Evaluation of different physical parameters that affect the clinical image quality for gamma camera by using different

radionuclides

F. A. Salah ,M.Sc*, W. A. Khalil ,Ph.D**, G. Ziada ,Ph.D*, M. A. Hejazy, Ph.D*. * Kasr El-Aini Center of Radiation Oncology & Nuclear Medicine-Cairo Univ. **Biophysics Dept.-Faculty of Science Cairo Univ. Main author email : Faten A.Salahm_hejazy2002@yahoo.com

Abstract Some scintillation camera manufactures adhere to standard code of performance specification established by National Electric Manufactures Association (NEMA). Items such as differential and integral uniformity, spatial resolution energy resolution, etc. are all calculated with reproducible methodology that allows the user reliable technique for creation of these standards to avoid any lack of clinical service that may violate the ethics of patient care. Because Tc-99m is the most frequently used radionuclide in nuclear medicine, many clinics perform the daily uniformity and weekly resolution checks using this radionuclide. But when other commonly used radionuclide such as Tl-201,Ga-67and I-131 are used, no standardized quality control is performed. So in these study we perform to evaluate the response of ADAC(digital) gamma camera and SELO(analogue) gamma camera to four radionuclide (Tl-201,Ga-67, I-131, and Tc99m) flood image acquired using different non-uniformity correction tables. In the planer study uniformity and resolution images were obtained using ADAC and SELO cameras, linearity was obtained only by ADAC camera, while in the SPECT study uniformity and contrast images were obtained using ADAC camera only. The response for using different non-uniformity correction tables acquired using different isotopes was different from gamma camera model to another. We can conclude that the most of the gamma camera quality control parameters (uniformity, resolution and contrast) are influenced by variation in the correction tables, while other parameters not affected by this variation like linearity. Key words : , gamma camera , clinical image quality, radionuclides.

Introduction The Anger Camera Performance characteristics and quality control: The Anger camera is not capable of producing "perfect" images of a radionuclides distribution. There are certain inherent imperfections that arise from the performance characteristics of the detector and its associated electronic circuitry and of the collimator. Image artifacts also can be caused by malfunctions of various camera components (Sorenson, et al. 1987). Methods and phantoms for scintillation camera performance testing and quality assurance have proliferated during quarter century of the camera's history. Measurement techniques have changed in response to improved imaging characteristics, reductions in measuring times, and the clinical availability of computer-assisted analysis. One comprehensive set of scintillation camera, performance measurements and routine quality assurance measurements, developed by the National Electrical Manufacturers Association (NEMA) in conjunction with users. Image Non Uniformity The non-uniformity in detector response, arises from several factors: (a) variations in PM tube response, (b) nonlinearity in X, positioning of pulses along the field of view, and (c) edge packing, factors (a) and (b) are the leading causes of deterioration in uniformity and special attention is needed to remedy them, or there can be general degradation of uniformity across the field of view (FOV) due to in appropriate spatial linearity or energy corrections. (IAEA- TECDOC-602). There are variations in the light production from -ray interaction in the detector, light transmission to PM tubes, and in the light detection and gains of the PM tubes. These variations in PM tube response contribute significantly to the non-uniformity of the detector response. This part of non-uniformity is corrected by correction factors that are obtained by a calibration method. Essentially, a collimated radioactive source is moved across the detector face with a constant PHA window setting, and variations in the photopeak values are determined as a function of their spatial positions. The energy correction factors are supplied by the manufacturer in the form of look-up table installed in me device and used to adjust the peak values for each location during imaging (Gopal B., 2001). If the camera is used with an energy window other than the one at which the PMT gain adjustments were made, the portions of the photopeak that are included in different areas of the detector are changed and striking non-uniformities may appease. Regular analyses of uniformity by a computer can facilitate detection of radial deterioration prior to any visible change. Uniformity can be different for different radionuclides and window settings It is thus important to ensure that uniformity is consistent for all radionuclides used on the gamma camera. Further, if non-standard or different window settings are introduced (e.g. narrow window, asymmetric window) heir effect on uniformity should be assessed before clinical studies are performed. A non-uniform area may be seen as either a hot or cold spot. These areas may be have diffuse distribution pattern through out the field of view, or they may be perceived as

discrete areas approximating the size of the photomultiplier tube serving that area of the crystal. These non-uniformities could simulate lesions in patients under examination. Thus cold areas could be read as tumors in the liver or infarct in the lung, while a hot area be read as a brain tumor (Muchllehuer et al., 1980).

Non-uniformity Correction Techniques

The basic premise of scintigraphic imaging is that regions of the image that exhibit a relative increase or decrease in count density have some meaningful anatomic or functional interpretation. Hence, considerable effort has been expended on the development of techniques for the correction of regional, spatial non-uniformities in Anger camera images. Historically, the causes of uniformity have been identified to originate from (1) regional variations in the energy response over the face of the camera crystal; (2) spatial distortion of counts from their true location; and (3) point source sensitivity variations. (Shabason et al. 1978, Knoll and Shrader 1982, Jaszczak and Coleman 1980)

Correction Tables
Modern gamma cameras include on-line corrections for variations of energy response ((photopeak)) and linearity across the crystal. Some cameras also include on-line uniformity corrections. These corrections are designed to provide a uniform energy response and good linearity a cross the FOV which are also prerequisites for good uniformity. As the camera slowly drifts over time, the correction tables have to be revised to again apply proper correction factors during collection of the image. In general, energy and on-line uniformity corrections require more frequent updating than linearity correction tables. Thus the operators usually collect information on energy and uniformity tables. There are several techniques have been developed to correct or compensate for image non-uniformities. One approach that is used in some camera "tuning" procedures to partially correct for image no uniformities is to adjust individual PM tube gains so that grater or lesser portions of the photopeak are included within the selected pulse-height analyzer window over different areas of the detector crystal. In effect, this adjusts detection used with an energy window other than the one at which the PM tube gain adjustments were made, the portions of the photopeak that are included in different areas of the detector are changed. Some cameras incorporate microprocessor-based computer circuitry to correct for image nonuniformities. With these cameras, a test image of a uniform radiation field, obtained prior to patient studies, is divided into a matrix of small, square elements (e.g. 64x64 elements) and stored in digital format (count/area) in the microprocessor device. The test image is used to generate a matrix of "correction factors", similar to the approach employed with Anger camera computer systems, except that the microprocessor performs the corrections in "real time" as the image data are collected, rather than by postprocessing of the image. The correction matrices are programmed into the microprocessor using calibration measurements performed at the factory before installation of the camera. The calibrations involve precise measurements using line sources of pulse-height spectra for individual PM tubes and of position distortions over

the entire area of the detector crystal. In subsequent imaging procedures the correction matrices are used on an event by event basis to compensate for regional differences in sensitivity by adjusting either the system gain or the pulse-height window, and to compensate for nonlinearities by accurately repositioning each event. Also in SPECT one of the most common and most severe reconstruction artifacts is the concentric ring or ((bull's eye)) artifacts caused by regional sensitivity variations in the projection images (Rodgers WL, et al. J Nucl Med 1982, Oppenheim BE, et al. J Nucl Med 1985). These variations are caused by camera spatial non linearitys, differences in crystal thickness or energy response, and collimator septum disparities (Halama JR, et al. 1992). The uniformity correction in (SPECT) can be evaluated by imaging a uniform phantom or a phantom with a uniform section. If the source distribution in the phantom is uniform, then the transverse slices through the phantom should also be uniform, but modified for the attenuation of the phantom itself. As mentioned above, the effects of the detector non-uniformity are circular ring artifacts.

