOV can be reduced in thephase-encoded direction, the voxel size will bedecreased and image detail will be maintained.The use of a rectangular FOV is a way of opti-mizing acquisition time and image detail if arectangular FOV works for the specific anatom-ical region that is being imaged.

By combining a reduced matrix size (in thephase-encoded direction) with a rectangularFOV, acquisition time can be reduced withoutaffecting image quality. This is one step in opti-mizing a procedure.

Half AcquisitionOne way of filling k space with a reduced num-ber of acquisition cycles is illustrated in Figure11-3. In addition to half acquisition, this methodmight be referred to by names such as half scanand half Fourier.

This method is based on the fact that in ausual acquisition there is symmetry between thetwo halves of k space. This occurs because of theway the phase-encoding gradients are steppedas illustrated in Figure 11-1. Generally, duringthe first half of the acquisition the gradient isapplied with a high strength for the first cycle.It is then stepped down from cycle to cyclereaching a very low, or zero, value near the cen-ter of the acquisition. During the second half ofthe acquisition the gradient strength is steppedback up to the maximum value for the last imag-ing cycle. In principle, the phase-encoding stepsin the second half are a mirror image of the stepsin the first half. Now let us return to Figure 11-3. When using the half acquisition method onlythe first half of k space is filled directly. Then, thedata that was acquired during the first half ismathematically “flipped” and used to fill thesecond half of k space. This makes it possible tofill all of the rows of k space in approximatelyhalf of the normal acquisition time. The actualacquisition time will be slightly more than halfbecause there must be some overlap in the datato make this process work. This is a method thatcan be used to reduce acquisition time. How-ever, it results in an increase in image noisebecause the image is being formed with asmaller number of acquired signals. It has theopposite effect of using the technique of aver-aging to reduce noise that was introduced inChapter 10. With respect to image noise, usinghalf acquisition is equivalent to setting the NSAto a value of one half.


Signal AveragingWe were introduced to the process of signalaveraging in Chapter 10 as a technique that isused to reduce image noise. We now return tosignal averaging to see how it affects acquisitiontime and how it can be combined with otherfactors to optimize an acquisition procedure. InFigure 11-4 we first see the general relationshipbetween the image acquisition cycles and thephase-encoding gradient steps in a basic acqui-sition. This is an acquisition where neither sig-nal averaging nor any of the fast imaging meth-ods to be discussed later are used. In this basicacquisition one imaging cycle, with duration ofTR, is used for each phase-encoding step. There-fore, the number of acquisition cycles is equal

to the number of lines in k space that must befilled, which is equal to the image matrix size inthe phase-encoded direction. We will now seethat there are acquisition techniques in whichthere is not this one-to-one relationshipbetween the acquisition cycles and the phase-encoding steps.

The first of these is the use of signal averag-ing, which is also illustrated in Figure 11-4. Theaveraging is achieved by repeating the cycle sev-eral times for each phase-encoding step. Thisprocess does not change the number of phase-encoding steps required, but it does increase thenumber of cycles in the acquisition. In theexample shown, the NSA protocol factor is setto 4. The cycle for each phase-encoding step isrepeated four times and the total acquisition


TR TR TR TR TR TR TR

ImageMatrix

“k” SpaceRows

Phase-encoding Gradient Steps

Cycles

Acquisition Time

Figure 11-3. Acquisition time is reduced by using half acquisition and duplicating the data to fill the second half of k space.

HALF ACQUISITION

time is increased by a factor of 4. When averag-ing is used, the acquisition time becomes:

Time = TR ¥ matrix size ¥ NSA.

Protocol Factor InteractionsUp to this point we have observed that each ofthe protocol factors that affect acquisition time(TR, matrix size, and NSA) also have an effecton the image quality characteristics. The rela-tionship becomes somewhat complex becausesome factors affect more than one characteris-tic. A good example is the factors that deter-mine voxel size. As we have seen in Chapter 10small voxels produce high image detail butresult in higher image noise.

In Figure 11-5 we consider three importantimaging goals that are affected by some of thesame protocol factors. These goals are high

image detail, low image noise, and acquisitionspeed. We have placed these goals at the threecorners of a triangle. We can think of this as atype of ball field with three goals. The objectiveis to move the ball closer to the goals. However,as we move the ball in the direction of one goal,we are moving away from the other goals. Inthis analogy the location of the ball representsthe operating point of a specific acquisition pro-tocol and is determined by the combination ofprotocol factors used. An optimized protocol isone in which the selected factors produce thebest balance among high detail, low noise, andspeed for a specific clinical procedure. In someprocedures, high detail might be the mostimportant characteristic, but it is obtained atsome sacrifice of achieving low noise. The noisecan then be reduced by averaging but at the costof an increased acquisition time.


TR TR TR TR TR TR TR TR TR TR TRTR

NSA=4


Phase-Encoding Gradient Steps

No Averaging

Averaging

Cycles

Cycles

Acquisition Time

Figure 11-4. Acquisition time is increased by the factor NSA when averaging is used.

SIGNAL AVERAGING

Developing An Optimized ProtocolAn acquisition protocol can be rather complexbecause of the large number of factors thatmust be adjusted and the interaction of manyof the factors with different image qualitycharacteristics and acquisition times. Oneapproach to developing a good protocol is toaddress the specific image characteristics inthis order:

∑ Contrast Sensitivity∑ Image Detail∑ Spatial Characteristics and Methods∑ Image Noise∑ Artifact Reduction


The first step is to select an imaging methodand factors (generally TR and TE) to give goodcontrast sensitivity and visualization for thespecific tissue characteristic (T1, T2, PD) ortype of fluid movement that is to be observed.

Also, consider if one of the selective signalsuppression techniques (Chapter 8) will behelpful in improving contrast.

Image Detail

The next step is to consider the visibility ofanatomical detail that is required for the spe-cific clinical procedure. This should lead to theselection of an appropriate slice thickness,which is both the largest dimension of a voxeland the one that can be adjusted over a con-siderable range.

FOV is generally determined by theanatomy that is to be covered. A rectangularFOV can be used to permit reduced matrix inthe phase-encoded direction without a reduc-tion in detail.

Using a reduced matrix percentage in thephase-encoded direction can optimize thematrix size.

It is important to not go for more detail thanis actually needed because, as we have seen, thesmall voxels cause an increase in image noise.

Spatial Characteristics and MethodsGenerally, the 2-D multiple slice acquisitionwould be used unless thin slices are requiredfor good detail and then the 3-D volume acqui-sition would be more appropriate.

The number of slices and slice orientationsare determined by the anatomical regions thatare to be covered and the desired views.

Image NoiseVoxel size (slice thickness, FOV, and matrix)has already been set to give the necessaryimage detail. TR and TE values have beenselected to give the desired contrast character-istics. Also, magnetic field strength is a fixedvalue for the system being used. All these fac-tors together, along with the anatomical vol-ume of tissue that is to be imaged and the typeof RF coil selected for the procedure, establisha basic level of image noise. This level of noise


Acquisition Speed

Voxel Size

Matrix(Phase Direction)

small large

512

256

128

1

(Halfscan)

2

4

8

1/2

Numbersof SignalsAveraged

ImageDetail

Signalto Noise

Figure 11-5. Several imaging goals must beconsidered together when selecting

protocol factors.

IMAGING GOALS

might or might not be acceptable. If it is notacceptable, averaging can be applied with theNSA parameter set to the minimum value thatwill give acceptable image quality, keeping inmind that this is increasing acquisition time.

Artifact ReductionOne of the final issues in developing a proto-col is to consider the types of artifacts thatmight occur and then include in the protocolone or more of the artifact reduction tech-niques that are described in Chapter 14.

Fast Acquisition MethodsUntil this point, we have seen that there arethree factors that determine acquisition time:TR, matrix size, and NSA. This applies to all ofthe general methods in which only one line ofk space is filled in each imaging cycle. We recallthat this is when only one phase-encoding stepoccurs in each cycle. There are several methodsthat are capable of filling multiple rows of kspace in one cycle. Two of these methods, echoplanar and GRASE, were described in Chapter7. Both of these are gradient echo methods. It isalso possible to fill multiple rows of k spacewith the spin echo imaging methods by usingthe technique that we are about to describe.When this is done, the methods are generallyknown as fast or turbo spin echo.

There are many reasons to want to speedup an image acquisition process. It can increasepatient throughput and reduce overall operat-ing cost; shorter procedures are more comfort-able for patients; there are less problems withpatient motion. Also, some dynamic studiesrequire the rapid acquisition of a series ofimages. Another advantage of using some ofthe fast acquisition methods is to compensate forother factors that increase acquisition time. Agood example is when using a fluid suppression

technique as described in Chapter 8. Thismethod requires very long TR values thatwould produce extremely long, and impracti-cal, acquisition times. However, by combiningit with a fast acquisition method, it becomes avery useful technique. The 3-D volume acqui-sition method that is used to produce thinslices and high image detail also requires a longacquisition time. This is because multiple repe-titions must be made since phase-encoding isused to cut the slices, as described in Chapter 9.This becomes a more practical method whenused with a fast acquisition technique.

Figure 11-6 illustrates the fast or turbotechnique. We recall from Chapter 6 that it ispossible to produce multiple spin echo signalswithin one cycle by applying several 180∞

pulses after each 90∞ excitation pulse. At thattime we were using each of the echo signals toproduce different images, each with a differentTE value. This is the multiple echo methodthat makes it possible to produce both a PDimage (with a short TE) and T2 images with thelonger TE values in the same acquisition.

Fast or turbo spin echo uses multiple spinecho signals but for a different purpose. Withthis technique, a different phase-encoding gra-dient strength (step) is applied for each of theecho signals. Therefore, each echo signal fills adifferent row of k space, and multiple rows arefilled within each cycle. The result is a reducedacquisition time. The reduction in timedepends on the number of echo signals pro-duced in each cycle and is one of the protocolvariable factors. This number becomes thespeed factor (or turbo factor) and acquisitiontime is now:

Time = TR ¥ matrix size ¥ NSA/speed factor.

Speed factors vary over a considerable range of possible values. In general, there is some


reduction in signal strength (and increasednoise) when high speed factor values are used.

With this technique k space is not filled ina progressive bottom-to-top order. The order offilling can be adjusted. One factor to consideris the TE value for the image. Since severalecho signals, each with a different TE value, areproduced in each cycle and used to form thesame image, what is the “correct” TE for the

image as far as contrast is concerned? Whensetting up the protocol for this method, an“effective” TE value is selected. This is the gen-eral TE value of the signals that are directed tothe more central rows of k space. The data inthese rows determine the general contrastcharacteristics of the image. The data in theouter rows of k space contribute more to thedetail characteristics of the image.


TR TR TR TR

180˚ 180˚ 18090̊

X

Phase-Encoding Gradient Steps

Multiple Echo Signals

Cycles

AcquisitionTime

Figure 11-6. The fast spin echo method reduces acquisition time by producing multiple phase-encoded signals within each cycle.

FAST SPIN ECHO

Mind Map SummaryAcquisition Time And Procedure Optimization


Matrix

Affects VoxelSize

Small Voxels

TR TR TR TR TR

TR TR TR TR TR

TR TR TR TR TR

NSA 4

4

“k” SpaceLines

SpeedFactor

ImageReconstruction

Acquisition Time

Acquisition Time

Acquisition Time

Imaging Cycles

Imaging Cycles

Imaging Cycles

Phase-EncodedDirection

ImageMatrix

The MR image acquisition process consists of an imaging cycle (like a heartbeat) that must berepeated many times. The total acquisition time is determined by the duration of each cycle; thatis the protocol factor, TR, and the number of cycles required. In a basic imaging procedure, thenumber of cycles is determined by the number of rows of k space that must be filled, which, inturn, is determined by the image matrix size in the phase-encoded direction. The matrix size canbe decreased but this results in an increased voxel size and reduced image detail. When setting upan imaging protocol, attention should be given to selecting a voxel size that provides an appro-priate balance between image detail and image noise.

Signal averaging is used to reduce image noise, but it also results in an increased acquisition timebecause cycles must be repeated according to the selected NSA value.

There are several imaging methods that are capable of filling multiple lines of k space duringone imaging cycle. These are some of the fast imaging methods that include fast (turbo) spin echo,echo planar, and GRASE. There is an adjustable speed factor associated with each of these methods.

12Vascular Imaging

123

Introduction And OverviewOne of the important characteristics of MRI isits ability to create an image of flowing bloodand vascular structures without having toinject contrast media. With MRI the contrastbetween the blood and the adjacent stationarytissue is produced by interactions between themovement of the blood and certain eventswithin the imaging process. This is very differ-ent from x-ray angiography where the imageshows the presence of blood (or contrastmedia). In MRI the image displays blood-filledvessels, but it is the movement of the floodthat actually produces the contrast.

It is a somewhat complex process becauseunder some imaging conditions the flowing

blood will produce increased signal intensity andwill appear bright, while under other conditionsvery little or no signal will be produced by theflowing blood and it will be dark. We will desig-nate these two possibilities as “bright blood”imaging and “black blood” imaging, as indicatedin Figure 12-1. There are several different physi-cal effects that can produce both bright bloodand black blood as indicated. These effects canbe divided into three categories:

1. Time effects.2. Selective saturation.3. Phase effects.

Under each of these categories there areseveral specific effects that can produce con-trast. Unfortunately, in addition to producing

useful image contrast, the movement of bloodoften produces undesirable artifacts within theimage. These will be considered in Chapter 14.

