IMAGING & THERAPEUTIC TECHNOLOGY 1541

AAPM/RSNA Physics Tutorialfor Residents: Topics in CTRadiation Dose in CT1

Michael F. McNitt-Gray, PhD

This article describes basic radiation dose concepts as well as thosespecifically developed to describe the radiation dose from computedtomography (CT). Basic concepts of radiation dose are reviewed, in-cluding exposure, absorbed dose, and effective dose. Radiation dosefrom CT demonstrates variations within the scan plane and along the zaxis because of its unique geometry and usage. Several CT-specificdose descriptors have been developed: the Multiple Scan Average Dosedescriptor, the Computed Tomography Dose Index (CTDI) and itsvariations (CTDI100, CTDIw, CTDIvol), and the dose-length product.Factors that affect radiation dose from CT include the beam energy,tube current–time product, pitch, collimation, patient size, and dosereduction options. Methods of reducing the radiation dose to a patientfrom CT include reducing the milliampere-seconds value, increasingthe pitch, varying the milliampere-seconds value according to patientsize, and reducing the beam energy. The effective dose from CT can beestimated by using Monte Carlo methods to simulate CT of a math-ematical patient model, by estimating the energy imparted to the bodyregion being scanned, or by using conversion factors for general ana-tomic regions. Issues related to radiation dose from CT are being ad-dressed by the Society for Pediatric Radiology, the American Associa-tion of Physicists in Medicine, the American College of Radiology, andthe Center for Devices and Radiological Health of the Food and DrugAdministration.©RSNA, 2002

Abbreviations: CTDI � Computed Tomography Dose Index, DLP � dose-length product, ICRP � International Commission on Radiological Pro-tection

Index terms: Computed tomography (CT), radiation exposure ● Dosimetry ● Radiations, exposure to patients and personnel ● Radiations, measure-ment

RadioGraphics 2002; 22:1541–1553 ● Published online 10.1148/rg.226025128

1From the Department of Radiology, David Geffen School of Medicine, University of California, Los Angeles, Box 951721, Los Angeles, CA 90095-1721. From the AAPM/RSNA Physics Tutorial at the 2001 RSNA scientific assembly. Received July 18, 2002; revision requested August 13 andreceived August 29; accepted August 30. Address correspondence to the author (e-mail: mmcnittgray@mednet.ucla.edu).

©RSNA, 2002

Radio

Gra

phic

s

IntroductionAlthough computed tomography (CT) representsonly a small percentage of radiologic examina-tions, it results in a significant portion of the ef-fective radiation dose from medical procedures(1). With the increasing use of CT for screeningprocedures (2) and advances in scanner technol-ogy, the trend for increased numbers of proce-dures performed with this imaging modality mayincrease. Although CT is clearly providing manyclinical benefits, the motivation to understandradiation dose in general as well as the specificconcepts related to CT grows with the prevalenceof this modality.

The purpose of this article is to review basicradiation dose concepts as well as those specifi-cally developed to describe the radiation dosefrom CT. This involves describing the basic con-cepts of exposure, dose, and effective dose andshowing how these are defined or estimated forCT. After the definitions are given, the effects ofvarious technical CT factors on radiation dose aredescribed. Next, some typical values for eachdose parameter are given for a few select imagingprotocols. Finally, some relevant resources andcurrent activities are discussed.

Radiation DoseMeasures: General Definitions

ExposureThe term exposure (3) describes the ability of xrays to ionize air. It is measured in roentgens (R);this unit is defined as the quantity of x rays thatproduces 2.580 � 10�4 C of charge collected perunit mass (kilograms) of air at standard tempera-ture and pressure (STP): 1 R � 0.000258 C/kgair. This term refers to the concentration, in air,of radiation at a specific point and is the ioniza-tion produced in a specific volume of air. It istypically measured with an ionization chamberand an electrometer. It essentially describes howmuch ionization is present in the volume, but itdoes not tell how much energy is absorbed by thetissues being irradiated.

Absorbed Radiation DoseAbsorbed radiation dose (3), often referred to asradiation dose, describes the amount of energy ab-sorbed per unit mass at a specific point. It is mea-sured in grays (1 Gy � 1 J/kg) or rads (1 rad �100 erg/g). The conversion between rads andgrays is 100 rad � 1 Gy. Absorbed dose essen-tially describes how much energy from ionizing

radiation has been absorbed in a small volumecentered at a point; it does not describe wherethat radiation dose is absorbed or reflect the rela-tive radiosensitivity or risk of detriment to thosetissues being irradiated.

Effective DoseEffective dose (3–5) (formerly referred to as theeffective dose equivalent [6]) takes into accountwhere the radiation dose is being absorbed (eg,which tissue has absorbed that radiation dose)and attempts to reflect the equivalent whole-bodydose that results in a stochastic risk that is equiva-lent to the stochastic risk from the actual ab-sorbed dose to those tissues irradiated in a non-uniform, partial-body irradiation such as a CTscan. It is a weighted average of organ doses, asdescribed in Equation (1):

E � �T�wT � wR � DT,R�, (1)

where E is the effective dose, wT is the tissue-weighting factor, wR is the radiation-weightingcoefficient (1 for x rays), DT,R is the average ab-sorbed dose to tissue T, T is the subscript foreach radiosensitive tissue, and R is the subscriptfor each type of radiation (here, only x rays arepresent). The weighting factors are set for eachradiosensitive organ in Publication 60 of the In-ternational Commission on Radiological Protec-tion (ICRP) (4). Effective dose is measured insieverts (Sv) or rems. The conversion betweensieverts and rems is 100 rem � 1 Sv.

Although methods to calculate the effectivedose have been established (ICRP Publications26 [6] and 60 [4]), these methods depend heavilyon the ability to estimate the dose to radiosensi-tive organs from the CT procedure (DT,R). How-ever, determining the radiation dose to these or-gans is problematic and direct measurement isnot possible.

Radiation DoseMeasures: CT Specific

Because of its geometry and usage, CT is aunique modality and therefore has its own set ofspecific parameters for radiation dose (7–13).This modality is unique because the exposure isessentially continuous around the patient, ratherthan a projectional modality in which the expo-sure is taken from one or two source locations.The modality typically uses thin sections—rang-ing from 0.5-mm to 20-mm nominal beam colli-mation. However, this modality also typically usesmultiple exposures along some length of the pa-tient to cover a volume of anatomy. In addition,

1542 November-December 2002 RG f Volume 22 ● Number 6

Radio

Gra

phic

s

these exposures may be done in sequences ofscans (eg, a series of scans such as pre- and post-contrast).