Computer correction of gamma camera

The random fluctuations arise largely because of the limited amount of light produced in the scintillation crystal. This light is converted into photoelectrons in the array of photomultiplier tubes coupled to the crystal. The number of photoelectrons is small enough so that statistical fluctuations in their number produce event-to-event variations in the apparent position and energy of recorded gamma rays. It is these fluctuations that generally limit both the spatial and energy resolution of the gamma camera.once the statistical fluctuations are the determining factor, there is little that can be done to further reduce random errors other than to increase the light yield in the scintillation crystal or quantum efficiency of the light-to-electron conversion (G. F. Knoll, et al. 1982). These deviations are reasonably stable and can be accurately measured, their effects may be removed from gamma ray images through appropriate computer processing. This processing can be either after the fact manipulation of the recorded image or on the fly correction of each event as it is recorded by the system. The latter has greater appeal since it allows the accumulation of a corrected image in the real time.

Development of Computerized Image Correction

One digital computers became widely available for use with scintillation cameras, efforts were soon made to apply corrections for some of these is the technique of uniformity correction by multiplying the recorded image by a matrix of correction coefficients derived from a previous recorded flood image (W. L. Ashburn, et al. J Nucl Med 1970 ). The correction coefficients are inversely proportional to recorded flood intensity at each corresponding point in the image, so the process is in essence a point to point normalization of the clinical image of the reference flood source is forced to be flat regardless of any inherent variations. More recently, similar corrections have been accomplished on the fly in some gamma ray cameras through the process of skimming. Again the reference flood image is used to correct all recorded images in a manner that would render the flood image perfectly flat. Skimming techniques require that a certain

fraction of all input pulses be discarded, with that fraction being the largest for hot areas in the flood image and reducing to zero for the cold areas. The net effect is virtually the same as multiplicative uniformity correction, although the skimming process does elimiate some useful counts and thereby slightly reduces the statistical information content of the image. This type of uniform correction has come into widespread use in camera computer systems. Improvements can be realized in the appearance of gamma ray images under many circumstances. The technique also has proved to be useful for correcting projection images before carrying out the reconstruction of single photon emission tomography images (R. J. Jaszczak, et al. 1980) However, there are potential problems in the skimming approach because it treats the symptom rather than the cause of the nonuniformities. It has been shown (T. N. Pedikal, et al. J Nucl Med 1976 ) that image artifacts can be produced by the technique under circumstances in which the amount of scattering is widely different between the recorded image and the reference flood image. This would be expected to be the case if some of the nonuniformity in the flood image were due to shifts in the energy acceptance window, and only if the amount of scatter remained constant would the multiplicative correction give uniform derived images. To the extent that the reference flood nonuniformities are due to spatial distortions, skimming also upsets the constancy of the point source sensitivity of the camera. Since skimming corrections can be as large as 25% or more, considerable error can enter the quantitative measurement of source intensity at different points in the image. Correction for spatial distortions It has been recognized for some time that the proper way to correct for the spatial distortions of gamma ray cameras is to reposition events back to their proper location. The first step in this direction was published by Spector, Brookeman, Kylstra, and Diaz in 1972 (S. S. Spector, et al. J Nucl Med 1972). In this study, an orthogonal grid source was used to accurately measure the degree of spatial nonlinearity recorded in a 50x50 digital image. Programs were then developed to use these data in the post processing of subsequent images recorded from the camera in an attempt to eliminate the effects of nonlinearity. These work first points out the difficulty of mapping a discrete digital image into a similar corrected image, when the mapping is nonlinear and the pixel size in each image is similar. The desired shifts are often of the same order of magnitude as the pixel dimension, and some scheme must be incorporated to allow shifts that are on a fine scale compared with the pixel size. In order to apply true event shifting, both the original image and the corrected image must be fine grained on a scale that is small compared to the pixel size. Then each sub-pixel element can be shifted to its proper location, and if the number of subpixels is large, a slight error due to digital truncation will not have a large effect on the content of any one pixel in the corrected image. Event-shifting method was suggested by Muehllehner (G. Muehllehner. U.S. Patent, 1973). However, it fails to recognize an additional essential feature of mapping from one digital image to another, and has apparently never been implements. Here it was suggested that a matrix of correction values could be derived from a calibration image of

a grid of point sources and used to shift events. However, because of practical limits on the size of such a matrix, the correction can be stored only on a scale that is similar to the image pixel size. Thus, despite efforts to fine-grained, but also the correction itself must be implemented on a scale that is fine compared with image pixel sizes. It is also important that shifting be smoothly varying from one subpixel element to the next. Therefore the amount of the shift must be individually calculated for each subpixel and is unique for each and every case. The first successful demonstration of an online system to carry out repositioning of events to correct for spatial distortion was reported by Knoll, Bennett, and Strange at the 1978 Annual Meeting of society of Nuclear Medicine (G. F. Knoll, et al. J Nucl Med 1978). It has been incorporated by Medical Data Systems as part of the MDS image preprocessor, together with a sliding energy window to accommodate variations in the camera energy signals. MATERIALS (1) Images were acquired with the use of dual detector camera angulations with optimized 90/180 SPECT system (ADAC, Vertex Plus version 1.31). the NaI(Tl) crystal of camera has thickness 9.5mm, 8l3-inch photomultiplier tube diameter, the number of photomultiplier tube is 54,and the detector UFOV dimensions is (50 x 38). The camera computer system work under UNIX operating system (Pegasys ultra HT), uses FBP, MLEM, OSEM, and Bayesiam reconstruction methods, and contain 3D volume and surface rendering option. (2) SELO ((HR 73/D UNITOMO)) Multi WB X-ray film photo camera also used in this study. This is Single head. The number of photomultiplier tube is 61 with tube diameter 2 inch and 12 with tube diameter 1.5 inch. Four-quadrant motor drive. The camera computer is sophy G computer complete of 190 MByte hard disk 5 inch floppy disk, 13 inch color monitor, 13 inch monochrome monitor, Key board, tracker ball. (3) Phantoms used are Four-quadrant bar phantom, which used to evaluate spatial resolution, bar width ((2.12mm, 2.54mm, 3.18mm, 4.23mm)), used with ADAC gamma camera. Circular bar phantom, which also used to evaluate spatial resolution but in SELO gamma camera, bar width ((3.25mm, 3.75mm, 4.25mm, 4.75mm), NEMA resolution /linearity phantom (NP) used to measure intrinsic spatial resolution and linearity per NEMA standards. Jaszczak ((PET/SPECT)) performance phantom, for testing Photon Emission CT systems. (4) Different point sources of different activities are used of (Tc-99m, Tl-201, I-131, Ga-67) METHODS (A) Planner study A uniformity correction of the camera with different isotopes was acquired by using four point sources of TC-99m (1mCi for each), TL-201,I-131 and Ga-67 each point source was positioned at distance of 5 useful field of view diameters from the crystal (intrinsic) (2.5 meter) using a 20% energy window width for each of the radio-nuclides. Intrinsic Uniformity Four sets of intrinsic flood images were obtained from each camera using 500 Ci point