Some of the flow effects can appear in vir-tually any image that contains blood vessels.In this chapter we will first explore the vari-ous effects associated with flowing blood andshow how they can be controlled and used.Then we will apply these effects to specificangiographic procedures.

Time EffectsThe time effects are related to the movementof blood during certain time intervals withinthe acquisition cycle. They are sometimesreferred to as the time-of-flight effects. The pro-duction of bright blood is related to the TRtime interval whereas the production of blackblood is related to the TE time interval.

Flow-Related Enhancement(Bright Blood)

The process that causes flowing blood to showan increased intensity, or brightness, is illus-trated in Figure 12-2. This occurs when the

direction of flow is through the slice, as illus-trated. It is also known as the in-flow effect andis used to produce contrast in in-flow contrastangiography. The degree of enhancement isdetermined by the relationship of flow velocityto TR. Three conditions are illustrated. Thearrow indicates the amount of longitudinalmagnetization at the end of each imaging cycle.Because of the slice-selection gradient the RFpulses affect only the blood within the slice.

Let us start by considering using a long TRin the absence of flow in the vessel. The longi-tudinal magnetization regrows to a relativelyhigh value during each cycle, as indicated atthe top of the figure. This condition produces arelatively bright image of both the blood and


Flowing Blood

Time Effects

Phase Effects

Saturation

In-Flow EnhancementFlow Void

Black BloodBright Blood

Intravoxel DephasingFlow CompensationEven-Echo Rephasing

Phase Contrast

Figure 12-1. The physical effects that producecontrast between flowing blood

and the surrounding tissue background.

Short TR

Long TRImage

Short TR

No Flow

No Flow

Flow

RF PulsesHigh Magnetization(Unsaturated)

Low Magnetization

(PartialSaturation)

90˚

Figure 12-2. Flow-related enhancementproduces bright blood because the

background tissue is partially saturated by theuse of short TR values.

FLOW EFFECTS FLOW ENHANCEMENTS

the stationary tissue and little contrast betweenthe two. With this as a reference condition, letus now see what happens when a short TRvalue is used. Each cycle will begin before thelongitudinal magnetization has approached itsmaximum. This results in reduced signal inten-sity and a relatively dark image because boththe non-flowing blood and tissue remain par-tially saturated.

The next step is to see what happens if theblood is flowing into the slice. The effect offlow is to replace some of the blood in the slicewith fully magnetized blood from outside theslice. The increased magnetization at the endof each cycle increases image brightness of theflowing blood. The enhancement increaseswith flow until the flow velocity becomesequal to the slice thickness divided by TR. Thisrepresents full replacement and maximumenhancement and brightness.

In order for blood to appear bright, the sur-rounding background stationary tissue must berelatively dark. Therefore, a part of any brightblood imaging process is to suppress the signalsfrom the surrounding tissue. With the in-floweffect, the background tissue is dark because itis partially saturated from the use of the short TRvalues. When long TR values are used, the back-ground tissue is not suppressed and there is lit-tle, if any, contrast with the flowing blood.

There are several factors that can have aneffect on flow enhancement. In multi-sliceimaging, including volume acquisition, thedegree of enhancement can vary with sliceposition. Only the first slice in the direction offlow receives fully magnetized blood. As theblood reaches the deeper slices, its magnetizationand resulting signal intensity will be reduced bythe RF pulses applied to the outer slices. Slowlyflowing blood will be affected the most by this.Faster flowing blood can penetrate more slicesbefore losing its magnetization. Related to this

is a change in the apparent cross-sectional areaof enhancement from slice to slice. A first slicemight show enhancement for the entire crosssection of a vessel. However, when laminarflow is present, the deeper slices will showenhancement only for the smaller area of fastflow along the central axis of the vessel. Anyother effects that produce black blood cancounteract flow-related enhancement. One ofthe most significant is the flow-void effect,which takes over at higher flow velocities.

Bright blood from flow-related enhance-ment is especially prevalent with gradient-echoimaging. There are two major reasons for this.The short TR values typically used in SAGEimaging increase the flow-related enhance-ment effect. Also, when a gradient rather thanan RF pulse is used to produce the echo, thereis no flow-void effect to cancel the enhance-ment. Therefore, gradient echo imaging is usedwhen using the in-flow process to producebright blood specifically, as in MR angiography.

Flow-Void Effect (Black Blood)Relatively high flow velocities through a slicereduce signal intensity and blood brightnesswith spin echo imaging. Figure 12-3 illustratesthis effect. The flow-void effect occurs becausethe blood flows out of the slice between the 90°and the 180° pulses. The arrow in the slice indi-cates the level of residual transverse magneti-zation that is present when the 180° pulse isapplied to form the spin echo signal. This isthe transverse magnetization produced by thepreceding 90° excitation pulse. The time inter-val between the 90° and the 180° pulses is one-half TE. If the blood is not moving, the bloodthat was excited by the 90° pulse will be withinthe slice when the 180° pulse is applied. Thisresults in maximum rephasing of the trans-verse protons at the echo event and relativelybright blood.

VASCULAR IMAGING 125

If the blood moves out of the slice betweenthe 90° and 180° pulses, complete rephasingand the formation of an echo signal will notoccur. This is because the 180° pulse can affectonly the blood within the slice. The spin echosignal is reduced, and the flowing bloodappears darker than blood moving with alower velocity. The intensity continues to dropas flow is increased until the flow velocityremoves all magnetized blood from the sliceduring the interval between the two pulses(one-half TE).

Selective SaturationAs we have observed on several occasions, sat-uration of the longitudinal magnetization witha 90° RF pulse produces a temporary darknessof a tissue or fluid because there is little mag-netization to produce a signal. We saw this

applied in Chapter 8 as a method for reducingartifacts from moving tissue. Selective satura-tion of a specific anatomical region is an effec-tive way of producing black blood.

This technique is generally known as pre-saturation because the saturation pulse isapplied before the imaging pulses in eachcycle, and is applied to the blood before itenters the image slice. Figure 12-4 illustratesthis concept. An RF pulse is selectively appliedto the anatomical region that supplies blood tothe slice. This pulse destroys the longitudinalmagnetization by flipping it into the trans-verse plane. When this saturated, or unmag-netized, blood flows into the slice a short timelater, it is not capable of producing a signal.Therefore, the image displays a void or blackblood in the vessels. The region, or slab, of pre-saturation can be placed on either side of theimaged slice. This makes it possible to selec-tively turn off the signals from blood flowing inopposite directions.

The presaturation technique can be usedto: (1) produce black-blood images; (2) selec-tively image either arterial or venous flow; and(3) reduce flow-related artifacts, as described inChapter 14.

Phase EffectsThere are two important phase relationshipsthat can be affected by movement of bloodduring the imaging process. One is the phaserelationship among the spinning protonswithin each individual voxel (intravoxel) andthe other is the voxel-to-voxel (intervoxel)relationship of the transverse magnetizations,as shown in Figure 12-5. Both of these con-cepts have been described before. We will nowsee how they are affected by flowing blood,how they contribute to image contrast, andhow they can be compensated for by the tech-nique of flow compensation.


SlowFlow

FastFlow

RF Pulses

Blood Excitedby 90˚ Pulse

No Excited BloodIn Slice

180˚90̊

RF Pulses

180˚90̊

Image

Image

Figure 12-3. The flow-void effect caused byblood flowing out of the slice between

the 90∞ and 180∞ pulses.

FLOW VOID


Flow

Flow

Saturation90˚

Excitation90˚

Figure 12-4. The application of a saturation pulse to eliminate the signals from flowing blood.

Flow Effects

IntervoxelPhase

TransverseMagnetization

Protons

In Phase Dephasing

Decrease andPhase Shift

FlowCompensation

Rephasing

IntravoxelPhase

Figure 12-5. Changes in both intervoxel and intravoxel phases produced by flow.

SATURATION

PHASE EFFECTS OF FLOW

Intravoxel Phase

In order to produce a signal, the protonswithin an individual voxel must be in-phase atthe time of the echo event. In general, dephas-ing and the loss of transverse magnetizationoccur when the magnetic field is not perfectlyhomogeneous or uniform throughout a voxel.A gradient in the magnetic field is one form ofinhomogeneity that produces proton dephas-ing. Since gradients are used for various pur-poses during an image acquisition cycle, thisdephasing effect must be taken into account.For stationary (non-flowing) tissue or fluid theprotons can be rephased by applying a gradi-ent in the opposite direction, as shown inFigure 12-6. We saw how a gradient was usedto diphase and then rephase protons to pro-duce a gradient echo event in Chapter 7. Let usnow consider this process in more detail andapply it to flowing blood.

In general, we are considering events thathappen within the TE interval; that is, betweenthe excitation pulse and the echo event. Werecall that the phase-encoding gradient isapplied during this time. We will use it as theexample for this discussion. When the gradi-ent is applied as shown, the right side of thevoxel is in a stronger magnetic field than theleft. This means that the protons on the rightare spinning faster than those on the left andquickly get out of phase. However, they can berephased by applying a gradient in the reversedirection as shown. Now the protons on theleft are in the stronger field and will be spin-ning faster. They will catch up with and comeinto phase with the slower spinning protonson the right. At the time of the echo event theprotons are in-phase and a signal is produced.The process just described is used in virtuallyall imaging methods to compensate for gradi-ent dephasing.


Excitation

FieldStrength

Flow Compensation

Flow

In-Phase

In-Phase

Dephased

Bright Blood

Dark Blood

Echo EventGradient

Dephasing

Non-movingTissue

FlowingBlood

FlowingBlood

GradientRephasing

Figure 12-6. Dephasing and rephasing produced by gradients and flow compensation.

PHASE CHANGES AND FLOW COMPENSATION

Flow Dephasing

Our next step is to consider what happens tothe protons in a voxel of flowing blood or otherfluid. This is also illustrated in Figure 12-6. Ifthe voxel is moving, the protons will not becompletely rephased by the second gradient.This is because the protons will not be in thesame position relative to the gradient and willnot receive complete compensation. The resultof this is that flow in the direction of a gradientwill generally produce proton dephasing andlittle or no signal at the time of the echo event.This is one possible source of black blood.

Flow Compensation

There are techniques that can be used torephase the protons within a voxel of flowingblood or other fluid. These techniques use asomewhat complex sequence of gradients toproduce the rephasing, as shown in Figure 12-6. The technique is generally called flow com-pensation or gradient moment nulling.

One of the selectable protocol factors con-trols the characteristics of the flow compensa-tion gradients. The technique can be turned onor off and adjusted to compensate for specifictypes of flow. Flow with a constant velocity is theeasiest to compensate. It is possible to compen-sate pulsatile (changing velocity or acceleration)flow by using a more complex gradient wave-form. Flow compensation is somewhat velocitydependent. That is, all velocities are not equallycompensated by the same gradient waveform.

One factor that must be considered is thetime required within the TE interval for theflow compensation process. Some of the morecomplex compensation techniques mightincrease the shortest TE that can be selected.Flow compensation gradients can also be usedto compensate for flow-induced intervoxelphase changes, which are described later.

There are two major applications of flowcompensation: One is to reduce flow-relatedartifacts (Chapter 14) and loss of signal; and theother is as a part of phase contrast angiography,which will be developed later in this chapter.

Even-Echo Rephasing

The phenomenon of even-echo rephasing canbe observed when a multi-spin echo techniqueis used to image blood flowing at a relativelyslow and constant velocity. This effect pro-duces an increase in signal intensity (bright-ness) in the even-echo images (second, fourth,etc.) compared to the odd-echo images.

This effect occurs when there is laminarflow through the vessel and the individual vox-els. This means that the different layers of pro-tons within a voxel are moving at differentvelocities and will experience different phaseshifts as they flow through a gradient. This is adephasing effect that will reduce transversemagnetization and signal intensity at the time ofthe first echo event. The key to even-echorephasing is that the second 180° pulse reversesthe direction of proton spin. Before the pulse,the protons in the faster layers had movedahead and gained phase on the slower movingprotons. However, immediately after the pulsethey are flipped so that they are behind theslower-spinning protons. This sets the stage forthem to catch up and come back into phase.This occurs at the time of the second echo eventand results in an increase in transverse magne-tization, higher signal intensity, and brighterblood than was observed in the first echo image.

The alternating of blood brightness betweenthe odd- and even-echo images will continuefor subsequent echoes, but there will be anoverall decrease in intensity because of the T2decay of the transverse magnetization. Bothturbulent and pulsatile flow tend to increaseproton dephasing within a voxel, which results


in a loss of signal intensity. Under some con-ditions this can counteract the effect of even-echo rephasing.

Intervoxel Phase

If a voxel of tissue moves during the imageacquisition process, the phase relationship of itsspinning transverse magnetization can beshifted relative to that of other voxels. There areboth advantages and disadvantages of this effect.

Phase Imaging

Most MRI systems are capable of producingphase images. The phase imaging that is de-scribed here is different from phase contrastangiography that will be described later. A phaseimage is produced with one of the conventionalimaging methods but with a different way ofcalculating the image from the signals. A phaseimage is one in which the brightness of a voxelis determined by its phase relationship ratherthan the magnitude of transverse magnetiza-tion. This is illustrated in Figure 12-7.

Consider a row of voxels across a vesselthrough which blood is flowing. We will assumelaminar flow with the highest velocity alongthe central axis of the vessel. As this bloodflows through a magnetic field gradient, thephase of the individual voxels will shift in pro-portion to the flow velocity. There, if we createan image at a specific time, we can observe thephase relationships. If the phase of a voxel’stransverse magnetization is then translatedinto image brightness (or perhaps color), wewill have an image that displays flow velocityand direction. This type of image is somewhatanalogous to a Doppler ultrasound image inthat it is the velocity that is being measuredand displayed in the image.