Variations within the Scan PlaneProjectional radiographic exposures are takenfrom one source position and the entrance skin

dose is much larger than the exit skin dose, creat-ing a large radiation dose gradient across the pa-tient (Fig 1). In contrast, the tomographic expo-sure of CT scans with a full 360° rotation resultsin a radially symmetric radiation dose gradientwithin the patient. That is, in a uniform circularobject, such as a test phantom, all of the points ata certain radius from the center have the same (ornearly the same) radiation dose (Fig 2). As weshall see, the magnitude of that dose gradient (thesize of the difference from center to periphery)will be affected by several factors, including thesize of the object, the x-ray beam spectrum, andthe attenuation of the material or tissue.

For example, in a typical CT dosimetry phan-tom that is 32 cm in diameter and made of poly-methyl methacrylate (PMMA)—usually referredto as the body phantom—measurements of CTdose, which will be defined later, obtained at thecenter are typically about 50% of the measuredvalue obtained at one of the peripheral positions.This result is illustrated in Figure 3, which showsthe center value obtained under specific condi-tions to be approximately 10 mGy while the pe-ripheral values are 20 mGy under those sameconditions. However, for a smaller-diameterphantom—the 16-cm-diameter phantom referredto as the head phantom—measured under theidentical exposure conditions, the center valuereading climbs to approximately 40 mGy, as do

Figure 1. Dose gradient resultingfrom a projectional radiographic ex-posure in which the source is station-ary at one position. The thicker linesrepresent the entrance skin dose,which is much larger than the exitskin dose, represented by the thinnerlines. This difference creates a lineargradient through the patient.

Figure 2. Dose gradient resulting from a full360° exposure from a CT scan. The thicker linesrepresent the entrance skin dose, which is muchlarger than the dose at the inner radius, repre-sented by the thinner lines. This difference re-sults in a radially symmetric radiation dose gradi-ent within the patient.

Figure 3. Typical dose measurements in a 32-cm-diameter (body) phantom from a single-de-tector CT scan. Values measured at the centerand periphery (1 cm below the surface) positionswithin a polymethyl methacrylate circular dosim-etry phantom demonstrate a radial dose gradientwith a 2:1 ratio from periphery to center. Techni-cal factors for the measurements were 120 kVp,280 mA, 1-sec scan (ie, 280 mAs), and 10-mmcollimation.

RG f Volume 22 ● Number 6 McNitt-Gray 1543

Radio

Gra

phic

s

the peripheral values (Fig 4). This indicates thatthe magnitude of the difference from center toperiphery is very much size dependent; it also in-dicates that the absolute values of the absorbeddoses are size dependent.

Z-Axis VariationsIn addition to the variations within the scanplane, there are variations along the length of thepatient or phantom. These can be characterizedby the z-axis dose distribution or radiation profile(Fig 5). This is the distribution of absorbed dosealong the axis of the patient due to a single axialscan (a full rotation at one table position). Theradiation profile is not limited to the primary areabeing imaged, and there are tails to this distribu-tion from the nonideal collimation of the x-raysource and from scatter of photons within the ob-ject being exposed. When multiple adjacent scansare performed, the tails of the radiation profilesfrom adjacent scans can contribute to the ab-sorbed dose outside of the primary area being im-aged. If these tails are significant and are nonzeroat some distance from the location of the originat-ing section, then these contributions can add up,creating additional absorbed dose in the primaryarea being imaged.

That is, the radiation dose in a specific sectionconsists of the sum of contributions to that sec-tion when that area is the primary area being im-aged as well as the contributions from the tails ofradiation profiles from adjacent sections whenother locations are the primary area being im-aged. The size of the contributions from adjacentsections is very directly related to the spacing ofsections and the width and shape of the radiationprofile.

To account for the effects from multiple scans,several dose descriptors were developed. One ofthe first was the Multiple Scan Average Dose(MSAD) descriptor (7,8,14). This is defined asthe average dose resulting from a series of scansover an interval I in length:

MSAD � �1/I� ��I/ 2

I/ 2

Dseries�z� dz, (2)

where I is the interval of the scan length andDseries(z) is the dose at position z parallel to the z(rotational) axis resulting from the series of CTscans.

Following this was the Computed Tomogra-phy Dose Index (CTDI) (15). This was defined

as the radiation dose, normalized to beam width,measured from 14 contiguous sections:

CTDI � �1/nT� ��7T

7T

Dsingle�z� dz, (3)

where n is the number of sections per scan, T isthe width of the interval equal to the selected sec-tion thickness, and Dsingle(z) is the dose at point zon any line parallel to the z (rotational) axis for asingle axial scan. This index was suggested by theFood and Drug Administration and incorporatedinto the Code of Federal Regulations (16).

However, to be measured according to thedefinition, only 14 sections could be measuredand one had to measure the radiation dose pro-file—typically done with thermoluminescent do-simeters (TLDs) or film, neither of which was

Figure 4. Typical dose measurements in a 16-cm-diameter (head) phantom from a CT scan.Values measured at the center and periphery (1cm below the surface) positions within a poly-methyl methacrylate circular dosimetry phantomdemonstrate essentially no radial dose gradient.Technical factors for the measurements were 120kVp, 300 mA, 1-sec scan (ie, 300 mAs), and5-mm collimation.

Figure 5. Radiation profile of a full-rotation CT scanmeasured at isocenter. This profile is the distribution ofradiation dose along the axis of the patient (the z axis)and is known as D(z).

1544 November-December 2002 RG f Volume 22 ● Number 6

Radio

Gra

phic

s

very convenient. Measurements of exposurecould be obtained with a pencil ionization cham-ber, but its fixed length of 100 mm meant thatonly 14 sections of 7-mm thickness could be mea-sured with that chamber alone. To measureCTDI for thinner nominal sections, sometimeslead sleeves were used to cover the part of thechamber that exceeded 14 section widths.

To overcome the limitations of CTDI with 14sections, another radiation dose index—CTDI100—was developed. This index relaxed the constrainton 14 sections and allowed calculation of the in-dex for 100 mm along the length of an entire pen-cil ionization chamber (14), regardless of thenominal section width being used. This index istherefore defined as follows (9):

CTDI100 � �1/NT� ��5 cm

5 cm

Dsingle�z� dz, (4)

where N is the number of acquired sections perscan (also referred to as the number of data chan-nels used during acquisition) and T is the nominalwidth of each acquired section (which is not nec-essarily the same as the nominal width of the re-constructed section width).