source of Tc-99m, TL-201,I-131 and Ga-67. Radionuclide photo-peak setting was adjusted carefully .A 20% energy window width for each of the radionuclides used. The point sources were positioned at distance of 5 useful field of view diameters from the crystal (UFOV) <<According to the NEMA protocol>> using (512x512) matrix with ADAC camera and 64x64 with SELO camera to a total count depth of 15 million counts. Four flood were taken with each radionuclide. First flood image was taken with the camera uniformity correction device adjusted on Tc-99m correction, second flood image was taken with the camera uniformity correction device adjusted on TL-201 correction, Third flood image was taken with the camera uniformity correction device adjusted on Ga-67 correction, Fourth image was taken with the camera uniformity correction device adjusted on I-131 correction. All flood field images of four radionuclide with different correction tables are collected to determine the integral uniformity and differential uniformity: I.U.= (MAX counts --- MIN counts) x 100 MAX counts + MIN counts D.U = (+ MAX counts) --- (- MIN counts) x 100 MAX counts + MIN counts A search for the pixels with maximum (MAX) and minimum (MIN) counts in the smoothed and masked UFOV image and in CFOV is quickly accomplished in Cartesian coordinates. Intrinsic spatial resolution Changes in the intrinsic spatial resolution of a scintillation camera are expected much less frequently than uniformity changes (practicality of NEMA). Localized changes in resolution are after coupled to degradation in camera uniformity. For these measurements the NEMA phantom is placed directly on crystal face and point source of Tc-99m is centered at least 5 UFOV diameter above the a above the phantom as described by the NEMA protocol. Center 20% window on the photopeak. The phantom is aligned with either the or camera axis. The above experiment is applied on other radionulides used in this study. A Full-Field of view image is acquired in (512x512) computer matrix size to density 15 millions counts <<each image is acquired with the 4 different correction tables of radionuclides used in the study>>. The line spread image used to measure intrinsic spatial resolution by making a profile crossing line in the digital image of intrinsic resolution phantom to generate line spread function that is used to characterize spatial resolution in terms of its FWHM and the results reported in mm for UFOV. These measurements are applied on ADAC gamma camera. In the case of SELO gamma camera intrinsic spatial resolution is determine by visual inspection of widths of the smallest bars of the quadrant bar-phantom that the scintillation

gamma camera can resolve in the x or y directions. A Full-Field of view image is acquired in (64x64) computer matrix size to density 15 millions counts <<each image is acquired with the 4 different correction tables of radio nuclides used in the study>>. So intrinsic spatial resolutions in x or y directions can be estimate in the terms of the full width at half maximum of the line spread function, using relationship: FWHM= 1.75B ((according to NEMA)). Where B is the width of the smallest bars that the camera can resolve. Intrinsic spatial linearity: Resolution/linearity phantom was used by placing it directly on the crystal face and point source of Tc-99m is centered at least 5 UFOV diameter a above the phantom as described by the NEMA protocol. Center 20% window on the photopeak. The phantom is aligned with either the or camera axis. The above experiment is applied on other radionulides used in this study and with ADAC gamma camera only. A Full-Field of view image is acquired in (512x512) computer matrix size to density 15 millions counts <<each image is acquired with the 4 different correction tables of radionuclides used in the study>>. The spatial linearity is determined by making horizontal profile crossing the image of the lines (UFOV) so we can determine the standard deviation of all the peak separations from each other. All results are given in millimeters. (B) SPECT STUDY Tomographic uniformity: The materials required for this study are Data Spectrum Corporation Jaszczak SPECT phantom ((a total performance phantom)). This phantom filled with 10-15 mCi of uniformity mixed solution of Tc-99m. The acquisition matrix used is 64x64 and the number of views to 64 over 360 <<using a circular orbit>>. Asymmetrical analyzer window to 20% is used and total time for a 64 angle acquisition is 32 minutes. Correction table of Tc-99m is used. The SPECT study repeated for times by same acquisition parameters with other used correction tables. The data are reconstructed by using a Hann or Hamming filter. integral uniformity is calculated by same equation used to calculate I.U. in planer study. All above steps repeated with different correction tables used in this study. Tomographic contrast: Tomographic contrast is an important indicator of how well system is performing with respect to detection of small lesions. To calculate the contrast we must select the transverse section where the cold spheres are most clearly defined <<noting the number of sphere that can be visualized>>. For each slice determine the mean counts of the sphere and the mean counts of background neighborhood of the sphere so Contrast = ( Mean sphere counts - Mean Bckg. Counts) ( Mean sphere counts + Mean Bckg. Counts) The calculation of the contrast value is done in image of Tc-99m for each correction table.

Results and discussions

Parameter (1) Uniformity (ADAC Camera) A) Acquisition of Tc-99m flood source versus different nonuniformity correction Maps: The measured Integral and Differential Uniformity (UFOV and CFOV) for Tc-99m flood source image after applied different non-uniformity correction tables are shown in Table (3.1 ) . The results are demonstrated using a histogram display for the effect of using different non-uniformity correction tables on planer image using Tc-99m flood source Fig.(3.1)
Table (3.1): Integral Uniformity and Differential Uniformity (UFOV AND CFOV), of ADAC CAMERA with (Tc-99m) at all variable flood correction.