Analytical software can be used in conjunc-tion with phase images to calculate selectedflow parameters.

Artifacts

Flow and other forms of motion can produceserious artifacts in MR images. Although thetechniques that can be used to suppress arti-facts will be described in Chapter 14, it is appro-priate to consider the source of one such arti-fact here. We recall that the process of phase-encoding is used to produce one dimension inthe MR image. A gradient is used to give eachvoxel in the phase-encoded direction a differ-ent phase value. This phase value of each voxelis measured by the reconstruction process(Fourier transform) and used to direct the sig-nals to the appropriate image pixels, asdescribed in Chapter 9. This process worksquite well if the tissue voxels are not movingduring the acquisition process. However, if avoxel moves through a gradient, the phase rela-tionship of its transverse magnetization will bealtered, as described above. This means that theRF signal will no longer carry the correct phaseaddress and will be directed to and displayed inthe wrong pixel location. The observable effectis streaking or ghost images in the phase-encoded direction. Although any type of tissuemotion can produce this type of artifact, it is a


Velocity

PHASE IMAGE PIXEL BRIGHTNESS

Blood Flow

Figure 12-7. Intervoxel phase changes can beused to measure and image blood flow.

significant problem with flowing blood. Flowcompensation as described above, and othertechniques to be described in Chapter 14 areused to reduce this type of artifact.

AngiographyThe MR angiography methods use a combina-tion of the effects described above to producevascular images. Generally, MR angiography isused to produce vascular images covering athick anatomical volume as opposed to a rela-tively thin slice as in most other imaging meth-ods. There are several different approaches usedto produce MR angiograms. Each method hasspecific characteristics that must be consideredwhen producing clinical images. The three gen-eral methods based on how the contrastbetween blood and background tissue are pro-duced are:

1. Phase Contrast Angiography.2. In-Flow Contrast Angiography.3. Contrast Enhanced Angiography.

The first two methods derive contrast fromthe movement of the blood and the floweffects that have just been described. The thirdmethod uses administered contrast media toenhance the vascular contrast.

Phase Contrast Angiography

We recall that when blood flows through amagnetic-field gradient, the phase relationshipis affected. This applies both to the phase rela-tionship of protons within a voxel and thevoxel-to-voxel relationship of transverse mag-netization. Both of these—intravoxel andintervoxel—phase effects can be used to pro-duce contrast of flowing blood. However, theintervoxel phase shift of the transverse mag-netization is the effect most frequently used toproduce phase contrast angiograms.

When blood flows through a gradient, thephase of the transverse magnetization changesin proportion to the velocity. When this tech-nique is used, the phase shift, not the magni-tude of the magnetization, is used to create theimage. Therefore, blood brightness is directlyrelated to flow velocity in the direction of theflow-encoding gradient.

Figure 12-8 illustrates the basic process ofcreating a phase contrast angiogram. At leasttwo image acquisitions are required. Oneimage is acquired with a flow phase-encodinggradient turned on. The phase of the magneti-zation is shifted in proportion to the flowvelocity. A flow compensation gradient isapplied during the acquisition of a secondimage to reset the phase of the flowing blood.The phase in the stationary tissue is notaffected and is the same in both images.

The mathematical process of phase or vec-tory subtraction used to produce the phasecontrast image is illustrated in Figure 12-9. Thevector subtraction process determines the dif-ference in phase between the flow-encodedblood and the flow-compensated blood, whichserves as a reference. The phase difference (and


Flowing

Images

PHASE SUBTRACTION

FlowSensitive

FlowCompensated

Stationary

Figure 12-8. The principle of phase contrastangiography. Blood brightness is related to

the phase change produced by flow.

blood brightness) increases with velocity. How-ever, when the phase reaches the 6 o’clockposition, additional increases in velocity willproduce a reduced difference in phase andbrightness as the phase returns to the 12 o’clockreference position. The contrast in this type ofimage is related to both the velocity and direc-tion of the flowing blood.

Velocity

Because the degree of phase shift and the result-ing contrast is related to flow velocity in acycling manner, the operator must set a veloc-ity value as one of the protocol factors. This willgive a specific relationship between phase val-ues and velocity for that particular imagingprocedure. Flow at this rate will produce maxi-mum contrast as shown in Figure 12-10. Theimportant thing to note is that at some veloci-ties, both below and above the set velocity, theblood will be black. This can produce an alias-ing artifact in which fast-flowing blood will bedark and will look like, or alias, slow-flowingblood. This must be taken into account bothwhen selecting a set velocity value and whenviewing images. An advantage of phase con-trast angiography is the ability to image specificranges of flow, including relatively slow flowrates if the proper factors are used.

Flow Direction

Phase contrast is produced only when the flowis in the direction of a gradient. To image


LowVelocity

HighVelocity

PixelBrightness

PHASE SUBTRACTION

Signal Intensity(Blood Brightness)

Low HighVelocity

Set Velocity

PHASE SUBTRACTION

Figure 12-9. The principle of phase subtraction.

Figure 12-10. The relationship of blood brightness to flow velocity in phase contrast angiography.

blood that is flowing in different directions,several image data sets must be acquired withgradients in the appropriate directions. Flow inall possible directions can be imaged by acquir-ing three images with three orthogonal gradi-ent directions.

The images for the different flow directionsare combined with the subtraction process toproduce one composite angiographic image.

In-Flow Contrast Angiography

With the in-flow contrast method an angio-graphic image is created by using the flow-related enhancement principle described pre-viously. This is also called the time-of-flighttechnique. Bright blood images are producedby using a gradient echo method with rela-tively short TR values. Because of the short TRvalues, the background tissue remains partiallysaturated (and dark) while the flowing bloodproduces a stronger (bright) signal because ofthe in-flow effect.

To produce an angiogram, the methodmust be extended to cover an anatomical vol-ume rather than a single slice. There are twoacquisition techniques that can be used toachieve this: 3-D volume acquisition or multi-ple 2-D slice acquisitions. Each has its specialcharacteristics that must be considered in clin-ical applications.

Three-Dimension (3-D) Volume Acquisition

A 3-D volume acquisition can be used to cre-ate an angiographic image, as shown in Figure12-11. We recall (from Chapter 9) that withthis acquisition method RF pulses are applied toand signals are acquired from an entire volumesimultaneously. During acquisition a phase-encoding gradient is applied in the slice selec-tion direction. The individual slices are thencreated during the image reconstructionprocess. The principal advantage is the abilityto create thin, contiguous slices with smallvoxel dimensions. This gives good image


Flow

RF Pulses

ANGIOGRAPHIC IMAGE

Figure 12-11. A 3-D volume acquisition showing decreased magnetization and blood brightnessproduced by saturation as the blood flows deeper into the volume.

detail. Also, in vascular imaging small voxelsreduce intravoxel dephasing and signal loss.

The principal disadvantage of this methodis that relatively slow-flowing blood becomessaturated as it passes through the acquisitionvolume. This can limit the volume size andvessel length, which is capable of producinggood image contrast. Blood flowing with a rel-atively high velocity, as along the central axisof a vessel, will be imaged deeper into theacquisition volume than slow-flowing blood.This is illustrated in Figure 12-11.

Two-Dimension (2-D) Slice Acquisition

With this technique the anatomical volume is covered by acquiring a series of single-sliceimages, as shown in Figure 12-12. This approachprovides relatively strong signals (bright blood)throughout the volume because each slice is

imaged independently and not affected byblood saturated in other slices. This makes itpossible to image relatively long vesselsextending over large volumes. Also, with thismethod, the signals from the flowing blood arerelatively independent of flow velocity.

Image Format

The image acquisition and reconstructionprocess results in data in the form of manyslice images covering a volume within thepatient’s body. This must then be reformattedinto a 3-D image of the vascular structureextending over this volume. There are severalmethods that can be used for this purpose.

Maximum Intensity Projection (MIP)

The maximum intensity projection (MIP) tech-nique is commonly used to create a single com-


Flow

RF Pulses

ANGIOGRAPHIC IMAGE

Figure 12-12. The acquisition of thin slices reduces the saturation and loss of signal intensity within the imaged volume. Compare to figure 12-11.

posite image from a stack of images, as shownin Figure 12-13. This is a mathematical processperformed by the computer. The stack ofimages is “viewed” along a series of pathwaysthrough the volume. The maximum signalintensity, or blood brightness, encountered inany slice along each pathway is then projectedonto the composite image. The resulting MIPimage is a 2-D image of the 3-D vascular struc-ture. This process generally enhances the con-trast between the flowing blood and the sta-tionary tissues.

It is possible to create images by projectingin different directions through the volume.This gives the impression of rotating the vas-cular structure so that it can be viewed from dif-ferent directions.

Surface Rendering

Images can also be formed by using surfacerendering programs. This is especially usefulfor large vessels such as the aorta.

Contrast Enhanced Angiography

Images of vascular structures can be producedby injecting contrast media and not relying onthe time and phase effects described previouslyto produce the image contrast. This overcomessome of the problems of signal loss producedby saturation in large areas and the limitedvelocity range than can be imaged with phasecontrast. It is, however, restricted in terms ofdepending on the presence of contrast mediain the vessels for continuous imaging.


Stack of Slice Images

Maximum Intensity Projection Image

Figure 12-13. The use of maximum intensity projection (MIP) to produce a composite image from a stack of slice images.


Mind Map SummaryVascular Imaging

90˚

MIP

3-D Volume Acquisition

2-D Slice Acquisition

In-flowEnhancement

PhaseContrast

IntravoxelDephasing

Black Blood

FlowVoid

Bright Blood

Time Effects

Saturation

Phase Effects

BackgroundPhase

Subtraction

Angiography

Velocity

Limits Coverage

Flowing Blood

MR is capable of producing images of flowing blood without the injection of contrast media.The visibility or contrast of the blood is produced by a variety of physical effects or interactionsbetween the moving blood and the MR process. Some effects produce black blood and others pro-duce bright blood. Black blood occurs when the flowing blood does not produce a signal. This canbe caused by the flow void effect, the application of saturation pulses to a vascular area, or anycondition that produces intravoxel dephasing.

Bright blood effects, used in angiography, are in-flow enhancement and phase contrast. In-flow enhancement produces bright blood by keeping the surrounding stationary tissue partiallysaturated. Blood flowing into the image area is not saturated and produces a bright signal. How-ever, consideration must be given to how the image slices are formed and acquired over an anatom-ical region. 3-D volume acquisition produces thin slices and high image detail, but results in par-tial saturation of the flowing blood and possible deterioration of the image. An alternative methodthat produces less saturation of the flowing blood is the 2-D slice acquisition process.

In phase contrast angiography the transverse magnetization of flowing blood experiences aphase shift when passing through a gradient. This phase change, which is proportional to veloc-ity, can be measured and translated into brightness in an image. Most MR angiographic imagesrepresent a 3-D anatomical region and are produced from a stack of slice images by the MIP processor by surface or volume rendering image processing.

13Functional Imaging

137

Introduction And OverviewMR functional imaging consists of several dif-ferent imaging methods that are used to visu-alize and, in some cases, quantify blood andfluid movement beyond the general vascularsystem. In the last chapter we learned thatthere are several techniques that can be usedto produce contrast and a visible image ofblood flowing through vessels. It is the perfu-sion throughout the capillary bed and then thediffusion of fluids throughout the tissue that isthe subject of most MR functional imagingprocedures. Specific and different methods areused to image both the perfusion and diffusionprocesses and to visualize tissue areas that aremetabolically active.

Diffusion ImagingDiffusion imaging is generally used to visualizetissue areas in which some pathologic processhas altered the motion of fluid that is outsideof the vessels and capillaries. This is the natu-ral, random, incoherent, Brownian motion ofthe proton-containing molecules. The motionis always there, but its rate is what changes andis the source of contrast in diffusion imaging.

Brownian Motion and DiffusionFree or unbound molecules in a material suchas fluid are in constant motion because of ther-mal activity. This is a random motion in whichthe molecules travel in one direction, collidewith other molecules and change direction,

and then move in another direction until thenext collision. This is known as Brownianmotion and is illustrated in Figure 13-1. It isthis motion that is the source of the generalprocess of diffusion. Diffusion is whataccounts for the net movement of a substancefrom an area of high concentration to an areaof lower concentration in the absence of otherpressures or forces. Within a region with a uni-form concentration, such as a voxel, the mol-ecules are still in motion even if there is no netmovement from one region to another.

The distance that the molecules move in aspecific time is very dependent on structuralcharacteristics of this tissue. The motion isgenerally greatest in fluids and somewhat lessin a cellular tissue environment.

Diffusion Coefficient

The diffusion coefficient, D, is related to thenet displacement of molecules in a giventime. The net displacement distance is muchless than the total path length traveled by amolecule during this time. This is a staticalprocess in which there is a range of distancestraveled by molecules in the same period oftime. The mean (rms) distance traveled iswhat is related to the diffusion coefficient asshown in Figure 13-1.

Apparent Diffusion Coefficient(ADC)

Diffusion coefficient values determined by MRImight be a composite from several structural


2Dt

2Dt

High Diffusion

Low Diffusion

DIFFUSION IN TWO TISSUECOMPARTMENTS OF A VOXEL

Figure 13.1. The average distance traveled by a molecule in time, t, depends on the diffusioncoefficient (D) for the specific tissue compartment within a voxel.

compartments (extracellular and intracellular)within a voxel. There could be different diffu-sion coefficient values within these variouscompartments. Therefore, the values deter-mined are designated as the apparent diffusioncoefficient, ADC.

