33
©2016 MFMER | slide-1 3 Tesla Zero TE Imaging of the Hip Benjamin M. Howe M.D. 1 , Stephen M. Broski M.D. 1 , David W. Stanley 2 B.S.R.T, Dan W. Rettmann B.S. 2 , Adam C. Johnson 1 M.D., Heidi A. Edmonson 1 Ph.D., Kimberly K Amrami 1 M.D. 1 Department of Radiology, Mayo Clinic, Rochester MN USA 2 GE Healthcare, Waukesha, WI, USA [email protected] Mayo Clinic, Rochester, MN International Skeletal Society 43 rd Annual Meeting September 7-9, 2016 Technical Report – TR_005

3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

  • Upload
    lamdat

  • View
    221

  • Download
    2

Embed Size (px)

Citation preview

Page 1: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-1

3 Tesla Zero TE Imaging of the Hip Benjamin M. Howe M.D.1, Stephen M. Broski M.D.1, David W. Stanley2 B.S.R.T, Dan W. Rettmann B.S. 2, Adam C. Johnson1 M.D., Heidi A. Edmonson1 Ph.D., Kimberly K Amrami1 M.D. 1Department of Radiology, Mayo Clinic, Rochester MN USA 2GE Healthcare, Waukesha, WI, USA [email protected] Mayo Clinic, Rochester, MN

International Skeletal Society 43rd Annual Meeting September 7-9, 2016 Technical Report – TR_005

Page 2: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-2

Disclosures • David W. Stanley B.S.R.T. is employed by GE

Healthcare • Dan W. Rettmann B.S. is employed by GE

Healthcare • PSD is a works in progress; exams performed

under IRB approved research protocol

Page 3: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-3

Objectives • Review cross sectional and 3D applications in

the structural evaluation of the hip • Discuss zero echo time (ZTE) bone imaging of

the pelvis and hip • Discuss the basics of 3D rendering

Page 4: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-4

Introduction Background

Ganz Clinical Orthopaedics & Related Research. 417:112-120, December 2003.

• Structural hip abnormalities have become a major focus in orthopedic surgery

• Hip dysplasia has long been known to contribute to accelerated osteoarthritis of the hip

• More recently, femoroacetabular impingement has been implicated as a mechanism for the development of osteoarthritis in non-dysplastic hips

Page 5: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-5

Introduction Femoroacetabular Impingement

• Cam-type FAI • Caused by an

aspherical femoral head

• Decreased femoral head-neck offset

• Anterolateral osseous bump results in “pistol grip” finding on radiographs

Page 6: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-6

Introduction Femoroacetabular Impingement

• Cam-type FAI • The alpha angle on

CT and MRI is an indicator of cam morphology

• Normal alpha angle on CT and MRI on axial oblique plane is <55°

Alpha Angle Measured on a Pelvic CT

Performed for the Evaluation of FAI

Page 7: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-7

Introduction Femoroacetabular Impingement

• Pincer type impingement • Acetabular over

coverage • General

• Coxa profunda • Acetabular

retroversion • Focal

• Anterior • Posterior

Page 8: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-8

Introduction Femoroacetabular Impingement

• Pincer type impingement • Various

measurements reported on cross sectional imaging studies

Acetabular Anteversion Measured on a Pelvic CT Performed for the

Evaluation of FAI

Page 9: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-9

Introduction Femoroacetabular Impingement

Banerjee et al. Curr Rev Musculoskelet Med. 2011 Mar; 4(1): 23–32

• The most common FAI type is mixed from cam and pincer morphology and the abnormalities can be complex

• Patients often undergo CT for: • Confirmation of structural abnormalities that

may be difficult to appreciate on radiographs • Creation of 3D images to aid in surgical

planning • Although some measurements can be made on

standard MR sequences, the 3D models cannot

Page 10: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-10

Zero Echo Time Bone Imaging 3D Imaging Processing

Friedman, Tamir, et al. Skeletal radiology 45.3 (2016): 307-321.