UFOV Integral uniformity %

UFOV Differential uniformity % ROW D.U % COLUMN D.U % CFOV Integral uniformity %

CFOV Differential uniformity% ROW D.U % COLUMN D.U %

Correction type
Tc99m correction 2.9 TL-201 correction 5.99 Ga-67 correction 7.34 I-131 correction 8.68

2.96 4.13 4.18 4.83

2.2 4.3 4.71 5.71

2.26 3.63 6.69 7.59

1.68 2.72 3.72 4.48

1.78 2.78 4.19 4.57

fig (3.3 ) the effect of increment of using different correction tables on(UFOV AND CFOV) with uniformity of TL-201 source in ADAC CAMERA

Tc-99m source
10

7 u n if o r m ityin%

6 I.U 5 ROW D.U COLUMN D U

2 UFOV CFOV UFOV CFOV UFOV CFOV UFOV CFOV

*Bars show means *Error bars show mean+/standard deviation

0 Tc-99m corr. TL-201 corr. correction type Ga-67 corr. I-131 corr.

Di Discussion
Qualitative analysis:- We found that the best intrinsic image quality was obtained when using the Tc-99m flood source with Tc-99m correction map. The Tc-99m flood image

versus TL-201 correction map showed a non-uniform image; while with Ga-67 and I-131 correction maps showed worst non-uniform images. Quantitative analysis:- - There are statistical significance differences in the mean uniformity values of Tc-99m flood image when applied Tc-99m correction map and in correlation with all other correction types used in the present study. From the collected data for both CFOV and UFOV it was found that the values of intrinsic flood field integral or differential uniformity in response to Tc-99m flood source varied significantly with each correction tables. B) Acquisition of Thallium-201 flood source versus different correction Maps : The results are demonstrated using a histogram display for the effect of using different non-uniformity correction tables on planer image using TL-201 flood source Fig. (3.3).
TL-201 source 6

4 Uniform ity in % I.U. 3 ROW D.U ColumnD.U 2 UFOV 1 CFOV UFOV CFOV UFOV CFOV UFOV CFOV

0 Tc-99m correction TL-201 correction Type of correction Ga-67 correction I-131 correction

*Bars show means *Error bars show mean+/- standard deviation

fig (3.3 ) the effect of increment of using different correction tables on(UFOV AND CFOV) with uniformity of TL-201 source in ADAC CAMERA

Discussion
Qualitative analysis:- We found that the best intrinsic image quality was obtained when using the TL-201 flood source with TL-201 correction map. The TL-201 flood image versus I-131 correction map showed the worst image; while TL-201 flood image with Tc99m correction map showed image look like image of TL-201 flood source with Ga-67 correction. Quantitative analysis:- There are statistical significance differences in the mean uniformity values of TL-201 flood image when applied TL-201 correction map and in correlation with all other correction types used in the present study. From the collected data for both CFOV and UFOV it was found that the values of intrinsic flood field integral or differential uniformity in response to TL-201 source also varied significantly with each correction tables. C) Acquiring Gallium-67 flood source versus different correction maps:-

The results are demonstrated using a histogram display for the effect of using different nonuniformity correction tables on planer image using Ga-67 flood source Fig.(3.5).
Ga-67 source
10

7 U n i f o r m i t yi n %

I.U. ROW D.U

COLUMN D U

3 UFOV 2 UFOV CFOV UFOV CFOV UFOV 1 CFOV CFOV

*Bars show means *Error bars show mean+/- standard deviation

0 Tc-99m correction TL-201 correcton Ga-67 correction I-131 correction Type of correction

fig (3.5 )the effect of increment of using different correction tables on uniformity of Ga-67 source with (UFOV AND CFOV) in ADAC CAMERA

Discussion
Qualitative analysis:- We found that the best intrinsic image quality was obtained when using the Ga-67source with Ga-67correction. The Ga-67 flood image versus I-131 correction map showed the worst non-uniform image in comparison with other images. Ga-67 flood image versus TL-201 correction map showed better image in comparison with image of Ga-67flood image versus Tc-99m correction, which showed a non-uniform image.. Quantitative analysis:- There are statistical significance differences in the mean uniformity values of Ga-67 flood image when applied Ga-67 correction map and correlation with all other correction types used. D) Acquiring Iodine-131 flood source versus different correction Maps : The results are demonstrated using a histogram display for the effect of using different non-uniformity correction tables on planer image using of I-131 flood source Fig.(3.7).
I-131 SOURCE
8

U n i f o r m i t yi n %

5 I.U ROW D.U. 4 COLUMN D U

UFOV

CFOV

UFOV

CFOV

UFOV

CFOV

UFOV

CFOV

*Bars show means *Error bars show mean+/standard deviation

0 Tc-99m correction TL-201 correction Ga-67 correction I-131 correction Type of correction

Fig.(3.7 )the effect of increment of using different correction table on uniformity of I-131 source with (UFOV and CFOV) in ADAC CAMERA

Discussion
Qualitative analysis:- We found that the best intrinsic image quality was obtained when using the I-131 flood source with I-131correction map. The I-131 flood image versus TL201 correction map showed the worst non-uniform image in comparison with other images. I-131 flood image versus Ga-67 correction map showed non-uniform image but better than the image of I-131source with Tc-99m correction. Quantitative analysis:- There are statistical significance differences in the mean uniformity values of I-131 flood image when applied I-131 correction map and in correlation with all other correction types used in these study. The mean uniformity of I131 flood source using TL-201 correction map was the worst than with Tc-99m correction than with Ga-67 correction. The qualitative and quantitative studies showed more or less the same finding. Parameter (2): Uniformity measurements using SELO Gamma Camera A) Acquisition of Tc-99m flood source versus different correction Maps: The measured mean Integral Uniformity and Differential Uniformity (UFOV and CFOV) for Tc-99m flood source image after applied non-uniformity correction tables are shown in Table (3.17). The results are demonstrated using a histogram display for the effect of using different correction tables on planer uniformity of Tc-99m source Fig. ( 3.9).

Discussion
Qualitative analysis:The qualitative and quantitative analysis of the images of radioisotopes represented in these study with SELO CAMERA showed that they agree in all points, so we can using for example only qualitative analysis of Tc-99m source. We found that the best intrinsic image quality was obtained when using the Tc-99m flood source with Tc-99m correction map. The Tc-99m flood image versus I-131 correction showed the worst non-uniform image. Tc-99m flood image versus TL-201 correction showed non-uniform image but better than the image of Tc-99m flood image versus Ga-67 correction.. Quantitative analysis:- There are statistical significance differences in the mean uniformity values of Tc-99m flood image when applied Tc-99m correction map and in correlation with all other correction types used in the present study. The values of mean integral uniformity and differential uniformity of Tc-99m source with Tc-99m correction (UFOV or CFOV) in SELO CAMERA are the optimum image for the clinical use, but the mean uniformity of Tc-99m flood source using I-131 correction map was the worst than with Ga-67 correction than with TL-201 correction. From the data we can say, any correction tables than Tc-99m correction with Tc-99m source are not accepted for the clinical use of this version of SELO CAMERA.