Diffusion Direction

In a specific tissue the diffusion rate might bedifferent in different directions because of theorientation of certain tissue structures. This isan important factor that must be taken intoaccount when producing diffusion images. As we will soon discover, the sensitivity forobserving diffusion is produced by applyingmagnetic field gradients. Only when a gradi-ent is applied in the direction of the diffusionwill it be imaged.

The Imaging Process

Diffusion imaging is based on the principlethat the diffusion motion of the moleculesproduces a dephasing of the spinning protonswithin a voxel and that this results in areduced signal intensity and image brightness.The dephasing is actually produced by applyingadditional diffusion-sensitizing gradients dur-ing the image acquisition cycle as shown inFigure 13-2. Here we see the gradients used inconjunction with a spin echo pulse sequence.Notice that there are actually two gradientpulses. One is applied before the 180∞ RF pulseand the second gradient is applied after thepulse. Let us recall that a gradient is a variationin field strength across each individual voxel.During the time of the gradient the spinningprotons will be in different field strengths and

FUNCTIONAL IMAGING 139

RFPULSE

180˚RF

PULSE

90˚

TE

DIFFUSION ACQUISITION

G

DT

Diffusion Sensitivity, b

Diffusion-SensitizingGradients

ECHOEVENT EPI

Acquisition

Figure 13.2. The application of gradients to produce dephasing and reduced signal intensity related to diffusion.

spinning at different rates, along the directionof the gradient. This produces a dephasing ofthe protons within the voxel. When the 180∞RF pulse is applied, it reverses the spin direc-tion. Now when the second gradient isapplied, it produces a rephasing on the spin-ning protons within the voxel. However, onlythe protons that have not moved or changedposition between the times of the two gradi-ents will be completely rephased. The protonsin molecules that have moved will be in a dif-ferent location and field strength during thesecond gradient and will not be completelyrephased. This results in reduced signal inten-sity from voxels containing the moving pro-tons. It is this reduction in signal intensity that

produces the contrast of the diffusing mole-cules with respect to the non-moving tissuestructures.

The reduction in signal intensity (SD/S0)produced by the diffusion depends on two fac-tors: the rate of diffusion expressed by thevalue of the diffusion coefficient, D, and thediffusion sensitivity, b, which is determined bycharacteristics of the gradients. The relation-ship is given by:

SD/S0 = e-bD

Diffusion Sensitivity

The diffusion sensitivity, b, is determined bythe strength, duration, and time separating the


I

DIFFUSION IMAGING

No Diffusion

Tissue with Low Diffusion

Tissue with High Diffusion

Diffusion-WeightedContrast

Calculate ADC

Diffusion Sensitivity, b

Rela

tive

Sign

al In

tens

ity

Figure 13-3. The relationship of signal intensity to the diffusion rates in tissues and the setting of the diffusion sensitivity protocol parameter.

two gradients. These factors are identified inFigure 13-2. The sensitivity is very dependenton the gradient strength, G. Maximum gradi-ent strength is a design characteristic of theimaging system. Diffusion imaging is generallylimited to systems equipped with strong, high-performance gradients.

The value of the diffusion sensitivity, b, isan adjustable protocol factor. Images can beacquired with different sensitivity values.

Diffusion-Weighted Images

Images in which the diffusion rate is a source ofcontrast can be obtained by applying the dif-fusion-sensitizing gradients as described. Thesensitivity can be adjusted by changing the bprotocol factor. As the value of b is increased,diffusion has an increased weighting and areaswith relatively high levels of diffusion becomedarker. This is illustrated in Figure 13-3.

ADC Map Images

An alternative to diffusion-weighted images isto calculate the apparent diffusion coefficient,ADC, for each voxel and display the values inthe form of an image. This can be donebecause the slope of the signal intensity ver-sus b value curve is determined by the ADC ofeach specific tissue voxel. By acquiring imagesat several (at least two) b values, the ADC canbe calculated. ADC map images have contrastthat is opposite to diffusion-weighted images.Areas of increased diffusion are bright.

Blood Oxygenation LevelDependent (BOLD) ContrastBlood oxygenation level dependent, or BOLD, isa source of contrast related to neural activity thatis endogenous and does not require the admin-istration of any contrast agent. Advantages are


CONTRAST

Oxyhemoglobin

Stimulation

IncreasedSignal

Deoxyhemoglobin(Paramagnetic)

Brain

BLOOD OXYGEN LEVEL DEPENDENT

Figure 13-4. The BOLD effect for visualizing areas of the brain activated by stimulation.

that it is not invasive and can be used continu-ously to observe dynamic effects.

BOLD is based on the fact that deoxyhe-moglobin is paramagnetic but oxyhemoglobinis not. This means that deoxyhemoglobin hasa higher magnetic susceptibility and producesmore local field inhomogeneities than oxyhe-moglobin does. This results in more rapid pro-ton dephasing in areas of high deoxyhemo-globin concentration.

The BOLD technique is used to visualizeactivated areas in the brain through a series ofevents as illustrated in Figure 13-4. Brain activ-ity produces an increase in metabolism andoxygen consumption. The associated vasodi-latation results in an increase in blood flow.When the flow (oxygen delivery) increasesmore than the oxygen consumption, there is

an increase in the local oxygen level. Theincrease in oxyhemoglobin in relationship tothe deoxyhemoglobin reduces the susceptibil-ity effect and rate of proton dephasing. Thisresults in an increase in the local T2* value.

When imaged with a T2*-weighted acquisi-tion, activated brain areas will produce anincreased signal intensity and brightness. Theincrease in signal intensity is small, on the orderof a few percent. However, by taking the differ-ence between images acquired with and withoutactivation, the areas of activation can beobserved.

Perfusion ImagingBlood flow through the microvasculature, ortissue perfusion, can be evaluated with MRI.


0 10 20 30

Time (sec)

PERFUSION IMAGING

Regional Blood VolumeRegional Blood FlowTransit Times

Voxel withLow Perfusion

Voxel withHigh Perfusion

Rela

tive

Sign

al In

tens

ity

DATA ANALYSIS

Figure 13-5. The reduction in signal intensity produced by the passage of a bolus of contrast media provides information on tissue perfusion.

In the previous chapter we learned that whenblood is flowing through relatively large ves-sels, compared to capillaries, there are severalphysical effects (time effects and phasechanges) that can be used as sources of con-trast and vessel visualization. These effects donot work for the small capillaries that distrib-ute the blood throughout the tissues. There-fore, other methods for producing contrastmust be used. The objective is not to visualizethe flow in individual vessels, as in angiogra-phy, but to evaluate the combined flowthrough many microvessels within a volumeof tissue such as a voxel.

A common method for imaging perfusionis illustrated in Figure 13-5. The process isstarted by administering a bolus of contrastmedia to the patient. As the bolus passes

through a specific tissue, there will be anincrease in local magnetic susceptibility. Thisproduces a decreased signal intensity in T2*-weighted images. Voxels with the highest per-fusion will experience the largest reduction insignal intensity as the bolus passes through.

The reduction in signal intensity with thepassage of the bolus through two different tis-sues is illustrated in Figure 13-5. Acquisition ofperfusion data requires fast imaging methods,such as EPI, because images must be acquiredevery few seconds to properly measure thecharacteristics of the bolus passage.

The area under the signal versus timecurve is proportional to the regional bloodvolume. By measuring the transit time of thebolus information on the regional blood, flowcan be obtained.


Mind Map SummaryFunctional Imaging


DIFFUSION PERFUSION ACTIVATION

INTRAVOXEL DEPHASING

Decreases Signal IntensityDecreasedDiffusion BOLD Contrast

CalculateImages

Increases Increases Decreases

Deoxyhemoglobin

Sign

al

Time

Diffusion imaging is based on the principle that the diffusion motion of molecules producesintravoxel dephasing when appropriate gradients are applied. The diffusion sensitizing gradientscan be adjusted to produce different levels of diffusion sensitivity, making it possible to measurethe apparent diffusion coefficient values in voxels of tissue. These values can be displayed as a mapimage. A diffusion-weighted image can also be produced in which areas with decreased diffusionwill appear bright.

Perfusion imaging is achieved by injecting a bolus of contrast media. As the bolus passes throughthe imaged tissue, the intravoxel dephasing is increased and the signal intensity drops. When thedata is collected as a rapid sequence of dynamic images, several different perfusion parameters canbe calculated and displayed as images.

BOLD is an imaging method that can be used to visualize activated areas within the brain. Thecontrast is produced by a shift in the deoxyhemoglobin-oxyhemoglobin ratio resulting from thevasodilatation in the activated area. The signal intensity in the activated area, with increased oxy-hemoglobin, increases because deoxyhemoglobin is a paramagnetic substance that contributes tointravoxel dephasing.

14Image Artifacts

145

Introduction And OverviewThere are a variety of artifacts that can appear inMR images. There are many different causes forartifacts, including equipment malfunctionsand environmental factors. However, most arti-facts occur under normal imaging conditionsand are caused by the sensitivity of the imag-ing process to certain tissue characteristics suchas motion and variations in composition.

There are many techniques that can beapplied during the image acquisition processto suppress artifacts. In this chapter we willconsider the most significant artifacts thatdegrade MR images and how the various artifactsuppression techniques can be employed.

An artifact is something that appears in animage and is not a true representation of anobject or structure within the body. Most MRIartifacts are caused by errors in the spatialencoding of RF signals from the tissue voxels.This causes the signal from a specific voxel tobe displayed in the wrong pixel location. Thiscan occur in both the phase-encoding and fre-quency-encoding directions, as shown inFigure 14-1. Errors in the phase-encodingdirection are more common and larger, result-ing in bright streaks or ghost images of someanatomical structures. Motion is the mostcommon cause, but the aliasing effect can pro-duce ghost images that fold over or wraparound into the image.

Errors in the frequency-encoding directionare limited to a displacement of just a few pix-els that can occur at boundaries between fat andnonfat tissues in which most of the protons arecontained in water.

Motion-Induced ArtifactsMovement of body tissues and the flow offluid during the image acquisition process isthe most significant source of artifacts. Theselection of a technique that can be used tosuppress motion artifacts depends on the tem-poral characteristic of the motion (periodic orrandom) and the spatial relationship of themoving tissue to the image area. Figure 14-2shows the types of motion that can produceartifacts and techniques that can be used toreduce them.

At this point we need to make a clear dis-tinction between blurring and artifacts. Werecall that the principal source of blurring inMRI is the size of the individual voxels. Undersome imaging conditions motion can produceadditional blurring, which reduces visibility of

detail and gives the image an unsharp appear-ance. This is especially true for motion thatcauses a voxel to change location from oneacquisition cycle to another. Blurring occurswhen the signals from an individual voxelcome from the different locations occupied bythe voxel. The signals are smeared over theregion of movement. Artifacts, or ghost images,occur when the signals are displayed at loca-tions that were never occupied by the tissue.

Most of the motion-induced artifacts areproduced by dephasing or phase errors. Werecall from Chapter 12 that flow and otherforms of motion can produce both intravoxeland intervoxel phase problems. Intravoxeldephasing generally results in reduced signalintensity, whereas intervoxel phase errorsproduce artifacts in the phase-encoded direc-tion. The phase-encoding process is especiallyaffected by body motion, which causes aparticular anatomical structure to be in a dif-ferent location from one acquisition cycle toanother. This contributes to the phase errorand to the production of artifacts.


Streaksand

Ghost

Pixel Displacement

Phase Encoding Frequency Encoding

Chemical ShiftWater-Fat ShiftFold Over

orWrap Around

AliasingMotion

ARTIFACTS

Figure 14-1. Classification of the most common MRI artifacts.

Phase-Encoded Direction

Motion artifact streaks and ghosting alwaysoccur in the phase-encoded direction. Prior toan acquisition the operator can select whichdirection in the image is to be phase-encodedas opposed to frequency-encoded. This makesit possible to place the artifact streaks in specificdirections. This is a very helpful technique forprotecting one anatomical area from motionand flow artifacts produced in another area. Itdoes not eliminate the artifacts but orientsthem in a specific direction.

Cardiac Motion

Cardiac activity is a source of motion in severalanatomical locations. The movement of theheart produces artifact streaks through the tho-racic region and blurring of cardiac structureswhen the heart is being imaged. In other partsof the body, pulsation of both blood and cere-brospinal fluid (CSF) can produce artifacts andloss of signal intensity.

Triggering

Synchronizing the image acquisition cycle withthe cardiac cycle is an effective technique forreducing cardiac motion artifacts. An EKG (elec-trocardiogram) monitor attached to the patientprovides a signal to trigger the acquisition cycle.

The R wave is generally used as the refer-ence point. The initiation of each acquisitioncycle is triggered by the R wave. Therefore, anentire image is created at one specific point inthe cardiac cycle. This has two advantages. Themotion artifacts are reduced and an unblurredimage of the heart can be obtained. The delaytime between the R wave and the acquisitioncycle can be adjusted to produce images through-out the cardiac cycle. This is typically done incardiac imaging procedures.

Maximum artifact suppression by this tech-nique requires a constant heart rate. Arrhyth-mias and normal heart-rate variations reducethe effectiveness of this technique.