(Note: The product of N � T is meant to reflectthe total nominal width of the x-ray beam duringacquisition. Therefore, if a multidetector scanneris used with four channels [rows] of 1.25 mmeach for scan acquisition, then, regardless of thereconstructed section width, NT � 4 � 1.25 mm �5 mm; similarly, if the same scanner is used withfour channels of 5-mm width for scan acquisition,then, regardless of the reconstructed sectionwidth, NT � 20 mm.)

Because the ionization chamber measures anintegrated exposure along its 100-mm length, thisis equivalent to the following formula:

CTDI100 � � f � C � E � L�/�NT�, (5)

where f is the conversion factor from exposure toa dose in air (use 0.87 rad/R), C is the calibrationfactor for the electrometer, E is the measuredvalue of exposure in roentgens acquired from asingle 360° rotation with a beam profile of NT (asdefined earlier), L is the active length of the pencilionization chamber, and N and T are as definedfor Equation (4).

Thus, the exposure measurement, performedwith one axial scan either in air or in one of thepolymethyl methacrylate phantoms for whichCTDI is defined, results in a calculated dose in-dex, CTDI100. This index can be measured andcalculated for the center location as well as atleast one of the peripheral positions (1 cm below

the surface) within the phantom to describe thevariations within the scan plane as well.

CTDIw was created to represent a dose indexthat provides a weighted average of the center andperipheral contributions to dose within the scanplane (9). This index is used to overcome thelimitations of CTDI100 and its dependency onposition within the scan plane. The definition isas follows:

CTDIw � �1/3��CTDI100�center

� �2/3��CTDI100�periphery.(6)

One final CTDI descriptor takes into accountthe parameters that are related to a specific imag-ing protocol, the helical pitch or axial scan spac-ing, and is defined as CTDIvol:

CTDIvol � CTDIw � NT/I, (7)

where N and T are as defined earlier and repre-sent the total collimated width of the x-ray beamand I is the table travel per rotation for a helicalscan or the spacing between acquisitions for axialscans.

For helical scans, the following formulas areused:

NT/I � 1/pitch (8)

and

CTDIvol � CTDIw/pitch, (9)

where pitch is defined as table distance traveled inone 360° rotation/total collimated width of thex-ray beam.

Another dose descriptor that is related toCTDI and is commonly reported on CT scannersand in the literature is the dose-length product(DLP) (9). This value is simply the CTDIvol mul-tiplied by the length of the scan (in centimeters)and is given in units of milligray-centimeters:

DLP � CTDIvol � scan length. (10)

This descriptor is used in one approach to obtainan estimate of effective dose that will be describedlater.

These CTDI descriptors are obviously meantto serve as an index of radiation dose due to CTscanning and are not meant to serve as an accu-rate estimate of the radiation dose incurred byan individual patient. Although the phantom

RG f Volume 22 ● Number 6 McNitt-Gray 1545

Radio

Gra

phic

s

measurements are meant to be reflective of anattenuation environment somewhat similar to apatient, the homogeneous polymethyl methacry-late phantom does not simulate the different tis-sue types and heterogeneities of a real patient. Inaddition, the CTDI100 calculation uses the f fac-tor (from Eq [5]) only to convert from exposureto a dose in air; other tissues have different f fac-tors. The f factor (in units of rads per roentgen) isdetermined by the ratio of the mass energy ab-sorption coefficient of a tissue to that of air:

f � 0.87 � ��t/t�/�a/a��, (11)

where t/t is the mass energy absorption coeffi-cient of the tissue (eg, bone, lung, soft tissue) anda/a is the mass energy absorption coefficient ofair. The mass energy absorption coefficient de-pends not only on the tissue but also on the en-ergy of the photons, especially in the energy rangeused by CT. Thus, the CTDI100 calculation pre-sents a very simplified condition for measuringradiation dose.

Factors That InfluenceRadiation Dose from CT

In general, there are some factors that have a di-rect influence on radiation dose, such as the x-raybeam energy (kilovolt peak), tube current (in mil-liamperes), rotation or exposure time, sectionthickness, object thickness or attenuation, pitchand/or spacing, dose reduction techniques such astube current variation or modulation, and dis-tance from the x-ray tube to isocenter. In addi-tion, there are some factors that have an indirecteffect on radiation dose—those factors that have adirect influence on image quality, but no directeffect on radiation dose; for example, the recon-struction filter. Choices of these parameters mayinfluence an operator to change settings that dodirectly influence radiation dose. These factorsare discussed in this section.

Beam EnergyThe energy of the x-ray beam has a direct influ-ence on patient radiation dose. This is selected bythe operator (technologist) when the kilovolt peakis chosen for the scan. However, it is also influ-enced by the filtration selected for the scan. Onsome scanners, the selection of filtration is ex-plicit; for others, it is implied (eg, by selection ofthe scan field of view [SFOV]). The influence ofbeam energy is shown in Table 1. When all othertechnical parameters are held constant and the

kilovolt peak is increased on a single-detector CTscanner, the CTDIw values also increase for boththe head and body CTDI phantoms. For ex-ample, when the kilovolt peak was increased from120 to 140 on a CT/i scanner (GE Medical Sys-tems, Milwaukee, Wis), the CTDIw increase was37.5% for the head phantom and 39% for thebody phantom.

Photon FluenceThe photon fluence, as influenced by the tubecurrent–time product (milliampere-seconds), alsohas a direct influence on patient radiation dose.As one might expect, the radiation dose is directlyproportional to the milliampere-seconds value.This is shown in Table 2, which gives the resultswhen the milliampere-seconds value is increasedand all other technical parameters are held con-stant on a single-detector CT scanner. Underthese conditions, the CTDIw values increase lin-early with milliampere-seconds for both the headand body CTDI phantoms.

Table 1Changes in CTDIw in Head and BodyPhantoms as a Function of Kilovolt Peak

BeamEnergy(kVp)

CTDIw

in HeadPhantom(mGy)

CTDIw

in BodyPhantom(mGy)

80 14 5.8100 26 11120 40 18140 55 25

Note.—All other factors were held constant at 300mA, 1 sec, and 10 mm. Results are from a single-detector CT scanner.

Table 2Changes in CTDIw in Head and BodyPhantoms as a Function of Milliampere-Seconds Setting

Tube Current–Time Product(mAs)

CTDIw

in HeadPhantom(mGy)

CTDIw

in BodyPhantom(mGy)

100 13 5.7200 26 12300 40 18400 53 23

Note.—All other factors were held constant at 120kVp and 10 mm. Results are from a single-detectorCT scanner.