Table (3.16): Integral Uniformity and Differential Uniformity (UFOV AND CFOV) of SELO CAMERA (Tc-99m) at all variable flood correction
Correction type Tc99mcorrection UFOV Integral uniformity % 3.3 UFOV Differential uniformity % 2.103 CFOV Integral uniformity % 2.933 CFOV Differential uniformity% 1.93

TL-20 correction

15.297

13.257

10.79

9.395

Ga-67 correction

16.037

14.47

11.303

9.593

I-131 correction

32.577

24.091

18.223

15.647

Tc-99m source
35

30

25

U n if o r m it yin%

20

I.U(UFOV) D.U(UFOV) I.U(CFOV) D.U(CFOV)

15

10

*Bars show means *Error bars show mean+/- standard deviation

0 Tc-99m corr. TL-201 corr. Correction type Ga-67 corr. I-131 corr.

fig(3.9): The effect of increment of using different correction tables on(UFOV and CFOV) uniformity of Tc-99m source in SELO CAMERA.

Parameter (3) Resolution (ADAC Camera)

A) Acquisition of Tc-99m flood source versus different non-uniformity correction Maps: Table (3.33) shows the mean spatial resolution of Tc-99m flood source in ADAC CAMERA using different non-uniform correction map. The histogram of effect of increment of using different non-uniformity correction tables on planer RESOLUTION measurements of the Tc-99m flood image in ADAC CAMERA is shown in Fig(3.15)
Table (3.33) Planer resolution of ADAC CAMERA (Tc-99m source) FWHM pixels 2.41 2.76 3.00 3.2 FWHM Mm 2.774 3.17 3.45 3.65

Type of correction Tc-99m correction TL-201 correction Ga-67 correction I-131 correction

3.5

R e s o l u t i o ni nm m

2.5

1.5

0.5

0 Tc-99m correction TL-201 correction Ga-67 correction I-131 correction Type of correction

Fig.( 3.15 ) the effect of increment of using different correction tables on spatial resolution of Tc-99m source in ADAC CAMERA.

Discussion
Qualitative analysis:- The best intrinsic resolution value was obtained when acquiring the Tc-99m flood using Tc-99m non-uniformity correction map. From In addition; the Tc-99m flood image corrected using I-131 non-uniformity map has showed the worst resolution image in comparison with other images. The acquisition of the Tc-99m flood using TL-201 and Ga-67 non-uniformity correction map showed images with nearly the same quality. Quantitative analysis:- Its obvious that there are no statistical significance differences for the measured resolution of the Tc-99m phantom image in correlation with Tc-99m or Tl201 correction tables. Our results showed that there are statistical significance differences (p= 0.010) with the measured resolution between Tc-99m and Ga-67 correction map. There are statistical significance differences for the mean resolution between Tc-99m and I-131 correction (p=0.00) B) Acquisition of TL-201 flood source versus different nonuniformity correction Maps: The histogram of effect of increment of using different non-uniformity correction tables on planer RESOLUTION measurements of the TL-201 flood image in ADAC CAMERA is shown in Fig. (3.17)
3.5 3.4

3.3

F W H M i n m m

3.2

3.1

2.9

2.8

2.7

Tc-99m correction

TL-201 correction

Ga-67 correction

I-131 correction

Type of correction

Fig.(3.17 ) the effect of increment of using different correction tables on spatial resolution of TL-201 source in ADAC CAMERA.

Discussion
Qualitative analysis:- The best intrinsic resolution quality was obtained when acquiring the TL-201 flood using TL-201non-uniformity correction map and Tc-99m nonuniformity correction map. The acquisition of the TL-201 flood using Ga-67, and I-131 non-uniformity correction map showed images with nearly the same quality. Quantitative analysis:- There are statistical significance differences (p=0.002) for the measured resolution of the Tl-201 phantom image in correlation with TL-201 or Tc-99m correction tables.. TL-201 correction only with TL-201 source can give optimum resolution for this version of gamma camera.

Parameter 4: Resolution

(SELO CAMERA)

A) Acquisition of Tc-99m flood source versus different non-uniformity correction Maps: Table (3.41) shows the mean spatial resolution of Tc-99m flood source in SELO CAMERA using different non-uniform correction map.The histogram of effect of increment of using different non-uniformity correction tables on planer RESOLUTION measurements of the Tc-99m is shown in Fig.(3.23).
Type of correction B The smallest bar that the camera resolve 3.25 3.25 4.25 4.25 FWHM in mm FWHM = 1.75B

Tc-99m correction TL-201 correction Ga-67 correction I-131 correction

5.6875 5.6875 7.4375 7.4375

Table (3.41 ) Planer resolution of SELO CAMERA (Tc-99m source)

F W H Min m m

0 Tc-99m correction TL-201 correction Ga-67 correction I-131 correction

Type of correction

Fig.( 3.23 ) the effect of increment of using different correction tables on spatial resolution of Tc-99m source in SELO CAMERA.

Discussion
Qualitative analysis:- The present study showed that the spatial resolution measured at the Tc-99m phantom that acquired with Tc-99m non-uniformity correction table and with TL-201 correction in SELO CAMERA were the same. The spatial resolution of Tc-99m phantom that acquired with Ga-67 correction and with I-131 correction in SELO CAMERA was the same, and showed worse resolution Repeating the test for several times give the same results so we can not do statistical calculations for this study and there are no significant variations at full width at half maximum with variation in correction tables.

(3): Test for linearity

Table (3.45) shows the measured mean spatial distortion of Tc-99m, TL-201, Ga-67 and I-131 phantom in ADAC CAMERA at all variable non-uniformity correction maps. The histogram of the effect of increment of using different correction tables with different radioisotopes on the linearity is shown in fig. (3.28 ). Qualitative analysis:-The qualitative analysis of the images represented in figure (3.29) showed that the sharpens of the lines unchanged with all sources used in these study at different correction tables. Quantitative analysis:-From the collected data it was found that: As the correction table varied with the same source, the spatial distortion remains with constant value which was equal to (= 1.18mm). There are no statistical data that can be found from the results of linearity test because all values were constant.