Cardiac triggering is also useful for reduc-ing artifacts from CSF pulsation. This can beespecially helpful in thoracic and cervical spineimaging.

Flow Compensation

The technique of flow compensation or gradi-ent moment nulling was described in Chapter12. In addition to compensating for blood floweffects, this technique can be used to reduceproblems arising from CSF pulsation, espe-cially in the cervical spine. It actually providestwo desirable effects. The rephasing of the pro-tons within each voxel increases signal inten-sity from the CSF, especially in T2-weightedimages. It also reduces the motion artifacts.

Respiratory Motion

Respiratory motion can produce artifacts andblurring in both the thoracic and abdominal

IMAGE ARTIFACTS 147

SOURCE OF MOTION ARTIFACTS

Regional Saturation

Serial Averaging

OrderedPhase-Encoding

TriggeringFlowCompensation

Flow Periodic Random

Figure 14-2. Various types of motion thatproduce artifacts in MR images

and correction techniques.

regions. Several techniques can be used to sup-press these motion effects.

Averaging

The technique of signal averaging is used pri-marily to reduce signal noise, as described inChapter 10. However, averaging has the addi-tional benefit of reducing streak artifacts aris-ing from motion. If a tissue voxel is moving atdifferent velocities and in different locationsduring each acquisition cycle, the phase errorswill be different and somewhat randomly dis-tributed. Averaging the signals over severalacquisition cycles produces some degree ofcancellation of the phase errors and the arti-facts. There are several different ways that sig-nals can be averaged. Serial rather than paral-lel averaging gives the best artifact suppression.Serial averaging is performed by repeating twoor more complete acquisitions and averaging.Parallel averaging is performed by repeating animaging cycle two or more times for eachphase-encoded gradient step. With serial aver-aging there is a much longer time between themeasurements made at each phase-encodedstep. This gives a more random distribution ofphase errors and better cancellation. As withnoise, increasing the number of signals aver-aged (NSA) reduces the intensity of artifactsbut at the cost of extending the acquisitiontime. The averaging process reduces artifactsbut not motion blurring.

Ordered Phase-Encoding

An artifact reduction technique that is usedspecifically to compensate for respiratorymotion is ordered phase-encoding. We recallthat a complete acquisition requires a largenumber of phase-encoded steps. The strengthof the phase-encoded gradient is methodically

changed from one step to another. In a normalacquisition the gradient is turned on with max-imum strength during the first step and is grad-ually decreased to a value of zero at the mid-point of the acquisition process. During thesecond half the gradient strength is increased,step by step, but in the opposite direction. Thebasic problem is that two adjacent acquisitioncycles might catch a voxel of moving tissue intwo widely separated locations. The location isalso somewhat random from cycle to cycle.This contributes to the severity of the artifacts.

Ordered phase-encoding is a technique inwhich the strength of the gradient for eachphase-encoded step is related to the amount oftissue displacement at that particular instant.This requires a transducer on the patient’sbody to monitor respiration. The signals fromthe transducer are processed by the computerand used to select a specific level for the phase-encoded gradient.


SaturatedRegion

Saturation EchoExcitation

RF Pulses

ImageArea

90̊ 90̊ 180̊

Figure 14-3. The use of regional presaturationto reduce motion artifacts.

REGIONAL SATURATION

Regional Presaturation

Regional presaturation is a technique that hasseveral different applications. In Chapter 12we saw how it could be used with flowingblood to eliminate the signal and produce ablack-blood image. An application of this tech-nique to reduce respiratory and cardiac motionartifacts in spine imaging is shown in Figure14-3. With this technique a 90˚ RF saturationpulse is selectively applied to the region ofmoving tissue. This saturates or reduces anyexisting longitudinal magnetization to zero.This is then followed by the normal excitationpulse. However, the region that had just expe-rienced the saturation pulse is still demagnet-ized (or saturated) and cannot produce a sig-nal. This region will appear as a black void in theimage. It is also incapable of sending artifactstreaks into adjacent areas.

FlowFlow is different from the types of motiondescribed above because a specific structure

does not appear to move from cycle to cycle.This reduces the blurring effect, but artifactsremain a problem.

The flow of blood or CSF in any part of thebody can produce artifacts because of thephase-encoding errors. Several of the tech-niques that have already been described can beused to reduce flow-related artifacts.

Regional Presaturation

Regional presaturation, as described in Chap-ter 12, is especially effective because it turnsthe blood black. Black blood, which producesno signal, cannot produce artifacts.

Figure 14-4 illustrates the use of presatura-tion to reduce flow artifacts. The area of satu-ration is located so that blood flows from itinto the image slice.

Flow Compensation

Flow compensation is useful when it is desir-able to produce a bright-blood image. It bothreduces intervoxel phase errors, the source of

IMAGE ARTIFACTS 149

Saturated Region Image Slice

Excitation

RF Pulses

Saturation Echo

90˚ 90˚ 180˚

Figure 14-4. The use of regional saturation to reduce flow artifacts.

REGIONAL SATURATION

streaking, and restores some of the intravoxelmagnetization and signal intensity.

Aliasing ArtifactsAliasing, which produces foldover or wrap-around artifacts, can occur when some part ofthe patient’s body extends beyond theselected field of view (FOV). The anatomicalstructures that are outside of the FOV appearto “wrap around” and are displayed on theother side of the image, as shown in Figure14-5. This occurs because the conventionalimaging process does not make a sufficientnumber of signal measurements or samples.Because of this undersampling the anatomi-cal structures outside of the FOV produce sig-nals with the same frequency and phase char-acteristics of structures within the image area.This phenomenon is known as aliasingbecause structures outside of the FOV take onan alias in the form of the wrong spatial-encoding characteristics.

Two techniques that can be used to elim-inate wraparound artifacts are illustrated inFigure 14-5. One procedure is to increase thesize of the acquisition FOV and then displayonly the specific area of interest. The FOV isextended by increasing the number of voxelsin that direction. This is described as over-sampling. Under some conditions these addi-tional samples or measurements will permit areduction in the NSA so that acquisition timeand signal-to-noise is not adversely affected bythis technique.

An alternative method of eliminatingwraparound or foldover artifacts is to applypresaturation pulses to the areas adjacent tothe FOV. This eliminates signals and theresulting artifacts.

Chemical-Shift ArtifactsThe so-called chemical-shift artifact causes amisregistration or pixel displacement betweenwater and fat tissue components in the fre-quency-encoded direction, as shown in Figure14-6. The problem occurs because the protonsin water and fat molecules do not resonate atprecisely the same frequency. The shifting ofthe water tissue components relative to fat canproduce both a void and regions of enhance-ment along tissue boundaries.


Presaturation

Oversampling

Saturated Areas

Wraparoundor

Foldover

FOV

DisplayedAcquired

Figure 14-5. The wraparound or foldoverartifact and methods for suppressing it.

FOLDOVER ARTIFACTS

There are several factors that determinethe number of pixels of chemical shift. Knowl-edge of these factors can be used to predict andcontrol the amount of chemical shift that willappear in a clinical image.

Field StrengthWe recall from Chapter 3 that the chemical shiftor difference in resonant frequency betweenwater and fat is approximately 3.3 ppm. This isthe amount of chemical shift expressed as a frac-tion of the basic resonant frequency. The prod-uct of this and the proton resonant frequencyof 64 MHz (at a field strength of 1.5 T) producesa chemical shift of 210 Hz. At a field strength of0.5 T the chemical shift will be only 70 Hz. Thepractical point is that chemical shift increaseswith field strength and is generally more of aproblem at the higher field strengths.

Bandwidth

In the frequency-encoded direction the tissuevoxels emit different frequencies so that theycan be separated in the reconstruction process.The RF receiver is tuned to receive this range offrequencies. This is the bandwidth of thereceiver. The bandwidth is often one of theadjustable protocol factors. It can be used tocontrol the amount of chemical shift (numberof pixels), but it also has an effect on othercharacteristics such as signal-to-noise.

In Figure 14-6 we assume a bandwidth of16 kHz. If the image matrix is 256 pixels in thefrequency-encoded direction, this gives 15pixels per kHz of frequency (256 pixels/16kHz). If we now multiply this by the chemicalshift of 0.210 kHz (210 Hz), we see that thechemical shift will be 3.4 pixels.

IMAGE ARTIFACTS 151

TissueWater Fat

EncodeFrequency

Image

256 Pixels

16 kHz

210 Hz at 1.5 T

70 Hz at 0.5 T

Fat

3.3 ppmSpectrum

Relative Resonant Frequency

16 Pixels/kHzImage Width (pixels)

Bandwidth (kHz)

Figure 14-6. The chemical-shift artifact is related to receiver bandwidth.

CHEMICAL-SHIFT ARTIFACT

The amount of chemical shift in terms ofpixels can be reduced by increasing the band-width. This works because the chemical shift,210 Hz, is now a smaller fraction of the imagewidth and number of pixels.

On most MRI systems the water-fat chemi-cal shift (number of pixels) is one of the proto-col factors that can be adjusted by the operator.On some systems it is designated as bandwidth.When a different value is selected the band-width is automatically changed to produce thedesired shift. Even though the chemical-shift

artifact can be reduced by using a large band-width, this is not always desirable. When thebandwidth is increased, more RF noise energywill be picked up from the patient’s body and the signal-to-noise relationship will bedecreased. Therefore, a bandwidth or chemi-cal-shift value should be selected that providesa proper balance between the amount of arti-fact and adequate signal-to-noise.

Fat suppression is useful for reducing thechemical shift artifact.


IMAGE ARTIFACTS 153

Mind Map SummaryImage Artifacts

P E R IODIC R ANDOM

4

ON

ON

ON

ON ON

Correct Pixel Location

Distorts Affected By

Flow

FlowCompensation

TriggeringGating

Regional Saturation FoldoverSuppression

Bandwidth(Pixel Shift)

Water Fat

Chemical Shift

Frequency-Encoded

Phas

e-En

code

d

PhaseFrequency

OrderedPhase-Encoding

ImageMatrix of Pixels

There are a variety of artifacts that can appear in MR images. Most artifacts occur when the sig-nal from a specific tissue voxel is not displayed in the correct image pixel location. It is possible forthe signal to be displaced in both the phase-encoded and frequency-encoded directions.

Motion of the tissue or fluid produces streaks or ghost images of a pixel in the phase-encodeddirection. This occurs because when motion is present, a specific tissue voxel is in different locationsduring the phase-encoding process from one imaging cycle to another. This produces errors in the phaseencoding. We can think of signals as being given an incorrect phase address. There are several tech-niques that can be turned on during an image acquisition to reduce the motion artifacts.

The chemical shift artifact causes signals from fat to be displaced or shifted with respect to thetissue water signals in the frequency-encoded direction. This happens because the protons in waterand fat resonate at slightly different frequencies. The amount of shift (number of pixels) in animage can be controlled by adjusting the RF receiver bandwidth. However, increasing the band-width to reduce the chemical shift artifact results in an increase in image noise.

Foldover is a type of artifact that can occur when the anatomical region is larger than the imagedregion. The image of the tissue outside the imaged area can be folded over and appear as an artifactin the image. There are techniques that can be turned on to suppress this type of artifact.

15MRI Safety

155

Introduction And OverviewMRI is not an inherently dangerous process foreither the patient or the MRI staff. However,the procedure does produce a physical envi-ronment and uses forms of energy that canproduce injury or discomfort if not properlycontrolled. The potential sources of injury arenot direct biological effects but interactions ofthe magnetic field with other objects, which inturn, might produce injury, and the applica-tion of RF energy to the patient’s body, whichproduces heat. During an imaging procedure apatient is subjected to:

∑A strong magnetic field∑Magnetic field gradients (that are rapidly

changing with time)

∑RF energy∑Acoustic noise (produced by the gradients).

There are potential hazards associated witheach of these if certain levels are exceeded orcertain conditions exist.

The purpose of this chapter is to provide theMRI staff with an understanding of the condi-tions that can produce injury or discomfort andto identify the actions to take (safety procedures)to produce a safe and comfortable examination.

Magnetic FieldsAs we have learned, a patient must be placedin a strong magnetic field in order to performthe imaging procedure. The magnetic field iscapable of producing several effects that can

lead to unsafe conditions. Before consideringthe potential hazards, let us review the basiccharacteristics of a magnetic field.

Physical CharacteristicsThe general characteristics of a magnetic fieldwere described in Chapter 2. Here we willreview those characteristics that relate to safety.

The strength of a magnetic field is a char-acteristic that must be considered. Most clini-cal imaging is performed with magnetic fieldsin the general range of 0.15 T to 1.5 T, butthere are some magnets with higher fieldstrengths now being used for research and forsome clinical imaging applications. As we willsee later, certain safety-related effects are de-pendent on field strength.

Let us recall that there are two generalareas of a magnetic field. One is the strong andrelatively uniform field area within the bore ofthe magnet where the patient is located. Theother is the external and somewhat weakerfield that surrounds the magnet. This externalfield varies in strength with location. It isstrongest where it meets the internal field andgradually becomes weaker with increased dis-tance from the magnet. This variation instrength is a natural gradient in the field.There is no precise point at which the fieldends. For all safety related purposes, the loca-tion at which the field strength drops to 5gauss (the 5 gauss line) is considered to be theouter boundary of the field. Undesirable effectscan be produced in both the internal andexternal field areas.

Magnetic fields can produce both mechan-ical and electrical effects on objects or materi-als located in the field. Both are potentialsources of unsafe conditions.