1546 November-December 2002 RG f Volume 22 ● Number 6

Radio

Gra

phic

s

Note that these results hold only while the tubecurrent–time product is varied and all other pa-rameters are held constant. This is an issue be-cause on some scanners, the user inputs a param-eter labeled “mAs,” but that parameter is reallythe effective milliampere-seconds value, which ismilliamperage � time/pitch. On these scanners,when pitch is varied, the milliampere-secondsvalue is varied in a corresponding fashion to keepthe effective milliampere-seconds value constant(17). The effects of pitch are discussed in the nextsection.

Helical PitchFor helical scans, the pitch parameter (defined astable distance traveled in one 360° rotation/totalcollimated width of the x-ray beam) has a directinfluence on patient radiation dose. This is essen-tially because as pitch increases, the time that anyone point in space spends in the x-ray beam isdecreased. The relationship between radiationdose and pitch has been shown previously by us-

ing phantoms and thermoluminescent dosimeters(18). On the basis of these results, the CTDIvol

(which is the only CTDI descriptor that takespitch into account) varies as shown in Table 3,which gives the results when the pitch is variedand all other technical parameters are held con-stant on a single-detector CT scanner.

X-ray Beam Collimation:Single-Detector ScannersThe collimation of the x-ray beam will both di-rectly and indirectly influence the patient radia-tion dose. The indirect effects will be describedlater. For a single section with all other technicalparameters held constant, more x-ray photonswill be transmitted when the collimator setting iswider (wider x-ray beam for a thicker section).However, exposure and absorbed radiation doseare defined on a per unit mass basis. The thickersection has more photons available but also moremass being irradiated than a thinner section, thusindicating that the radiation dose for thick andthin sections may be close to equivalent (the dif-ference might be attributed to the higher scatterexpected in the thicker section). This equivalencewould also assume that the radiation profiles(and, as shown earlier, the overlap between adja-cent exposures) are equivalent between narrowand wide collimation settings. However, previouspublications have shown that this is not quite truefor single-detector scanners and that thinner colli-mations typically result in a greater degree ofoverlap and higher CTDI values (18). The resultsfrom measuring CTDIw are shown in Table 4,which gives the results when the collimation isvaried and all other technical parameters are heldconstant on a single-detector CT scanner.

X-ray Beam Collimation:Multiple-Detector ScannersAlthough the effects of beam collimation weresmall for a single-detector scanner, current expe-rience shows that this is not the case with multi-detector scanners. In fact, early reports from earlyversions of multidetector scanners showed signifi-cant dependence on x-ray beam collimation (19).These effects result from differences in x-raybeam collimation—even when the same recon-structed section thickness is used. That is, onmany multidetector scanners, there are severalways to scan and reconstruct images that have thesame section thickness. For example, on one mul-tidetector CT scanner (LightSpeed Qx/i; GEMedical Systems), one can perform axial scans of

Table 3Changes in CTDIvol in Head and BodyPhantoms as a Function of Pitch

Pitch

CTDIvol

in HeadPhantom(mGy)

CTDIvol

in BodyPhantom(mGy)

0.5 80 360.75 53 241.0 40 181.5 27 122.0 20 9

Note.—All other factors were held constant at 120kVp, 300 mA, 1 sec, and 10 mm. Results are from asingle-detector CT scanner.

Table 4Changes in CTDIw in Head and BodyPhantoms as a Function of Collimationfor a Single-Detector Scanner

Collimation(mm)

CTDIw

in HeadPhantom(mGy)

CTDIw

in BodyPhantom(mGy)

1 45 193 41 185 40 187 40 18

10 40 18

Note.—All other factors were held constant at 120kVp and 300 mAs.

RG f Volume 22 ● Number 6 McNitt-Gray 1547

Radio

Gra

phic

s

4 � 1.25 mm (5-mm beam width), 4 � 2.5 mm(10-mm beam width), and 4 � 5 mm (20-mmbeam width) to make a 5-mm-thick reconstructedsection. For each of these modes, when CTDIw

values are measured, there is a surprising differ-ence in absorbed dose. These results are shown inTable 5, which gives the results when all othertechnical parameters are held constant on thatmultidetector CT scanner.

These results show that the difference in beamcollimation, not the reconstructed section width,makes a significant difference in CTDIw. Thesedifferences may be as much as 55% in the headphantom and 65% in the body phantom, with thehigher doses coming when narrower beam colli-mation is used.

Effects of Object (and Patient) SizeIn each of the preceding sections, we reportedresults for both the head and body phantoms.These phantoms, as described earlier, are made ofthe same soft-tissue–equivalent material but are16-cm-diameter and 32-cm-diameter right circu-lar cylinders, respectively. To produce each table,we used the same technical factors for each phan-tom. Therefore, the primary difference in resultsbetween the head and body phantoms is size.Each of the tables shows that when the same tech-nical parameters are used, the appropriate indexshows that the smaller object always absorbs thehigher dose and that the difference is at least afactor of two. Thus, for the same exposure fac-tors, smaller patients would be expected to absorbmuch higher amounts of radiation dose thanlarger patients. This has significant implicationsfor pediatric patients and small adults.

This is primarily because tissues are being ex-posed with both entrance radiation (as the tube ispositioned directly over the tissue) and exit radia-tion (as the tube moves to the other side of thepatient) as the source moves around the patient.For smaller patients, the exit radiation has beenattenuated by less tissue and therefore is closer tothe entrance radiation in its intensity, resulting ina much more uniform dose distribution (nearlyequal at all locations in a 16-cm-diameter phan-tom). For the larger patient, the exit radiation ismuch less intense due to its attenuation throughmore tissue. This results in a difference within thescan plane with the higher radiation dose valuesoccurring near the periphery, where entrance ex-posure is highest.

The effect of patient or object size on radiationdose has brought significant discussion into theproper selection of protocols for imaging pediatric

patients as well as adjusting technical factors forpatients according to size (20–28).

Other Options for Reducing Scan DoseIn addition to the technical parameters discussedearlier, manufacturers have recently provided us-ers with other means to reduce patient dose. Oneof these is an option to make changes in tube cur-rent based on the estimated attenuation of thepatient at a specific location. Thus, the tube cur-rent will be programmed to a maximum value andcan be reduced when there is information that alocation along the patient is expected to be lessattenuating than the most attenuating location tobe imaged. This is determined by using both an-teroposterior and lateral planning projectionviews. From these views, the tube current will beprogrammed to vary by location along the lengthof the patient and even as the tube is rotatingaround the patient. The exact details of the op-tion vary by manufacturer.