Table (3.45 ) the collected data for linearity test:

Type of correction Type of source Spatial distortion (linearity) in pixels
1

Spatial distortion (linearity) in mm

1.18

Tc-99m correction

Tc-99m source TL-201 source Ga-67 source I-131 source Tc-99m source TL-201 source Ga-67 source I-131 source Tc-99m source TL-201 source Ga-67 source I-131 source Tc-99m source TL-201 source Ga-67 source I-131 source

TL-201 correction

1.18

Ga-67 correction

1.18

I-131 correction

1.18

Fig (3.28 ): the effect of increment of using different correction tables with different radioisotopes in ADAC CAMERA on the linearity
Planer Linearity Of ADAC CAM ERA
1.4

1.2

s p a t i a ld i s t o r t i o n ( l i n e a r i t y )i n m m

0.8

Tc-99m source TL-201 source Ga-67 source I-131 source

0.6

0.4

0.2

Tc-99m correction

TL-201 correction

Ga-67 correction

I-131 correction

Type of correction

Tc-99 correction

I-131 correction

Fig. (3.29): Image samples of Linearity test of Tc-99m source with different correction tables in ADAC CAMERA.

Part 2: Test for SPECT images

Parameter 5: SPECT Uniformity in (ADAC CAMERA) A Acquisition of Tc-99m SPECT Phantom versus different non-uniformity correction Maps: The measured
Integral and Differential Uniformity (UFOV and CFOV) for Tc-99m flood source image after applied different non-uniformity correction tables are shown in Table 3.46, 3.47. 50The results are demonstrated using a histogram display for the effect of using different non-uniformity correction tables on planer image using Tc-99m flood source Fig. (3.30, 3.31).
Table (3.46) SPECT uniformity of ADAC CAMERA (Tc-99m source) for UFOV

Type of correction Tc-99m correction TL-201 correction Ga-67 correction I-131 correction

Integral uniformity % 4.68 5.78 8.25 10.96

Differential uniformity Row D.U. % Column D.U. % 3.72 4.00 4.19 4.32 6.08 7.8 6.26 8.4

Table (3.47) SPECT uniformity of ADAC CAMERA (Tc-99m source) for CFOV

Type of correction Tc-99m correction TL-201 correction Ga-67 correction I-131 correction

Integral uniformity % 3.82 4.8 6.94 8.46

Differential uniformity Row D.U. % Column D.U. % 3.34 3.5 3.94 4.75 6.48 3.99 5.17 6.90

Tc-99m source
9

10

SPECT UNIFORMITY IN %

Uniformity in %

6 I.U 5 ROW D.U ColumnD.U 4

I.U D.U ROW

D.U COLUMN

0 Tc-99m correction Tl-201 correction Ga-67 correction I-131 correction

0 Tc-9m correction TL-201 correction Ga-67 correction I-131 correction

Type of correction

Type of correction

Fig (3.30) the effect of increment of using different correction tables on uniformity test for UFOV with Tc-99m source in SPECT STUDY (ADAC CAMERA)

Fig (3.31 ) the effect of increment of using different correction tables on uniformity test for CFOV with Tc-99m source in SPECT STUDY (ADAC CAMERA)

Discussion Qualitative analysis:- The effect of using different correction tables with Tc-99m
SPECT phantom (hot and linearity / uniformity), showed that the best image set is obtained when using Tc-99m correction map , while the image with I-131 correction showed an undesired sharpens which effects the image quality for the two inserts. The quality of the image when using Tc-99m phantom with TL-201 correction is decrease in comparison with the image with Tc-99m correction, but still better than the image with I131 correction. The Ga-67 correction showed that the image quality nearly like the image quality of TL-201 correction. Quantitative analysis:- The values of mean integral uniformity and differential uniformity of Tc-99m source with Tc-99m correction (UFOV or CFOV) in ADAC camera is the optimum with this correction table for obtaining an optimum uniform image for the clinical use of this version of gamma camera. There is statistical significance differences for the mean uniformity between Tc-99m correction and TL-201 correction, but in the case of DIFFERENTIAL UNIFORMITY (COLUMN) in both UFOV and CFOV there are no statistical significance. Also there are statistical significance differences for the mean between Tc-99m correction and Ga-67 correction, and also there are statistical significance differences for the mean between Tc99m correction and I-131 correction.

Parameter 6: Test for Contrast for SPECT images:

Table (3.51) shows the measured mean contrast of Tc-99m phantom in ADAC CAMERA, at all variable flood correction. The histogram of effect of increment of using different correction tables on contrast of Tc-99m source in ADAC CAMERA for SPECT STUDY is shown in Fig. (3.34 ). Table (3.51) the collected data for contrast Type of correction Tc-99mcorrection TL-201 correction Ga-67 correction I-131 correction Sphere mean count (count/pixel) 4632 4035 6198 6166 Bckg mean count (count/pixel) 5420 4809 7536 7805

Contrast -0.0784 -0.0875 -0.0974 -0.1173

Tc-99m source
0 Tc-99m correction TL-201 correction Ga-67 correction I-131 correction

-0.02

-0.04

co n tra st v a lu e

-0.06

contrast

-0.08

-0.1

*Bars show Means *Error Bars show Mean +/- standard deviation
Type of correction

-0.12

Fig (3.34 ): The effect of increment of using different correction table with Tc99m source on the contrast values in SPECT.

Discussion
Qualitative analysis:- Qualitative analysis of the images for tc-99m SPECT phantom with different acquired at different non-uniformity correction tables, showed no variation in this study. So we preferred to apply quantitative analysis to be more accurate Quantitative analysis:- Our results showed that the contrast of Tc-99m phantom varied with different correction table. For sure the contrast of Tc-99m source with Tc-99m correction is the best. While the contrast with TL-201 source was decreased. There are no significance difference for the mean contrast between the Tc-99m correction and TL-201 correction; while there are statistical significance difference for the mean contrast between Tc-99m correction and Ga-67 and I-131 correction.