Biological EffectsA major question in MRI today concerns thepotential of undesirable effects of a magnetic

field on the human body. At this time, there areno indications of any irreversible biological effectsproduced by the magnetic fields used for generalclinical imaging. However, in many minds thereis not a feeling of complete safety. A magneticfield, like x-radiation, is an invisible environ-ment that carries with it a certain mystique inrelation to biological effects. In the early days ofx-ray imaging (over a century ago), a false senseof security led to unsuspected injury to personsreceiving high levels of exposure. It has takenmany years for us to develop a reasonably goodunderstanding of the effects of x-ray and otherforms of ionizing radiation. It is this experiencethat causes many to ask: How safe is exposure tomagnetic fields? Are there unknown biologicaleffects from a magnetic field that are going toappear sometime in the future?

First, we must recognize that there is a majorphysical difference between magnetic fields andionizing radiation. A static magnetic field is nota form of energy that is directly transferred tohuman tissue like the various forms of radiation.We can think of it more as an environmentsomewhat like gravity. There is no basis forassuming that the undesirable effects producedby exposure to high levels of ionizing radiations(such as burns, mutations, and cancer-induc-tion) will also apply to magnetic fields.

Magnetic fields, like gravity, do producesome effects, although they are distinctly dif-ferent from the effects produced by radiation.The observed short-term effects of magneticfields have occurred at relatively high fieldstrengths beyond the general range of mostclinical systems.

Biological effects arising from physicalagents generally are of two forms: deterministicand stochastic. Deterministic effects are thosethat generally occur and are often observable ator shortly after the exposure; it is the severity ofthe effect that increases with the strength of the


physical agent. Burns would be an example. Astochastic effect is one that is not necessarilyproduced by every exposure but where there isa statistical probability of the effect occurring.Generally, it is the probability or the risk of theeffect occurring that increases with the level ofexposure. The induction of mutations and can-cer are the two primary stochastic effects asso-ciated with exposure to ionizing radiation.These effects have not been demonstrated tooccur with exposure to magnetic fields.

Internal Objects

If a patient’s body contains internal ferromag-netic objects, there is a potential for injurywhen the patient is placed within the mag-netic field. As we know, a magnetic field exertsa force or torque on ferromagnetic objects. Thepulling and twisting on objects within thebody can tear tissue, rupture blood vessels, andproduce fatal injuries.

The most common types of objects thatmust be considered are the medically im-planted ones such as prosthetic devices andsurgical clips. In addition, there is the possi-bility of a person having embedded bullets,shrapnel from war wounds, swallowed objects,or pieces of metal embedded during an acci-dent. Even very small metal fragments or par-ticles can produce serious injury if they arelocated in a sensitive site such as the eye. Thereis a well-known case of a metal worker wholost his sight because of a small metal fragmentthat was under his eyelid and ruptured his eye-ball when he was placed in a magnetic field foran imaging procedure.

External Objects

Ferromagnetic objects that are brought intothe external magnetic field will experience aforce that either attracts them to the sides ofthe magnet or propels them into the bore.

Objects are pulled along the natural field gra-dients, from the weak to the strong areas of thefield. This produces the so-called projectileeffect where the objects are accelerated to highvelocities and can injure anyone in their path,typically the patient.

This can occur with objects of all sizes.Accidents have occurred with large objectssuch as forklift tongs, metal chairs, and mopbuckets. Small objects such as medical instru-ments carried in pockets and hairpins can eas-ily become flying projectiles.

Magneto-electrical EffectsOne of the laws of physics is that an electri-cal current will be induced or generated in aconducting material that is moved through amagnetic field, or is located in a changingmagnetic field. In fact, this is the principle ofthe electric generator and the transformer.There is a possibility that motion within thebody, such as cardiac activity, will generatesome internal electrical currents. However,

MRI SAFETY 157

The principal safety procedure to safe-guard against injury from internal objectsis to screen patients by interview andreview of medical records to determine ifthere are internal objects from prior sur-geries or injuries.

The next step is to determine if anydetected objects are potential hazards. Thisis sometimes difficult. The only assurance iswhen a specific type of device or object(such as a surgical clip) has been evaluatedand found to be safe. Current informationon many such devices is generally availableon the World Wide Web.

Safety Procedures

these currents do not appear to produce anyundesirable effects, such as fibrillation, at thefield strengths used for clinical imaging.

The most common effect of this type is anelevated T wave in the EKG signal from apatient being monitored in the magnetic field.This is the time within the cardiac cycle whenthe blood flow velocity is the highest. A volt-age is generated by this flow through the mag-netic field that adds to the normal T wave,making it appear to be enhanced. It does notrepresent a change in cardiac activity.

Activation of or Damage toImplanted Devices

Some implanted electronic devices, such ascardiac pacemakers, have magnetically acti-vated switches. The reason for this is so physi-cians can temporarily turn off the pacemakerfor testing by placing a small magnet on thesurface of the patient near the pacemaker site.The concern is that if a patient with this typeof pacemaker should enter a magnetic field,the pacemaker would be turned off, whichcould be potentially fatal. This type of activa-tion could also occur in the external field sur-rounding a magnet. This is why it is a common

practice to exclude patients with pacemakersfrom the field areas within the 5 gauss lineunless it has been determined that the fieldwill not have an adverse effect on their device.

There is a possibility that some implantedelectronic devices could be damaged by thehigh field strength within the magnet.

GradientsWe recall that a gradient is a non-uniformityin the magnetic field produced by turning onthe various gradient coils. A gradient is a vari-ation in field strength over space, which initself is not a safety problem. However, poten-tial problems are produced at the times whenthe gradients are being turned on and off. Dur-ing these times the magnetic field is changingrapidly with respect to time. A time-varyingmagnetic field induces or generates electrical


The principal safety procedure is propertraining of all persons who enter the roomcontaining the magnet.

Signs should be posted informing patientsand staff about the presence of a strongmagnetic field and the associated hazards.

The imaging staff should monitor all otherpersons who enter the room and who mightbring hazardous objects into the room.

The magnet room should be properlysecured when not under the direct observa-tion and supervision of the imaging staff.

Special attention should be given to ensurethe safety of cleaning and maintenance staffwho might not have adequate safety trainingand who often work during non-operationalhours with minimum supervision.

Safety Procedures

Exclude persons with implanted electronicdevices from the magnetic field unless ithas been determined that there is noadverse effect on a particular device.

Safety Procedures

currents through conductive materials locatedwithin the changing field, as illustrated in Fig-ure 15-1. This includes human tissue as well asany implanted materials or devices.

For a specific material or device theamount of current produced depends on therate at which the magnetic field is changed.This is expressed as the factor, dB/dt, where Bis the symbol for field strength. This is in theunits of tesla per second. The U.S. Food andDrug Administration (FDA) provides guide-lines on dB/dt values that can be considered tobe safe operating levels. A maximum value of6 T/sec applies to an imaging system, but

other, higher values apply to specific imagingconditions.

All MR systems do not have gradients thatchange at the same rate (dB/dt). This is actu-ally the gradient slew rate described in Chap-ter 2. The trend in equipment development isto go to higher values to perform certain typesof acquisitions, such as echo planar imaging.

RF EnergyDuring an imaging procedure pulses of RFenergy are transmitted to the patient’s body.Most of the energy is absorbed by the tissueand is converted to heat that has the potentialof increasing the temperature within the body.This is the physical principle of the microwaveoven. In the microwave oven high-frequencyradio energy (microwaves) are transmitted at ahigh power level to a relatively small mass offood. This high concentration of energy canraise the temperature by hundreds of degrees.

It is assumed that gradient design is with-in appropriate safety limits.

MRI SAFETY 159

Time (t)

MAGNETIC FIELD GRADIENTS

Induced Electrical Currents

Rate of Change of Magnetic Field(dB/dt) in tesla per second

Fiel

d St

reng

th (

B)

Figure 15-1. Electrical currents are generated by rapidly changing magnetic fields (the gradients).

Safety Procedures

In MRI the RF energy applied to a patient’sbody is at a much lower power and is distrib-uted over a larger mass of tissue than in themicrowave oven example. The elevation intemperature produced during MRI dependson many factors. Variations among patientsthat will affect this include body size andblood circulation, which distributes heat andpromotes cooling.

The principal issue associated with animaging procedure is the rate at which theenergy is transferred to the body. The factorsthat determine this are illustrated in Figure 15-2. Recall that the rate of energy transfer is thephysical quantity power and is expressed in theunits of watts. The rate at which temperatureis increased is related to the concentration ofpower in tissue.

Specific Absorption Rate (SAR)

In MRI, the rate at which energy (heat) isdeposited in a unit mass of tissue is expressedin terms of the Specific Absorption Rate (SAR) inunits of watts per kilogram of tissue. For spe-cific tissue conditions, the rate of temperaturerise will be proportional to the SAR. The totalincrease in temperature is determined by theSAR, the duration of the exposure to the RFpulses, and various tissue cooling and heat dis-tribution factors.

Limits have been established that arethought to represent insignificant risk to patientsif the examination times are within establishedranges. These are shown in Figure 15-2.

The SAR is determined by a combinationof factors associated with the MR system and


RADIO FREQUENCY HEATING

Specific Absorption Rate (watts/kilo)Maximum Limits

Power(watts)

Unit Time

Magnetic Field StrengthPulse Flip AnglesPulses Per CycleTR

3.2Averaged

Over Head

8Peak

In AnyGram

of Tissue

0.4AveragedOver Body

Figure 15-2. The factors that determine the SAR, or rate of heat production during an image acquisition.

the acquisition protocol as listed in Figure 15-2. Each RF pulse delivers a quantity of energyto the body. Therefore, the SAR is determinedby the energy in each pulse and the rate atwhich the pulses are applied.

Magnetic Field Strength

The field strength is a factor because theamount of RF energy required to produce apulse with a specific flip angle increases withfield strength. Therefore, a 90∞ pulse deliversmore energy in a 1.5 T field than in a 0.5 T field.

Pulse Flip Angles

The energy of a pulse increases with flip angle. A180∞pulse delivers more energy than a 90∞pulse.

Pulses per Cycle

We have observed that there are many combi-nations of RF pulses used by the differentimaging methods, or pulse sequences. A sim-ple spin echo acquisition might use just twopulses (90∞ and 180∞), while other methods,such as fast or turbo spin echo acquisitions usemany 180∞ pulses. This produces more energydelivered during each imaging cycle.

TR

Since TR is the duration of the imaging cycle,it affects the rate at which energy is deliv-ered. Reducing TR concentrates the pulsesinto a shorter time and results in an increasein the SAR.

Determining the SAR for a Patient

We have just seen that there are many variablefactors that determine the SAR for a specificpatient undergoing an imaging procedure.These include both the size of the patient andthe specific imaging protocol that is being used.

It would be difficult to take all of the factorsand manually calculate the SAR value. Fortu-nately, most MR systems perform this calcula-tion during the preparation of the patient andof the protocol. The SAR value might be dis-played for the operator to visually check and toautomatically prevent the system from pro-ceeding with an imaging protocol that willexceed established SAR limits.

Surface Burns

It is possible for metal objects, such as monitor-ing leads and electrodes that are in contact withthe patient’s body to act as an antenna and topick up some of the RF energy. There have beencases where the concentration of energy at thepoint of contact has produced burns.

Acoustic NoiseThe noise produced by the gradients is loud,especially for high performance gradients per-forming certain types of acquisitions. This isoften uncomfortable for patients, and can be apotential source of longer term effects to hearing.

Be aware of the potential problem anduse only devices and lead configurationsthat have been demonstrated to be safe.Investigate any reports from the patientof discomfort.

MRI SAFETY 161

Safety Procedures

Provide hearing protection and audiodiversions as appropriate.

Safety Procedures


Mind Map SummaryMRI Safety

MagneticField

Heat

SAR

Time

Gradients

ElectricalCurrents

90˚ 180˚ 180˚

Radio Frequency Energy

MagneticField

The MR imaging procedure is generally safe for both patients and staff if certain precautionsare observed. However, there are three, specific physical environments that, if not properly con-trolled, could produce injury. They are: a strong magnetic field, rapid-changing magnetic fieldgradients, and RF energy.

The strong static (non-changing) magnetic field does not appear to produce any undesirabledirect biological effects. However, injury can result from the effect of the magnetic field on metalobjects, both external to and within the patient’s body. Objects, such as surgical clips and otherimplanted devices, can be pulled or rotated by the magnetic field. Magnetic-susceptible objectsbrought into the external field can be pulled into the magnetic field and become projectiles thatcan injure patients.

A gradient, that is a rapidly changing magnetic field when it is being turned on and off, caninduce electrical currents within a patient’s body or equipment attached to the patient. This is whylimits have been established on the rate of change for the gradient fields.

The RF energy applied to a patient’s body is absorbed and converted to heat. This can increasetissue temperatures. The SAR is the quantity that expresses the rate at which energy is beingdeposited. SAR is determined by several factors: the number of pulses in a time interval, the flipangles of the pulses, and the strength of the magnetic field. Limits have been established to min-imize adverse effects.