In the near future, manufacturers may providereal-time (or close to real-time) tube currentmodulation (29–31), so that tube current can bevaried (reduced) as the scan is actually occurring,eliminating the need for both planning projec-tions for dose reduction purposes (they may stillbe needed for planning purposes).

Indirect EffectsIn addition to the direct effects that collimationhas, as described earlier, there are some indirecteffects that both it and the reconstruction algo-rithm may have on radiation dose. This is be-cause, when thinner reconstructed image thick-nesses are used, with all other factors held con-stant, there will be more noise in the image(where noise is defined as the standard deviationof the CT number). Therefore, noise typically

Table 5Changes in CTDIw in Head and BodyPhantoms as a Function of Collimationfor a Multidetector Scanner

Collimation(mm)

TotalBeamWidth(mm)

CTDIw

in HeadPhantom(mGy)

CTDIw

in BodyPhantom(mGy)

4 � 1.25 5 62 332 � 2.5 5 62 331 � 5 5 62 334 � 2.5 10 46 242 � 5 10 46 244 � 5 20 40 20

Note.—All other factors were held constant at 120kVp and 300 mAs.

1548 November-December 2002 RG f Volume 22 ● Number 6

Radio

Gra

phic

s

increases with 1/√T, where T is the nominal sec-tion thickness. Therefore, a 10-mm-thick sectionis expected to have 3.2 times less noise than a1-mm-thick section. Often when noisy images areobtained, the kilovolt peak or milliampere-sec-onds value or both are increased to offset the in-crease in noise due to narrower sections. Similarbehavior is observed for the effects of the recon-struction algorithm. Algorithms that enhancehigher spatial frequencies and improve spatialresolution (such as required for lung or skeletalimaging) also increase the noise in the image. Toovercome this increase in noise, the kilovolt peakor milliampere-seconds value or both may be in-creased. This increase in kilovolt peak or milliam-pere-seconds value will result in an increase inradiation dose. Therefore, although changing thealgorithm or section thickness may not have adirect effect on radiation dose, the selection oftechnical factors to offset the resulting increase inimage noise may result in an increase in radiationdose.

Methods to Reduce Patient DoseFrom the preceding discussion, it appears thatthere are several mechanisms to reduce the radia-tion dose to a patient (32,33). However, each ofthem has some resulting trade-off involved. Eachof these is discussed below (27).

Reducing the Milliampere-Seconds Value.—From the results presented earlier, the radiationdose is linear with the milliampere-seconds valuewhen all other factors are held constant. So, if themilliampere-seconds value is reduced by 50%,the radiation dose will be reduced by the sameamount. However, this reduction will increaseimage noise by 1/√(mAs), which means that a50% reduction in the milliampere-seconds valueresults in a noise increase of 41% (1/√ 2 � 1.41, a41% increase). Depending on the requirements ofthe clinical application, this reduction may bereadily accepted; in other cases, this type of re-duction in milliampere-seconds may compromisethe diagnostic quality of the imaging examination.For example, detection of high-contrast objects inthe lung may not require a low-noise imaging pro-tocol and the reduction in milliampere-secondsmay be well tolerated. On the other hand, imag-ing low-contrast lesions in the liver does require alow-noise imaging protocol and the reduction inmilliampere-seconds may limit the ability to de-tect these lesions.

Increasing Pitch.—The radiation dose is in-versely proportional to pitch when all other fac-tors are held constant. Therefore, increasing pitchis one consistent way to reduce radiation dose.The trade-off in increasing pitch is an increase in

effective section thickness, which results in in-creased volume averaging, which in turn may re-duce the image signal (contrast between someobject and background). The ability to use thistype of dose reduction again depends on the clini-cal application.

Varying the Milliampere-Seconds Value byPatient Size.—CT is an example of a digital mo-dality in which the image quality continues toimprove as the exposure increases. This is con-trasted with analog projectional film, in which toohigh of an exposure results in an overexposed (toodark) film. Thus, when pediatric patients or smalladult patients are imaged with CT using full-sizedadult techniques, there is no penalty to imagequality; in fact, the image quality is better underthese conditions, as more photons reach the de-tector and image noise is reduced. However, theradiation dose to the smaller patient is potentiallyhigher than is necessary to obtain a diagnosticimage. Therefore, significant effort has recentlybeen put into developing size- and weight-basedimaging protocols to reduce radiation dose to pe-diatric patients and small adult patients, so thatradiation dose can be reduced while still achievingsufficient diagnostic image quality (21–25,27).This has typically been in the form of a reducedmilliampere-seconds value for reduced patientsize and has led to the development of suggestedtechnique charts for pediatric patients.

Reducing Beam Energy.—As discussed earlier,reducing the beam energy results in reduced ra-diation dose when all other factors are held con-stant. This will increase the image noise, and con-trast changes will occur with a change in kilovoltpeak, increasing with lower kilovolt peak for mosttissue interfaces but decreasing or changing verylittle for others. From CTDI100 results, and notfrom the CTDIw results, we can observe that theradiation dose gradient is larger from periphery tocenter in the body phantom at lower kilovolt peaksettings. This implies a greater relative skin dosefor patients when lower kilovolt peak settings areused.

EstimatingEffective Dose from CT

The definition of effective dose was given earlier(Eq [1]) (4) as the weighted sum of organ dosesresulting from the examination, where the radio-sensitive organs were defined along with theirtissue-weighting factors. Although it appearsstraightforward to estimate effective dose, it is

RG f Volume 22 ● Number 6 McNitt-Gray 1549

Radio

Gra

phic

s

actually difficult to accurately estimate the dose toan individual organ from a CT scan. This is evenmore difficult when attempting to estimate theeffective dose for each patient when each one hasunique characteristics of height, weight, age, gen-der, and composition. Still, several differentmethods for estimating effective dose—primarilyto standard patient models—have been devel-oped.

The first is based on Monte Carlo simulationsperformed several years ago (34–37). This work,performed by members of the United Kingdom’sNational Radiological Protection Board (NRPB),used Monte Carlo methods to simulate CT scan-ning around a previously developed mathematicalpatient model (MIRD V [38]). This hermaphro-dite patient model uses geometric shapes (cylin-ders, spheres, cones) to approximate the shapeand location of all radiosensitive organs in thebody; other approximations were used for thebrain and lens of the eye. Physical measurementsof CTDI in air on many scanners were collectedas part of a national dosimetry program (36);these data were used to convert from the normal-ized Monte Carlo data to absolute dose values.The simulation also involved modeling scannergeometry, spectrum, section thickness, and otherproperties with the patient model in the beam—accounting for primary and scatter radiation.From these simulations, radiation doses for eachorgan in each imaged section were calculated andcould be tallied to estimate the organ dose foreach organ. These organ doses could then becombined with appropriate weighting factors toestimate effective dose (or effective dose equiva-lent from ICRP 26). This work was completedbefore helical and multidetector scanners werecommonplace and did not explicitly model helicalscanning or the effects of multidetector scanners.