Summary & Conclusion

The present study showed the following : This study presents a complete description of using different non-uniformity correction tables acquired using different radionuclides and its effect on image quality produced by gamma cameras. Most of the gamma camera quality control parameters (uniformity, resolution and contrast) are influenced by variation in the correction tables, while other parameters not affected by this variation like linearity. Our study conclude that the variation in uniformity measurement is not significant and accepted in clinical use when using a digital camera,.(Elpida S. et al. 1987). It is obvious also that the spatial resolution of both types of gamma camera (ADAC, SELO) was influenced by the change in correction tables. There were a significance differences when using different radionuclide sources except in the case of Tc-99m correction and TL-201 correction only with using Tc-99m source gave optimum resolution for this version of ADAC gamma camera. The quality control parameter that is not influenced by the change in the correction table is the linearity, but it is accepted for the clinical use of this version of ADAC CAMERA which is equal to= 1.78mm For SPECT study, both parameters uniformity and contrast that were checked in this study were influenced by using different correction tables with the same source. So for Tc-99m phantom with Tc-99m correction, the uniformity values for either integral or

differential uniformity are the optimum for obtaining an ideal uniform image for the clinical use of this camera. The uniformity of Tc-99m images with TL-201 correction is also accepted in the clinical work, with Ga-67 and I-131 correction the uniformity values for either integral or differential became worse. Different statistical parameter was calculated using computerized statistical software program (SPSS, version 11). i) The ANOVA test (Analysis of variance): From this test, we calculated the mean value for each physical parameter and the standard deviation for each physical parameter. ii) POST HOC test: from this test we calculated the significance of difference knowing that the value is significant at the level > 0.05 Finally; we conclude that, for the optimum use , its recommended to image patients using non-uniformity correction map the same radionuclide that injected for the patients.

References
(1) AAPM. Rotating Scintillation Camera SPECT Acceptance Testing and Quality Control. Published for the American Association of Physicists in Medicine by the American Institute of Physics, New York. AAPM ReportNo.22[1987]. (2) Australian and New Zealand Society of Nuclear Medicine (ANZSNM) Technical Standards Subcommittee. Minimum Quality Control Requirements for Nuclear Medicine Equipment. Version 57 November 1999. (3) Ayyangar K, Patel J, Park CH: Quality control of gamma Cameras using computer. MedPhys 6:335, 1979 (abs). (4) Bhagwat D. Ahluwalia. Tomographic Methods in Nuclear Medicine: Physical Principles, Instruments, and Clinical Application. 1st. edition. Raven Press 1989. (5) Bieszk JA. Performance changes of an anger camera in magnetic fields up to 10G.JJVycAferf1986;27:l902.-1907. (6) Bone FJ, Graham KD, Dowdey JE. Image aberrations produced by multichannel collimators for a scintillation camera. Radiology 1971 ;98: 329-334. (7) Brookeman VA: Computer and quality control in nuclear medicine. Semin NuclMed 8:113-124, 1978. (8) Brooks R A, Di Chiro G. Theory of image reconstruction in computed tomography. Radiology; 117:561-572, (1975). (9) Brown PH, Royal HD. Re: Distortion of bar phantom image by collimator [letter]. JNuclMedW^ 20:1100-1101. (10) Busemann-Sokole E, Cradduck TD: Interpretation of the NEMA protocols for Scintillation Camera Performance, JNuclMed 24:973-974, 1983. (11) Bushberg JT, Seibert JA, Leidholdt EM Jr, Boone JM. The Essential Physics of Medical Imaging. Baltimor: Williams and Wilkins; 1994. (12) Cormack AM. Representation of a function by its line integrals, with some radiological applications. JApplPhys 1963;34:2722-2727. (13) Cormack AM, Reconstruction of densities from their projections, with applications in radiological physics. Phys Med. Siol. 1973; 18:195-217. (14) Croft BY. Single Photon Emission Computed Tomography. Chicago: Year Book Medical Publishers; 1986. (15) English RJ, Brown SE. SPECT: single photon emission computed tomography: a primer. New York, NY: Society of Nuclear Medicine; 1986. (16) English RJ, Brown SE. SPECT: single photon emission computed tomography: a primer. New York, NY: Society of Nuclear Medicine; 1990. (17) Erickson J. Imaging systems. In: Harbert J, da Rocha AFG, eds. Textbook of Nuclear Medicine, Volume I: Basic Science Philadelphia: Lea and Febiger; (1984).

(18) Faber TL, Stokely EM, Templeton G H, et al. Quantification of three-dimensional left ventricular segmental wall motion and volumes from gated tomographic radionuclide ventriculograms . J. Nucl. Med.; 30:638-649. (1989). (19) G.F. Knoll, N.C. Bennett, and D.R. Strange, "Real-Time Correction of Radioisotope Camera Signals for Nonuniformity and Nonliearity", J. Nucl. Med, 19,746 (1978). (20) Gerd MuehIIehner, Robert, H. Walks and Richard sans. J. Nucl. Med .22-.72-77,(1981). (21) Gullberg GT. An analytical approach to quantify uniformity artifacts for circular and noncircular detector motion in single photon emission computed tomography imaging. Med. Phys.; 14:105-114, (1987). (22) Gopal B. Saha, Physics and Radiobiology of Nuclear Medicine. Second Edition. Cleveland- 2000. (23) Graham LS, Olcha, Usier JM, et al: Automatic field uniformity corrections in scintillation cameras. SPIE 152; 127-131, 1978. (24) Graham LS. A rational quality assurance program for SPECT instrumentation. In: freeman LM, ed. Nuclear Medicine Annual 1989 New York, NY: Raven Press, Ltd.; 81-108: 1989. (25) Groch MW, Marshall RC, Erwin WD, et, al. Quantitative Gated blood pool SPECT for the assessment of coronary artery disease at rest. J. Nucl. Cardiol. . 5:567-573, (1998). (26) Halama JR, Henkin RE. Quality assurance in SPECT imaging. Applied Rad. 41-50, (1987). (27) Halama JR, Madsen MT. Is your gamma camera.- SPECT system installed and operating optimally. Applied Rad. 35-41, (1992). (28) Harkness BA, Rogers W.L., Clinthorne H.N., Keyes JW Jr. SPECT: quality control procedures and artifact identification. J Nuc Med Tech. 11; 55-60, (1983). (29) Hasegawa BH, Kirch DL, Lefree MT, et al: Quality control of scintillation camera using aminicomputer. J Nuct Med 22: 1075-1080, (1981). (30) Henry Rusinck , SPECT reconstruction techniques. In Elissa, Lipcon, Joseph J. Sanger, editors. 2nd ed Clinical SPECT Imaging. Raven New York, 43-78, (1995). (31) Hine GJ. An ortho test pattern for quality assurance in nuclear medicine. Washington D.C., Nuclear Associates Pamphlet No. 265K., (1986). (32) Hines H, Kayayan R, Colsher J, Hashimoto D, Schubert R, Femando J, Simcic V, Vemon P, and Sinclair RL. National Electrical Manufacturers Association recommendations for implementing SPECT instrumentation quality control. J Nucl Med Tech. 27(1): 67-72,(1999).[Medline]. (33) Hounsfield GN. Computerized transverse axial scanning (tomography). I. Description of the systemBrJRadiol. 46:1016-1022,(1973). (34) Button BF, Cormack J: Computer- assisted quality control of gamma cameras. AustNZMed 9:624, 1979 (abs). (35) IPSM " Quality Control of Gamma Camera and Associated Computer System" (Institute of Physical Scinences in Midicine. New York) 1992; report no.65, 1-13. (36) IPSM " Quality Control of Gamma Camera and Associated Computer System" (Institute of Physical Scinences in Midicine. New York) 1992; report no. 66, 18-56. (37) Jansson LG, Parker RP: Pitfalls in gamma camera field uniformity correction. BrJRadiol. 48: 408409, (1975) (Letter to the Editor). (38) Jaszczak RJ, Murphy P.H.,.Huard D., Burdine J.A, RadionucHde emission computed tomography of the head with Tc-99m and scintillation camera. J. Nucl. Med. vol.18: 373-380, (1977). (39) Jaszczak RJ, Coleman RE, " Selected Processing Techniques for Scintillation Camera Based SPECT Systems", in Single Photon Emission Computed Tomography and other Selected Topics, Sorenson JA, Ed., Society of Nuclear Medicine, New York. pp. 45-59, (1980). (40) Jaszczak RJ, Whitehead FR., Lim CB., Coleman R.E. Lesion detection with single- photon emission computed tomography (SPECT) compared with conventional imaging. JNuclMed. 23:97-102 (1982). (41) Jaszczak RJ, Coleman RE., Single photon emission computed tomography (SPECT): principles and instrumentation. Invest Radiol. 20: 897-910, (1985). (42) Joint NCDRH and State quality assurance surveys in nuclear medicine: Phase 1 scintillation cameras and dose calibrators. U.S. Dept. of Health and Human Services, HHS Publications FDA 83-8204, (1983). (43) Keyes JW, Gazella GR, Strange DR: Image analysis by on-line minicomputer for improved camera quality control. J Nucl Med 13:525-527, (1972). (44) Keyes JW. Section 3, Operational Guidelines: Single-Photon Emission Computed Tomography (SPECT). In Wagner HN, Szabo Z, Buchanan JW (eds), Principles of Nuclear Medicine. Philadelphia; Saunders, 1995;332-341.