MRI SAFETY 163

165

A Abdominal regions, 148 Acoustic noise and safety, 161 Acquisition, 49-50, 76, 79

volume, 93-94 Acquisition control, 21-22 Acquisition time, 11, 83, 94, 104, 113-22

definition, 113-15 fast acquisition methods, 120-21 half acquisition, 116-17 matrix size, 115-16

rectangular field of view (FOV), 116 reduced matrix in phase-encoded direction,

116 optimized protocol development, 119-20

artifact reduction, 120 contrast sensitivity, 119 image detail, 119 image noise, 119-20 spatial characteristics and methods, 119

protocol factor interactions, 118-19 signal averaging, 117-18 time of repetition (TR), 115

Active shielding, 20 ADC. See Apparent diffusion coefficient Adipose. See Fat Aliasing artifacts, 150 Anatomical structures, 11, 150 Angiography (vascular flow), 3, 6, 18, 69, 86

x-ray, 123 See also Vascular imaging

Angled slice orientation, 92 Apparent diffusion coefficient (ADC), 138-39, 141 Arrhythmias, 147 Artifacts, 9, 18, 32, 81, 145-53

aliasing artifacts, 150 chemical-shift artifacts, 100, 150-52

bandwidth, 151-52 field strength, 151

foldover, 105 intervoxel phase, 130-31

motion-induced, 86, 87, 146-50 cardiac motion, 147 flow, 149-50 phase-encoded direction, 147 respiratory motion, 147-49

reduction of, 120 Audio diversions, 161 Averaging, 109-10

respiratory motion, 148 Axial images, 6

B Bandwidth in chemical-shift artifacts, 151-52 Bandwidth of receiver, 109 Binary multiples, 116 Biological effects, 156-57

deterministic, 156-57 stochastic, 157

Biological functions, 1 Black blood, 123, 125 Blood

black and bright, 123, 125 movement of, 123

Blood oxygenation level dependent (BOLD)contrast, 141-42

Blurring (detail), 8-9, 10, 11, 103-12, 146 See also Image detail and noise

Body coils, 20, 108 Body of patient, 106 BOLD contrast. See Blood oxygenation level

dependent (BOLD) contrast Bone marrow, 86 Bound proton pool, 85, 86 Bound states, 40 Brightness, 4, 5, 6, 37, 39, 40, 43, 53, 57

of blood, 69, 123, 125 of fat and fluid, 81

Brownian motion and diffusion, 137-38 Buildings, 19 Burns, 161

Index

C Calcium, 27 Cancer, 157 Capillaries, 143 Carbon, 27, 28 Cardiac activity, 157 Cardiac motion, 147

flow compensation, 147 triggering, 147

Cerebrospinal fluid (CSF), 39, 40, 147 Chelation, 46 Chemical elements, 27 Chemical shift, 3-4, 32, 100 Chemical-shift artifacts, 100, 150-52 Circularly polarized mode, 21 Claustrophobia, 16 Coils

gradient, 17 quadrature, 21 radio frequencey (RF), 20, 28

Components. See System components Computed tomography (CT) imaging, 8 Computer functions, 21-22 Computers, 19, 106

image reconstruction, 99-100 Contrast

low, 75 mixed, 76

Contrast agents, 45 diamagnetic material, 44, 45 ferromagnetic material, 19-20, 44, 46 negative, 46 paramagnetic material, 44, 45-46

Contrast angiography. See Contrast enhancedangiography; In-flow contrast angiography;Phase contrast angiography

Contrast enhanced angiography, 131, 135 Contrast enhancement, 75-76 Contrast sensitivity, 7-8, 9, 53-57, 73-75, 119 Contrast (windowing), 22 Contrast-detail curve, 104 Control of image quality characteristics, 7-11 Coolants, 16 Copper, 15, 16 Coronal slice orientation, 92 Crystalline structure, 46 CSF. See Cerebrospinal fluid CT. See Computed tomography (CT) imaging “Curtain of invisibility,” 103, 104 Cycle repetition time. See Time of repetition (TR)

D D. See Diffusion coefficient dB/dt values, 159 Deoxyhemoglobin, 142 Detail (blurring), 8-9, 10, 11, 103-12, 146

See also Image detail and noise Deterministic biological effects, 156-57 Diamagnetic material, 44, 45 Diffusion, 3 Diffusion coefficient (D), 138 Diffusion direction, 139 Diffusion imaging, 137-41 Diffusion sensitivity, 140-41 Diffusion-weighted images, 141 Disease, 1 Display of images, 22, 25 Doppler ultrasound image, 130

E Earth’s magnetic field, 15 Echo event and signals, 53 Echo planar imaging (EPI) method, 76-77, 120,

143, 159 Echo planar imaging (EPI) speed factor, 77 Echoes, 28 Eddy currents, 18 EKG. See Electrocardiogram Electric generators, 157 Electric motors, 21 Electrical conductors, 106 Electrical power, 16 Electrically conducted enclosure, 21 Electrocardiogram (EKG), 147, 158 Electronic equipments, 19 Electrons, unpaired, 46 Electrosurgery units, 106 Elevated T wave, 158 Embedded bullets, 157 EPI method. See Echo planar imaging method Equipment, 19, 21 Equipment operator, 3, 7, 161 Even-echo rephasing in vascular imaging, 129-30 Excitation, 29-30, 31, 36, 38, 53

selective, 91, 92 Excitation/saturation-pulse flip angle, 71-73, 74 Excited state, 29, 36 External field, 18, 19 External objects and safety, 157


F Fast acquisition methods, 120-21, 143 Fast spin echo, 120 Fat tissue, 32, 39, 40, 100, 146

magnetization transfer in, 86 suppression of images, 81, 82-83 and water chemical shift, 152

FDA. See United States Food and DrugAdministration

Ferromagnetic material, 19-20, 44, 46 Fibrillation, 158 Field direction, 14-15 Field strength, 14, 15, 31, 106-7

in chemical-shift artifacts, 151 Field uniformity. See Homogeneity Field of view (FOV), 105, 106, 109, 116, 150 Flip angles, 36, 61

excitation/saturation-pulse, 71-73, 74 Flow compensation, 129

in cardiac motion, 147 vascular imaging, 129

Flow dephasing in vascular imaging, 129 Flow direction in phase contrast angiography,

132-33 Flow in motion-induced artifacts, 149-50 Flow-related enhancement (bright blood) in

vascular imaging, 124-25 Flow-void effect in vascular imaging, 125-26 Fluid movement and image types, 3 Fluids, 39

magnetization transfer in, 86 suppression of images, 81, 83-84 See also Water

Fluorescent lights, 21, 106 Fluorine, 27, 28, 31 Foldover artifacts, 105 Fourier transformation, 22, 91, 98, 100, 130 FOV. See Field of view Free induction decay (FID), 60, 70, 71, 76 Free proton pool, 85, 86 Free states, 40 Frequency, 91 Frequency-encoding, 89, 94-96 Frequency-encoding direction in artifacts, 145,

146 Frequency-encoding gradient, 98 Functional imaging, 18, 137-44

blood oxygenation level dependent (BOLD) contrast, 141-42

diffusion imaging, 137-41 ADC map images, 141 apparent diffusion coefficient (ADC), 138-39

Brownian motion and diffusion, 137-38 diffusion coefficient (D), 138 diffusion direction, 139 diffusion sensitivity, 140-41 diffusion-weighted images, 141 imaging process, 139-40

perfusion imaging, 142-43

G G. See Gauss Gadolinium, 45, 46 Gadolinium contrast media, 84, 143 Gadolinium diethylene triamine penta-acetic acid

(GaDTPA), 46 GaDTPA. See Gadolinium diethylene triamine

penta-acetic acid Gamma-ray, 20 Gauss (G), 15 Ghost images, 100, 145, 146, 147 Gradient coils, 17 Gradient cyle, 98-99 Gradient echo, 17-18, 28, 51, 53, 65, 115 Gradient echo imaging methods, 69-80

advantages and limitations, 69 echo planar imaging (EPI) method, 76-77 gradient and spin echo (GRASE) method, 77-78 magnetization preparation, 78-79 process, 69-71 small angle gradient echo (SAGE) method, 71-

76, 78 contrast enhancement, 75-76 contrast sensitivity, 73-75 excitation/saturation-pulse flip angle, 71-73,

74 Gradient moment nulling, 129 Gradient pulse, 59 Gradient and spin echo (GRASE) method, 77-78,

120 Gradient strength, 98 Gradients, 16-18, 91

and safety, 158-59 GRASE method. See Gradient and spin echo

(GRASE) method Gray matter, 40 Gyromagnetic ratio, 31

HHalf acquisition, 116-17 Half Fourier, 116 Half scan, 116

INDEX 167

Head coils, 20, 108 Hearing effects, 161 Hertz (Hz), 94, 151 Homogeneity, 15, 18, 85

See also Inhomogeneities Hydrogen, 5, 27, 28, 31, 107 Hz. See Hertz

I Image acquisition time, 11 Image artifacts. See Artifacts Image brightness. See Brightness Image characteristics, 91 Image detail and noise, 103-12, 119-20

averaging, 109-10 detail sources, 104-6 field strength, 106-7 noise sources, 106 radio frequency (RF) coils, 108-9 receiver bandwidth, 109, 152 signal-to-noise considerations, 106, 107 tissue characteristics, 107 TR and TE, 107-8 voxel size, 106 See also Detail (blurring); Noise

Image format in angiography, 134-35 maximum intensity projection (MIP), 134-35 surface rendering, 135

Image quality, 7-11, 103, 113 Image reconstruction, 22, 49, 50, 91, 99-100 Image reconstruction phase, 89, 90, 116 Image storage and retrieval, 22 Imaging cycle, 51-53 Imaging equipment, 19 Imaging methods, 50-51 Imaging process, 49-58

acquisition, 49-50 contrast sensitivity, 53-57

proton density (PD) contrast, 55-56, 57 T1 contrast, 53-55, 57 T2 contrast, 56-57

imaging cycle, 51-53 echo event and signals, 53 excitation, 53 time of repetition (TR), 49, 52, 53 time to echo (TE), 52, 53, 57

imaging methods, 50-51 k space, 49 protocol, 50 reconstruction, 49, 50

Implanted objects and devices, 157, 158

In-flow contrast angiography, 133-34 three-dimensional (3-D) volume acquisition,

133-34 two-dimensional (2-D) volume acquisition, 134

In-flow effect, 124 In vivo spectroscopic analysis, 26 Inhomogeneities, 18-19, 60, 70, 75

See also Homogeneity Internal field, 18, 19 Internal objects and safety, 167 Intervoxel phase, 126, 127, 130-31, 146

artifacts, 130-31 phase imaging, 130 vascular imaging, 130-31

Intravoxel phase, vascular imaging, 126, 127, 128 Inversion pulse, 79 Inversion recovery, 64-65, 79

short time inversion recovery (STIR) fatsuppression, 82-83 Ionizing radiation, 156 Iron, 19-20, 45, 46 Iron oxide particles, 46 Isotopic abundance, 27, 28

K k space, 49, 76, 77, 78, 98, 100, 114, 116, 121

L Larmor frequency, 31 Linear polarized mode, 21 Liquid helium, 16 Liquid state, 27 Liver, 40 Longitudinal magnetization and relaxation, 36,

37-40, 51-52, 78-79 See also T1 (longitudinal relaxation time)

Longitudinal relaxation time. See T1 (longitudinalrelaxation time)

Low contrast, 75

M Magnesium, 27 Magnetic direction, 36, 37 Magnetic field, 14-15 Magnetic field shielding, 15, 19-20 Magnetic fields and safety, 155-58 Magnetic flipping, 36-37, 38 Magnetic moment, 4, 26, 29 Magnetic muclei (protons), 4, 25-28


Magnetic relaxation times, 3, 4 Magnetic resonance imaging, 1-12, 14

control of image characteristics, 7-11 definition, 1 fluid movement and image types, 3 hydrogen, 5 protons (magnetic nuclei), 4 radio frequency signal intensity, 4 spatial characteristics, 6-7 spectroscopic and chemical shift, 3-4 tissue characteristics and image types, 3, 5-6 tissue magnetization, 2, 4 See also Safety; System components

Magnetic susceptibility, 44-46 Magnetization, 4 Magnetization preparation, 78-79 Magnetization transfer, 85, 86 Magnetization transfer contrast (MTC), 85-86 Magnetization vector, 35 Magnetized, 25 Magneto-electrical effects and safety, 157-58 Magnets, 15-16 Magnification (zooming), 22 Maintenance staff, 158 Manganese, 45 Mass numbers, 27 Matrix, reduced in phase-encoded direction, 116 Matrix size, 105-6, 114, 115-16 Maximum intensity projection (MIP), 134-35 Mean (rms) distance traveled, 138 Medical equipment, 21 Megahertz (MHz), 31, 95, 151 Metal ions, 45-46 Metal objects, 19, 85, 161 MHz. See Megahertz Microvasculature perfusion, 142-43 Microvessels, 143 Microwave ovens, 159 Milliseconds (msec), 18 Millitesla per meter (mT/m), 18 MIP. See Maximum intensity projection Mixed contrast, 76 Modulators, 21 Molecular oxygen, 45 Molecular size, 39-40 Motion of patient, 113 Motion-induced artifacts, 86, 87, 146-50 Movement of blood, 123 MRI. See Magnetic resonance imaging msec. See Milliseconds mT/m. See Millitesla per meter Multi-slice imaging, 92-93 Multiple spin echo, 61, 63-64

Muscle, 40 Musical instrument strings, 31 Mutations, 157

N Nb-Ti. See Niobium-titanium Negative contrast agents, 46 Neutron-proton composition, 26 Niobium-titanium (Nb-Ti), 16 Nitrogen, 27, 28 Nitroxide free radicals, 45 NMR. See Muclear magnetic resonance Noise, 9, 10, 11, 103-12