This work formed the basis for several softwareprograms that have taken the results of thesesimulations and put an interface on them to allowusers to input some technical parameters and cal-culate an effective dose. These include programssuch as CTDOSE (from John Le Heron at NewZealand’s National Radiation Laboratory[johnleh@nrl.moh.govt.nz]) and the Microsoft(Redmond, Wash) Excel–based product from theImaging Performance Assessment of CT (Im-PACT) scanners group in the United Kingdom(39). The latter software package matches thedesired scanner to one of those used in the origi-nal study on the basis of CTDI in air as well asCTDI in phantom measurements.

In a related but independent effort, Zankl et al(40) from the Gesellschaft fur Strahlen- und Um-

weltforschung (GSF) performed simulations ontwo different mathematical sex-specific phan-toms, “Adam” and “Eva” (41), which were basedon the ICRP data on Reference Man. Similar tothe MIRD V phantom, the GSF mathematicalphantoms were described by using combinationsof spheres, ellipsoids, cones, and parts and com-binations of these shapes. The Monte Carlosimulations were performed modeling photontransport through the simulated patient by usingthree different energy spectra. The section thick-ness was 1 cm, and radiation doses were calcu-lated for single sections at positions varying con-tiguously from the bottom of the trunk to the topof the head. The result was mean organ conver-sion factors for each organ and per single CT sec-tion of 1 cm width centered at a given location inthe phantom. The mean dose to a single organresulting from a particular CT scan was estimatedby summing up the contribution to the organdose from each relevant section. The relative or-gan doses were converted to absolute organ dosesby using the air kerma free-in-air on the axis ofrotation. As in the National Radiological Protec-tion Board data, these organ doses can be com-bined with appropriate weighting factors to esti-mate effective dose (or effective dose equivalentfrom ICRP 26).

The GSF work formed the basis for a softwareprogram with a graphical user interface (42). Thispackage (WinDose; Scanditronix/Wellhofer,Schwarzenbruck, Germany, www.wellhofer.com)has taken the results of the simulations and put aninterface on them to allow users to input sometechnical parameters and calculate an effectivedose. The output is an estimate of the effectivedose given the scanner characteristics, body re-gion, scan range, and other parameters. This pro-gram also outputs the effective dose in millisie-verts as well as in units of natural backgroundradiation per year.

Another approach was that of estimating theenergy imparted, developed by Atherton andHuda (43–47). Energy imparted is the measureof the total ionizing energy deposited in the pa-tient during the CT examination and here is sym-bolized by �. On the basis of the simulation dataof Jones and Shrimpton (36,37), the energy im-parted was calculated for a mathematical anthro-pomorphic phantom (38). For different CT sys-tems, the ratio of effective dose equivalent (HE,defined in ICRP 26 [6]) to energy imparted(HE/�) and the ratio of effective dose (E, definedin ICRP 60 [4]) to energy imparted (E/�) werecalculated and plotted as a function of the posi-tion along the anthropomorphic phantom. Highpeaks in this function showed the position of or-gans with a higher radiosensitivity like the gonads,breast, or thyroid. Therefore, in this approach,

1550 November-December 2002 RG f Volume 22 ● Number 6

Radio

Gra

phic

s

effective dose is estimated by first estimating theenergy imparted to the body region being scannedand then multiplying by the ratio factor for thatparticular region.

One other method of note to estimate the ef-fective dose involves conversion factors for a gen-eral anatomic region as described by the Euro-pean Guidelines on Quality Criteria for Com-puted Tomography (9), which are based on thework of Jessen et al (48). In this approach, theCTDIvol and distance are used to estimate theDLP, which is then multiplied by a region-spe-cific conversion factor to estimate the effectivedose. These conversion factors range from 0.0023mSv/mGy � cm for the head region to 0.017 mSv/mGy � cm for the chest region and 0.019 mSv/mGy � cm for the pelvis. This approach obviouslydoes not take into account any patient-specific oreven examination-specific factors but provides aneasily estimated value of effective dose.

For a typical chest scan performed with a single-detector scanner at 120 kVp, 250 mAs, 5–7-mmcollimation, and a pitch of 1, the CTDI100, center

would be 10 mGy, the CTDI100, periphery would be18 mGy, the CTDIw would be 15 mGy, and theCTDIvol would be 15 mGy. If a length of scan of25 cm is assumed, the DLP would be 375 mGy �cm; when the conversion factor for the chest isused, one estimate of effective dose would be 6.4mSv.

For a typical head scan performed with a single-detector scanner at 120 kVp, 300 mAs, 5-mmcollimation, and a pitch of 1, the CTDI100, center

would be 40 mGy, the CTDI100, periphery would be40 mGy, the CTDIw would be 40 mGy, and theCTDIvol would be 40 mGy. If a length of scan of17.5 cm is assumed, the DLP would be 700mGy � cm; when the conversion factor for thehead is used, one estimate of effective dose wouldbe 1.6 mSv. Thus, the effective dose for the headscan is considerably less than that for the chestscan, even though the CTDI values for the headscan are much higher. This is because fewer ofthe radiosensitive organs are irradiated.

Relevant Activities and ResourcesThere have been recent activities regarding radia-tion dose in many different agencies and organi-zations. These include an ongoing focus on radia-tion dose by the Society for Pediatric Radiology,where the dose from CT to pediatric patients isan issue of significant concern. These concernswere presented in several articles including refer-ences 20–28, which reflected the concern thatCT techniques being used for pediatric patientswere similar to those used for adults and were notbeing adapted to the small size of pediatric pa-tients. These articles also reported that pediatric

patients may have increased sensitivity to ionizingradiation. Since that time, the Society for Pediat-ric Radiology has held sessions at its annual con-ferences dedicated to CT and radiation dose re-duction techniques.