• CT is optimal for the creation of 3D images of osseous anatomy:

• High spatial resolution • Rapid acquisition and

availability • High contrast

between bone and soft tissues

3D Image Created from CT

Page 11: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-11

Zero Echo Time Bone Imaging 3D Imaging Processing

Friedman, Tamir, et al. Skeletal radiology 45.3 (2016): 307-321.

• Generation of 3D model from CT requires little time and technologist training

• The drawback of CT is the radiation exposure in the young and typically healthy population with hip pain

3D Image Created from CT

Page 12: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-12

Zero Echo Time Bone Imaging

Du, Jiang, et al. Journal of Magnetic Resonance 207.2 (2010): 304-311 Delso, Gaspar, et al. Journal of Nuclear Medicine 56.3 (2015): 417-422

• Challenges in MR of cortical bone • Cortical bone has a low proton density

(15-20% free water by volume) • Cortical bone has a fast decay time

(T2* of 390 µS at 3T) • High resolution, isometric voxels are

required for 3D rendering

Page 13: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-13

Zero Echo Time Bone Imaging

Wiesinger, Florian, et al. Magnetic resonance in medicine 75.1 (2016): 107-114.

• ZTE Pulse sequence • Nonselective pulse excitation • 3D center out radial sampling • Gradients are not ramped down between

repetitions • Encoding starts immediately at the time of

excitation, TE=8µs

Page 14: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-14

Zero Echo Time Bone Imaging Parameters

• Imaging was performed on 3.0T Discovery MR750W (GE Healthcare, Waukesha, WI, USA)

• 32 channel body array • 40 cm • TR ~ 400ms • Matrix 320 x 320 • Nex = 4 • BW 62.5 kHz • Resolution 1.3-1.6 mm • Scan time ~ 4-6 minutes

Page 15: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-15

ZTE image prior to logarithmic rescaling of the normalized image intensities

Page 16: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-16

Zero Echo Time Bone Imaging

Wiesinger et al. Magnetic resonance in medicine 75.1 (2016): 107-114..

• The intrinsic properties of ZTE creates two histogram intensity peaks:

1. Lower Peak: • Cortical Bone • Gas

2. Upper Peak: • All other tissues

(muscle, fat, tendons, ligaments, fluid)

Page 17: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-17

Zero Echo Time Bone Imaging

Wiesinger, Florian, et al. Magnetic resonance in medicine 75.1 (2016): 107-114 Delso, Gaspar, et al. Journal of Nuclear Medicine 56.3 (2015): 417-422

• Automatic Post Processing: • A histogram-based intensity correction

algorithm is applied as described by Wiesinger et al.

• A logarithmic rescaling is used to enhance cortical bone

• This aids in threshold segmentation

Page 18: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-18

• Non-rescaled axial ZTE image of the pelvis

• Slice Thickness 1.3mm

• Acquisition time 6:11

• Logarithmic rescaled image with greyscale inversion

Page 19: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-19

• Non-rescaled axial ZTE image of the pelvis

• Slice Thickness 1.6mm to shorten time and increase SNR

• Acquisition time 3:58

• Logarithmic rescaled image with greyscale inversion

Page 20: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-20

Standard Cross Sectional Measurements Performed on 2D reformatted ZTE images

Acetabular Anteversion 19.5°

Alpha Angle 55.8°

Page 21: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-21

2D ZTE Femoral Neck Radial Reformats for Head/Neck Morphology

Page 22: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-22

Zero Echo Time Bone Imaging 3D Imaging Processing

Friedman, Tamir, et al. Skeletal radiology 45.3 (2016): 307-321.