See also Image detail and noise; Visual noise Non-ionizing radiation, 20 Normal tissues, 3 NSA. See Number of signals averaged Nuclear alignment, 28-29 Nuclear magnetic effect, 44 Nuclear magnetic resonance (NMR), 14, 25-34

chemical shift, 32 field strength, 31 Larmor frequency, 31 magnetic nuclei, 25-28

isotopic abundance, 28 radio frequency (RF) intensity, 26-27 relative sensitivity and signal strength, 28 relative signal strength, 27 spins, 26 tissue concentration of elements, 27

nuclear magnetic interactions, 28-30 excitation, 29-30, 31 nuclear alignment, 28-29 precession and resonance, 29, 30-31 relaxation, 30

Nuclear medicine procedures, 4 Number of signals averaged (NSA), 110, 116, 117,

120

O Odd-echo images, 129 Optimization of procedure, 113-22 Optimized protocol development, 11, 119-20 Ordered phase-encoding in respiratory motion,

148 Orthogonal directions, 17 Oxygen, 27, 28

molecular, 45 Oxygen level, 142 Oxyhemoglobin, 142

INDEX 169

P Pacemakers, 158 Paramagnetic material, 44, 45-46 Parts per million (ppm), 15, 151 Passive shielding, 19-20 Pathologic conditions, 3, 11, 27, 81, 83, 137 Patient discomfort, 161 Patient environments, 16 Patient motion, 113 Patient safety. See Safety Patient’s body, 106 PD. See Proton density (PD) images Perfusion, 3 Perfusion imaging, 142-43 Permanent magnets, 16 Phase, 91 Phase contrast angiography, 131-33

flow direction, 132-33 velocity, 132

Phase effects in vascular imaging, 126-31 Phase imaging, intervoxel, 130 Phase-encoded direction in motion-induced

artifacts, 147 Phase-encoding, 89, 96-98 Phase-encoding gradient, 98 Phased array, 20 Phosphorus, 27, 28, 31 Pixels, 6, 7, 27, 57, 99, 100, 104, 150 Polarization, 21 Post processing, 22 Postal analogy, 100 Potassium, 27 Power, 160 Power amplifiers, 21 Power monitoring circuit, 21 ppm. See Parts per million Precession and resonance, 29, 30-31 Presaturaion technique, 126 Procedure optimization, 113-22

See also Acquisition time Projectile effect, 157 Prosthetic devices, 157 Protocol factors, 49, 52, 113, 118, 152

interactions of, 118-19 and safety, 161

Protocols, 50, 104 in the acquisition process, 22 optimization of, 11, 119-20

Proton density (PD), 3, 5, 6, 27, 52, 63 contrast, 52, 53, 55-56, 57, 61, 62 gradient echo methods, 71, 75, 78

Proton dephasing, 43-44

Protons (magnetic nuclei), 4 Pulse inversion, 79 Pulse sequences, 51 Pulses, 36, 38

saturation, 78

Q Quadrature coils, 21 Quality of image, 7-11, 103, 113 Quench, 16

R R wave, 147 Radiation, ionizing, 156 Radio broadcast stations, 94 Radio communications devices, 21 Radio frequency (RF) coils, 20, 28, 108, 108-9 Radio frequency (RF) energy, 21, 106

safety, 159-61 Radio frequency (RF) intensity, 26-27 Radio frequency (RF) pulse, 69 Radio frequency (RF) sequence, 61, 63 Radio frequency (RF) signal intensity, 4 Radio frequency (RF) silent condition, 41 Radio frequency (RF) system, 20-21 Radio receivers, 31 Radio signals, 25 Radio transmitters, 106 Radioactive nuclide imaging, 4 Receiver bandwidth, 109 Receiving system, 21, 106 Reconstruction of image. See Image reconstruction Rectangular field of view (FOV), 116 Reduced matrix in phase-encoded direction, 116 Reduction in signal intensity formula, 140 Regional presaturation

and flow in motion-induced artifacts, 149 respiratory motion, 148, 149

Regional saturation, 86-87 Relative sensitivity and signal strength, 28 Relative signal strength, 27 Relaxation, 3, 30, 37, 38, 55, 59, 60

See also Tissue magnetization and relaxation Resistive magnets, 16 Resonance, 25, 29, 30-31

tissue, 14 Resonant characteristic, 25 Resonant frequency, 31, 84, 95-96 Respiratory motion, 147-49

averaging, 148


ordered phase-encoding, 148 regional presaturation, 148, 149

Risetime, 18 rms. See Mean (rms) distance traveled

S Safety, 16, 21, 155-63

acoustic noise, 161 gradients, 158-59 magnetic fields, 155-58

biological effects, 156-57 external objects, 157 implanted devices, 158 internal objects, 167 magneto-electrical effects, 157-58 physical characteristics, 156

radio frequency (RF) energy, 159-61 specific absorption rate (SAR), 160-61 surface burns, 161

training for, 158 warning signs, 158

SAGE method. See Small angle gradient echo(SAGE) method Sagittal images, 6 Saturation, 36, 37, 38, 53

selective, 86 Saturation pulses, 78 Selective excitation, 91, 92 Selective saturation, 86

vascular imaging, 126 Selective signal suppression, 81-88

magnetization transfer contrast (MTC), 85-86 regional saturation, 86-87 spectral presaturation with inversion recovery

(SPIR) fat suppression, 84-85 T1-based fat and fluid suppression, 81-85

fluid suppression, 83-84 short time inversion recovery (STIR) fat

suppression, 82-83 Semi-solid structure, 86 Sheet metal, 21 Shielding

magnetic field, 15, 19-20 in RF system, 21, 106

Shimming, 15, 18-19, 85 Short time inversion recovery (STIR) fatsuppression, 82-83 Shrapnel, 157 Signal acquisition, 89, 91 Signal averaging, 117-18 Signal-to-noise considerations, 106, 107 Silent condition in radio frequency (RF), 41

Single shot EPI, 77 Slew-rate, 18 Slice selection, 91-94 Slice selection gradient, 98 Slices, 6, 7, 17, 91 Small angle gradient echo (SAGE) method, 71-76,

78, 125 Sodium, 27, 28, 31 Solid state, 27 Solid structure, 86 Spatial characteristics of the magnetic resonance

image, 6-7, 11, 89-102 frequency-encoding, 94-96 gradient cyle, 98-99 gradients, 91 image characteristics, 91 image reconstruction, 91, 99-100 phase-encoding, 96-98 resonant frequency, 95-96 signal acquisition, 91 slice selection, 91-94

multi-slice imaging, 92-93 selective excitation, 92 volume acquistion, 93-94

Spatial characteristics and methods, 119 Specific absorption rate (SAR), 160-61 Spectral presaturation, 32 Spectral presaturation with inversion recovery

(SPIR) fat suppression, 84-85 Spectroscopic and chemical shift, 3-4, 32 Spectroscopy, 14, 25, 32 Speed factor, 120-21

EPI, 77 Spin echo, 28, 43, 51, 53, 115

fast and turbo, 120 Spin echo imaging methods, 59-67, 76, 79

inversion recovery, 64-65 method, 61-64

multiple spin echo, 61, 63-64 proton density (PD) contrast, 62 T1 contrast, 62, 65-66 T2 contrast, 63

process, 59-62 radio frequency (RF) sequence, 61, 63

Spin-spin interactions, 43, 60 Spins, 26, 42 SPIR. See Spectral presaturation with inversion

recovery (SPIR) fat suppression Spoiling and T1 contrast enhancement, 76 Staff safety. See Safety STIR. See Short time inversion recovery (STIR) fat

suppression Stochastic biological effects, 157

INDEX 171

Storage and retrieval of image, 22 Strings on musical instruments, 31 Superconducting magnets, 15-16 Superparamagnetic materials, 46 Surface burns, 161 Surface coils, 20, 108 Surface rendering of image format in angiography,

135 Surgical clips, 157 Susceptibility, magnetic, 44-46 System components, 13-24

computer functions, 21-22 acquisition control, 21-22 image reconstruction, 22 image storage and retrieval, 22 viewing control and post processing, 22

gradients, 16-18 eddy currents, 18 functions, 17-18 orientation, 17 risetime and slew-rate, 18 strength, 18

magnetic field, 14-15 field direction, 14-15 field strength, 15 homogeneity, 15, 18

magnetic field shielding, 15, 19-20 active shielding, 20 passive shielding, 19-20

magnets, 15-16 permanent, 16 resistive, 16 superconducting, 15-16

radio frequency (RF) system, 20-21 coils, 20 polarization, 21 receiver, 21 shielding, 21 transmitter, 20-21

shimming, 15, 18-19 tissue magnetization, 13 tissue resonance, 14 See also Magnetic resonance imaging

T T1 (longitudinal relaxation time), 3, 5, 6, 38-40,

46 contrast, 39, 40, 52, 53-55, 57 fat and fluid suppression, 81-85 gradient echo methods, 73, 75, 76, 78 short time inversion recovery (STIR) fat

suppression, 82- 83, 84

spin echo methods, 61, 62, 65-66 values for various tissues, 40 See also Longitudinal magnetization and

relaxation T2 (transverse relaxation time), 3, 5, 6, 46

contrast, 41-43, 52, 56-57 fluid images, 81, 83-84 gradient echo methods, 69, 71, 75, 78 spin echo method, 60, 61, 63, 65 T2* value, 43-44, 45, 60, 63, 70, 75, 76, 142,

143 tissue characteristics, 40, 43 values for various tissues, 40 See also Transverse magnetization and

relaxation T. See Tesla TE. See Time to echo Television reception, 106 Television transmitters, 106 Tesla (T), 15, 159 Thermal energy of the body, 29 Thoracic regions, 148-48 Three dimensional (3-D) volume, 87, 92, 93, 94,

119 Three-dimensional (3-D) volume acquisition, in-

flow contrast angiography, 133-34 TI. See Time interval Time after saturation (TS), 79 Time effects in vascular imaging, 124-26 Time interval (TI), 79 Time of repetition (TR), 21-22, 49, 52, 53, 61, 62,

63, 65, 71, 77, 78, 107-8, 115 Time to echo (TE), 52, 53, 57, 61, 63, 65, 71, 83,

107-8 Time-of-flight effects, 124 Tissue characteristics

image detail and noise, 107 and image types, 3, 5-6

Tissue concentration of elements, 27 Tissue magnetization, 2, 4, 13 Tissue magnetization and relaxation, 2, 4, 13, 35-

48 longitudinal magnetization and relaxation, 36,

37-40, 51- 52, 78-79 magnetic field strength effect, 40, 41 molecular size, 39-40 T1 contrast, 39, 40

magnetic susceptibility, 44-46 contrast agents, 45 diamagnetic materials, 45 ferromagnetic materials, 46 paramagnetic materials, 45-46 superparamagnetic materials, 46


tissue magnetization, 36-37 magnetic direction, 36, 37 magnetic flipping, 36-37, 38

transverse magnetization and relaxation, 36,37, 38, 40-44, 51-52, 59, 60

proton dephasing, 43-44 T2 contrast, 41-43

Tissue perfusion, 142-43 Tissue resonance, 14 Tomographic imaging process, 6 TR. See Time of repetition Training for safety, 158 Transaxial slice orientation, 92 Transformers, 157 Transmitter, 20-21 Transverse magnetization and relaxation, 36, 37,

38, 40-44, 51-52, 59, 60 See also T2 (transverse relaxation time)

Transverse relaxation time. See T2 (transverserelaxation time)

Triggering in cardiac motion, 147 TS. See Time after saturation Turbo factor, 120-21 Turbo spin echo, 120 Two dimensional (2-D) volume, 92, 94, 98, 100,

119 Two-dimensional (2-D) volume acquisition, in-

flow contrast angiography, 134

U Ultrasound, 20 Ultrasound image, 130 United States Food and Drug Administration

(FDA), 159 Unpaired electrons, 46

V Vascular flow (angiography), 3, 6, 18, 69, 86 Vascular imaging, 123-36

angiography, 131-35 image format, 134-35 in-flow contrast angiography, 133-34 phase contrast angiography, 131-33

phase effects, 126-31 even-echo rephasing, 129-30 flow compensation, 129 flow dephasing, 129 intervoxel phase, 126, 127, 130-31, 146 intravoxel phase, 126, 127, 128

selective saturation, 126

time effects, 124-26 flow-related enhancement (bright blood),

124-25 flow-void effect, 125-26

See also Angiography (vascular flow) Vehicles, 19 Velocity in phase contrast angiography, 132 Vertical magnetic fields, 16 Viewing control, 22 Visual noise, 103

See also Noise Volume acquistion, 93-94 Voxel size, 104-5, 106, 109, 115, 119 Voxels, 6, 7, 11, 17, 27, 35, 39, 44, 59, 60, 70, 91,

95, 96, 98, 99, 100

W Warning signs for safety, 158 Water, 32, 39, 40, 100, 146

and fat chemical shift, 152 See also Fluid

Watts, 160 Weighted images, 53

T1, 3, 41, 46 T2, 3, 41-43 See also T1; T2

White matter, 40 Windowing (contrast), 22 Work-station computers, 22 World Wide Web, 157

X x, y, and z directions, 17 X-ray angiography, 123 X-ray imaging, 8, 20, 156

Z Zooming (magnification), 22

INDEX 173

Documents

Magnetic Resonance Imaging - sprawls.org Magnetic Resonance Ima… · Magnetic Resonance Imaging Principles, Methods, and Techniques Perry Sprawls, Ph.D., FACR, FAAPM, FIOMP Distinguished