The American Association of Physicists inMedicine (AAPM) has convened a task group(Task Group 23: CT Dosimetry) under its Diag-nostic X-ray Imaging Committee to address is-sues related to CT radiation dose. This taskgroup is addressing issues relating to standardiz-ing descriptors, measurement methods, and edu-cation activities related to these. This is in addi-tion to the task group on reference values createdunder the AAPM’s Radiation Protection Com-mittee (Task Group 7: Reference Values for Di-agnostic X-ray Examinations), which reports val-ues for head CT scans. Both of these task groupsoperate under the AAPM’s Science Council.

In addition, the CT accreditation program ofthe American College of Radiology (ACR) re-quires sites applying for accreditation to measureand calculate CTDI100, CTDIw, CTDIvol, DLP,and effective dose for pediatric body, adult head,and adult body techniques. Although this will notbe a pass-fail criterion at first, the ACR’s programwill eventually use radiation dose as part of itspass-fail criteria. This program does provide ac-ceptable value limits for each protocol.

The Center for Devices and RadiologicalHealth (CDRH) of the Food and Drug Adminis-tration has made some public recommendationson CT scanning for pediatric and small adult pa-tients, which they have published on their Website (27). In addition, the agency has completedits Nationwide Evaluation of X-ray Trends(NEXT) for CT, which was undertaken in 2000(1). Preliminary results were presented at the2001 conference of the Radiological Society ofNorth America. In this survey, CTDI values werereported from many different scanners placed atdifferent institutions across the United States.From this survey, mean values as well as ranges ofvalues are being obtained for different types ofscans. Publications on the results of this surveyare being prepared.

ConclusionsThis article describes the basic concepts of radia-tion dose in CT. The basic concepts of exposure,absorbed dose, and effective dose were describedin general and then described in the context ofCT-specific descriptors. The various CTDI de-scriptors were defined, and the relationships be-tween them were described as well. Once these

RG f Volume 22 ● Number 6 McNitt-Gray 1551

Radio

Gra

phic

s

CTDI parameters were defined, factors that influ-ence radiation dose, both directly and indirectly,were described and the magnitudes of those ef-fects were reported. The effective dose was de-scribed, and several methods for estimating thisimportant parameter were presented. Some rel-evant activities were identified.

Although there are many dose descriptors andmethods for measuring radiation dose and effec-tive dose from CT, there are still areas for furtherresearch. Recent investigations include examiningthe effects of patient size on radiation dose (withspecific implications for pediatric patients) andmethods for estimating patient-specific radiationdose (49,50), rather than estimating based onphantoms or standard mathematical patients suchas the MIRD V. In the near future, there may bemethods to estimate the radiation dose, organdose, and effective dose from arbitrary CT scan-ning protocols by using patient-specific models.

References1. Stern SH, Kaczmarek RV, Spelic DC, Suleiman

OH. Nationwide Evaluation of X-ray Trends(NEXT) 2000–2001 survey of patient radiationexposure from computed tomographic (CT) ex-aminations in the United States (abstr). Radiology2001; 221(P):161.

2. Center for Devices and Radiological Health, Foodand Drug Administration. Whole body scanningusing computed tomography (CT). Available at:http://www.fda.gov/cdrh/ct/. Accessed April 2002.

3. Bushberg JT, Seibert JA, Leidholdt EM, BooneJM. The essential physics of medical imaging. 2nded. Philadelphia, Pa: Lippincott Williams &Wilkins, 2001.

4. International Council on Radiation Protection.1990 recommendations of the International Com-mission on Radiological Protection. Publication60, Annals of the ICRP 1991; 21. Oxford, En-gland: Pergamon, 1991.

5. McCollough CM, Schueler BA. Calculation ofeffective dose. Med Phys 2000; 27:838–844.

6. International Council on Radiation Protection.Recommendations of the International Commis-sion on Radiological Protection, 1977. Publication26, Annals of the ICRP 1(3) (reprinted with addi-tions in 1987). Oxford, England: Pergamon, 1977.[superseded by ICRP Publication 60]

7. American Association of Physicists in Medicine.Standardized methods for measuring diagnosticx-ray exposure. Report no. 31, 1990. Available at:www.aapm.org/pubs/reports.

8. American Association of Physicists in Medicine.Specification and acceptance testing of computedtomography scanners. Report no. 39, 1993. Avail-able at: www.aapm.org/pubs/reports.

9. European Guidelines on Quality Criteria for Com-puted Tomography (EUR 16262 EN, May 1999).Available at: www.drs.dk/guidelines/ct/quality/index.htm. Accessed July 2002.

10. Shrimpton PC, Edyvean S. CT scanner dosim-etry. Br J Radiol 1998; 71:1–3.

11. Rothenberg LN, Pentlow KS. CT dosimetry andradiation safety. In: Goldman LW, Fowlkes JB,eds. Syllabus: a categorical course in diagnosticradiology physics—CT and US cross-sectionalimaging. Oak Brook, Ill: Radiological Society ofNorth America, 2000; 171–188.

12. McCollough CM, Zink FE. Performance evalua-tion of CT systems. In: Goldman LW, FowlkesJB, eds. Syllabus: a categorical course in diagnosticradiology physics—CT and US cross-sectionalimaging. Oak Brook, Ill: Radiological Society ofNorth America, 2000; 189–207.

13. Nagel HD, ed. Radiation exposure in computedtomography. 2nd ed. Hamburg, Germany: CO-CIR, 2000. [available through cocir@zvei.org]

14. Jucius RA, Kambic GX. Radiation dosimetry incomputed tomography. Appl Opt Instrum EngMed 1977; 127:286–295.

15. Shope TB, Gagne RM, Johnson GC. A methodfor describing the doses delivered by transmissionx-ray computed tomography. Med Phys 1991;8:488–495.

16. Department of Health and Human Services, Foodand Drug Administration. 21 CFR Part 1020: Di-agnostic x-ray systems and their major compo-nents; amendments to performance standard; Fi-nal rule. Federal Register 1984, 49, 171.

17. Mahesh M, Scatarige JC, Cooper J, Fishman EK.Dose and pitch relationship for a particular multi-slice CT scanner. AJR Am J Roentgenol 2001;177:1273–1275.

18. McNitt-Gray MF, Cagnon C, Solberg TD, ChettyI. Radiation dose from spiral CT: the relative ef-fects of collimation and pitch. Med Phys 1999;26:409–414.

19. McCollough CH, Zink FE. Performance evalua-tion of a multi-slice CT system. Med Phys 1999;26:2223–2230.

20. Brenner DJ, Elliston CD, Hall EJ, Berdon WE.Estimated risks of radiation-induced fatal cancerfrom pediatric CT. AJR Am J Roentgenol 2001;176:289–296.