(45) King MA, Schwinger RB, Doherty PW, et al. Two-dimensional filtering of SPECT images using the Metz and Wiener filters. J Nucl Med. 25:1234-1240. (1984). (46) Knoll GF, Bennett MC, Strange DR: Real-time correction of radioisotope camera signals for nonuniformity and nonlinearity. J Nucl Med. 19:746;1978(abs). (47) Knoll GF, Bennett MC, Koral KF, et al: processing. In VI International Conference on Information Processing in Medical Imaging, Paris, in Press. (48) Koral KF, Schrader ME, Knoll GF: A measure of Anger-camera linearity: results with and without a corrector. J Nucl Med 22:1069-1074, (1981). (49) Kuhl DE, Edwards RQ. Image separation radioisotope scanning. Radiology. 80:653-662, (1963). (50) Kuhl DE, Edwards RQ. Cylindrical and section radioisotopes scanning of the liver and brain. Radiology. 83:926-935, (1964). (51) Kuhl DE, Edwards RQ, Reorganizing data from transverse section scans of the brain using digital processing. Radiology. 91:975-983, (1968). (52) Lalush DS, Tsui BM. Performance of ordered-subset reconstruction algorithms under conditions of extreme attenuation and truncation in myocardial SPECT . J Nucl. Med. 41:737-744, (2000), (53) Madsen MT, Park CH. Enhancement of SPECT images by Fourier filtering the projection image set. JNuclMed. 26:395-402, (1985). (53) Minimum quality control requirements for nuclear medicine practices(Australin Society of Nuclear Medicine) (54) Morrisou LM, Bruno FP, Mauderliw: source of gamma camera image inequalities. JNuclMed 12:785-791, (1971). (55) Muehllehner G, Wake RH, Sano R: Standards for performance measurements in scintillation cameras. J Nucl Med11\ 72-77, (1981). (56) Murphy PH. Acceptance testing and quality control of gamma cameras, including SPECT: J Nucl Med. 28:1221-1227, (1987) [Medline]. (57) NEMA standards publication NUI- 1994 (P.I) performance measurements of scintillation camera Washington, DC National Electrical Manufacturers Association, (1994). (58) NEMA standards publication NUI - 1994 (P.2) performance measurements of scintillation camera Washington, DC National Electrical Manufacturers Association, (1994). (59) O'Conner MK, Oswald WM. The line resolution pattern: A new intrinsic resolution test pattern for nuclear medicine. JNucl Med 29: 1856-1859, (1988). (60) O'Conner MK, Vermeersch C: Critical examination of the uniformity requirements for singlephoton emission computed tomography. Med Phys. 18: 190-197, (1991). (60) Operating Manual for the Modumed System. Ann Arbor, MI, Medical Data Systems, (1976). (61) Oppenheim BE, Appledom CR. Uniformity correction for SPECT using a mapped cobalt-57 sheet source. J Nucl Med. 26:409-415, (1985).[Medline] (62) Padikal TN, Ashare AB, Kereiakes JG: Field flood uniformity correction: benefits or pitfalls. J Nucl Med 17:653-656, (1976). (63) Paras P, Hine GJ, Adams R. BRH test pattern for the evaluation of gamma camera performance. J Nucl Med. 22: 468-470, (1981). (64) Phelps ME. Emission computed tomography. Semin Nucl Med. 7:337-365. (1977). (65) Quantitation of SPECT performance: Report of Task Group 4, Nuclear Medicine Committee (USA). (66) Quality Control of Scintillation Cameras. U.S. Dept. of Health, Education and Wetfar, HEW publications FDA 76-8046, (1976). (67) Rodgers WL, Clinthorne NH, Harkness BA, Koral KF, et al. Field-flood requirements for emission computed tomography with an anger camera. J Nucl M^.(1982):23:162.-168.[Medline] (68) Rodgers WL, Clinthorne NH, Slamos J, et at. Performance evaluation of SPRINT: a single-photon ring tomograph for brain imaging. JNuc Med 1984:25:1013.-1018.[Medline] (69) Roger R Fulton, Brain F. Hutton. The computer and its application in nuclear medicine. In I.P.C. Murray, P.J. Ell, editors. Nuclear Medicine in clinical in Clinical Diagnosis and Treatment. 2nd edition (1998), 1571-1601. (70) Rollo FD. Evaluating imaging devices. In: Rollo FD. ed. Nuclear Physics, Instrumentation and Agents. St. Louis: Mosby; (1977);436. (71) Rollo FD, Harris CC. Factor's affecting image formation. In; Rollo FD, ed, Nuclear Physics, Instrumentation and Agents. St. Louis: Mosby; (1977):387.