• CT is optimal for the creation of 3D images of osseous anatomy:

• High spatial resolution • Rapid acquisition and availability • High contrast between bone and soft tissues

• Generation of 3D model from CT requires little time and technologist training

• The drawback of CT is the radiation exposure in the young and typically healthy population with hip pain

Page 23: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-23

3D Generated Models Maximum Intensity Projections

• Requires minimal processing time

• Lack detail of 3D models created from CT

ZTE MIP

Page 24: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-24

Zero Echo Time Bone Imaging 3D Imaging Processing

Friedman, Tamir, et al. Skeletal radiology 45.3 (2016): 307-321.

• Segmentation: • Refers to the process of converting pixels

from a 2D image set into a 3D object • Requires:

• Isometric voxels • High resolution • High contrast between the bones and soft

tissues

Page 25: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-25

Zero Echo Time Bone Imaging 3D Imaging Processing

• Threshold: • Accepting or rejecting

the pixels within a dataset based on greyscale pixel value

• Limited in standard MR techniques

• Starting point for ZTE bone segmentation

Raw threshold of the hemipelvis from ZTE

Page 26: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-26

3D Image of the Pelvis and Hips Generated from a 1.3 mm ZTE Acquisition

Created with Materialise 3-Matic® and Mimics ® (Materialise, Leuven, Belgium)

Page 27: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-27

3D Image of the Pelvis Generated from a 1.3 mm ZTE Acquisition

Created with Materialise 3-Matic® and Mimics ® (Materialise, Leuven, Belgium)

3D Generated Model with Wrapping and Smoothing

3D Generated Model with Overlying Mesh

Page 28: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-28

3D Image of the Proximal Femur Generated from a 1.3 mm ZTE Acquisition

Created with Materialise 3-Matic® and Mimics ® (Materialise, Leuven, Belgium)

Page 29: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-29

Zero Echo Time Bone Imaging 3D Imaging Processing Limitations

• Fascial Plane Artifacts: • Occur at muscle

interfaces and fascial planes

• Artifact is easy to remove in post processing as it is remote from cortex

Page 30: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-30

Zero Echo Time Bone Imaging 3D Imaging Processing Limitations

• False Positive Bone Voxels:

• Occurs at the surface of the cortex

• Difficult to remove in 3D post processing

• 3D smoothing and wrapping algorithms are helpful, but lead to artifactual bone contours

Page 31: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-31

Zero Echo Time Bone Imaging 3D Imaging Processing Limitations

• Thin Cortical Bone: • Some areas of thin

cortical bone are not detected

• This results in cortical defects in 3D models

Thin Cortex of Posterior Femoral Head

Page 32: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-32

Conclusion • This ePoster reviews our initial

experience with ZTE in osseous evaluation of the pelvis and hips

• As this has only been attempted in several volunteers, further evaluation is required

Page 33: 3 Tesla Zero TE Imaging of the Hip · PDF file3 Tesla Zero TE Imaging of the Hip ... standard MR sequences, the 3D models cannot ©2016 MFMER ... • Non-rescaled axial ZTE

©2016 MFMER | slide-33

References • Ganz, Reinhold, et al. "Femoroacetabular impingement: a cause for

osteoarthritis of the hip." Clinical orthopaedics and related research 417 (2003): 112-120.

• Du, Jiang, et al. "Qualitative and quantitative ultrashort echo time (UTE) imaging of cortical bone." Journal of Magnetic Resonance 207.2 (2010): 304-311.

• Delso, Gaspar, et al. "Clinical evaluation of zero-echo-time MR imaging for the segmentation of the skull." Journal of Nuclear Medicine 56.3 (2015): 417-422.

• Wiesinger, Florian, et al. "Zero TE MR bone imaging in the head." Magnetic resonance in medicine 75.1 (2016): 107-114.

• Friedman, Tamir, et al. "3D printing from diagnostic images: a radiologist’s primer with an emphasis on musculoskeletal imaging—putting the 3D printing of pathology into the hands of every physician." Skeletal radiology 45.3 (2016): 307-321.

• Banerjee et al. Femoroacetabular impingement: a review of diagnosis and management Curr Rev Musculoskelet Med. 2011 Mar; 4(1): 23–32