21. Paterson A, Frush DP, Donnelly LF. Helical CTof the body: are settings adjusted for pediatric pa-tients? AJR Am J Roentgenol 2001; 176:297–301.

22. Donnelly LF, Emery KH, Brody AS, et al. Per-spective: minimizing radiation dose for pediatricbody applications of single-detector helical CT—strategies at a large children’s hospital. AJR Am JRoentgenol 2001; 176:304–306.

23. Robinson AE, Hill EP, Harpen MD. Radiationdose reduction in pediatric CT. Pediatr Radiol1986; 16:53–54.

24. Kamel IR, Hernandez RJ, Martin JE, SchlesingerAE, Niklason LT, Guire KE. Radiation dose re-duction in CT of the pediatric pelvis. Radiology1994; 190:683–687.

25. Rogalla P, Stover B, Scheer I, Juran R, GaedickeG, Hamm B. Low-dose spiral CT: applicability topaediatric chest imaging. Pediatr Radiol 1998;29:565–569.

26. Huda W, Chamberlain CC, Rosenbaum AE, Gar-risi W. Radiation doses to infants and adults un-dergoing head CT examinations. Med Phys 2001;28:393–399.

1552 November-December 2002 RG f Volume 22 ● Number 6

Radio

Gra

phic

s

27. Feigal DW. FDA Public Health Notification: re-ducing radiation risk from computed tomographyfor pediatric and small adult patients. Center forDevices and Radiological Health, Food and DrugAdministration. Available at: http://www.fda.gov/cdrh/safety/110201-ct.html. Accessed November2001.

28. Shrimpton PC, Wall BF. Reference doses for pae-diatric computed tomography. Radiat Prot Do-simetry 2000; 90:249–252.

29. Gies M, Suess C, Wolf H, Kalender WA, MadsenMT. Dose reduction in CT by anatomicallyadapted tube current modulation. I. Simulationstudies. Med Phys 1999; 26:2235–2247.

30. Kalender WA, Wolf H, Suess C, Gies M, BautzWA. Dose reduction in CT by anatomicallyadapted tube current modulation. II. Phantommeasurements. Med Phys 1999; 26:2248–2253.

31. Kalender WA, Wolf H, Suess C, Gies M, GreessH, Bautz WA. Dose reduction in CT by on-linetube current control: principles and validation onphantoms and cadavers. Eur Radiol 1999; 9:323–328.

32. Ravenel JG, Scalzetti EM, Huda W, Garrisi W.Radiation exposure and image quality in chest CTexaminations. AJR Am J Roentgenol 2001; 177:279–284.

33. Huda W, Scalzetti EM, Levin G. Technique fac-tors and image quality as functions of patientweight at abdominal CT. Radiology 2000; 217:430–435.

34. Shrimpton PC, Jones DG, Hillier MC, Wall BF,Le Heron JC, Faulkner K. Survey of CT practicein the UK. Part 2: Dosimetric aspects. NRPBR249. Chilton, England: National RadiologicalProtection Board, 1991.

35. Jones DG, Shrimpton PC. Survey of CT practicein the UK. Part 3: Normalised organ doses calcu-lated using Monte Carlo techniques. NRPBR-250. Chilton, England: National RadiologicalProtection Board, 1992.

36. Jones DG, Shrimpton PC. Normalised organdoses for x-ray computed tomography calculatedusing Monte Carlo techniques. NRPB SR-250.Chilton, England: National Radiological Protec-tion Board, 1992. [computer software report]

37. Shrimpton PC, Jones DG. Normalised organdoses for x-ray computed tomography calculatedusing Monte Carlo techniques and a mathematicalanthropomorphic phantom. Radiat Prot Dosim-etry 1993; 49:241–243.

38. Cristy M. Mathematical phantoms representingchildren of various ages for use in estimates of in-ternal dose. NUREG/CR-1159 ORNL/NUREG/TM-367. Oak Ridge, Tenn: Oak Ridge NationalLaboratory, 1980.

39. Imaging Performance Assessment of CT (Im-PACT) CT Patient Dosimetry Calculator, version0.99m. Created 07/01/2002. Available at: http://www.impactscan.org/ctdosimetry.htm.

40. Zankl M, Panzer W, Drexler G. The calculation ofdose from external photon exposures using refer-ence human phantoms and Monte Carlo methods.Part VI: Organ doses from computed tomographicexaminations. GSF-Bericht 30/91. Neuherberg,Germany: Gesellschaft fur Strahlen- und Umwelt-forschung, 1991.

41. Kramer R, Zankl M, Williams G, Drexler G. Thecalculation of dose from external photon expo-sures using reference human phantoms and MonteCarlo methods. Part I: The male (Adam) and fe-male (Eva) adult mathematical phantoms. GSF-Bericht S-885. Munich, Germany: Gesellschaftfur Strahlen- und Umweltforschung, 1982.

42. Kalender WA, Schmidt B, Zankl M, Schmidt M.A PC program for estimating organ dose and ef-fective dose values in computed tomography. EurRadiol 1999; 9:555–562.

43. Atherton JV, Huda W. CT doses in cylindricalphantoms. Phys Med Biol 1995; 40:891–911.

44. Huda W, Atherton JV. Energy imparted in com-puted tomography. Med Phys 1995; 22:1263–1269.

45. Atherton JV, Huda W. Energy imparted and effec-tive doses in computed tomography. Med Phys1996; 23:735–741.

46. Huda W, Scalzetti EM, Roskopf M. Effectivedoses to patients undergoing thoracic computedtomography examinations. Med Phys 2000; 27:838–844.

47. Ware DE, Huda W, Mergo PJ, Litwiller AL. Ra-diation effective doses to patients undergoing ab-dominal CT examinations. Radiology 1999; 210:645–650.

48. Jessen KA, Shrimpton PC, Geleijns J, Panzer W,Tois G. Dosimetry for optimisation of patient pro-tection in computed tomography. Appl Radiat Isot1999; 50:165–172.

49. Beifuss U, McNitt-Gray MF, Cagnon CH, De-Marco JJ. Monte Carlo methods to simulate radia-tion dose from spiral computed tomography:phantom studies (abstr). Radiology 2000; 217(P):488.

50. Jarry G, DeMarco JJ, McNitt-Gray MF. MonteCarlo dose verification of a commercial CT scan-ner with applications for patient specific dosimetry(abstr). Med Phys 2002; 29:1344. [presented atthe 2002 annual conference of the American Asso-ciation of Physicists in Medicine]

RG f Volume 22 ● Number 6 McNitt-Gray 1553

Radio

Gra

phic

s