Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Murthy, K.
Right arrow Articles by Gagnon, J. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Murthy, K.
Right arrow Articles by Gagnon, J. H.
(Radiology. 2000;215:280-285.)
© RSNA, 2000

Technical Developments

Positron Emission Mammographic Instrument: Initial Results1

Kavita Murthy, MSc, Marianne Aznar, MSc, Alanah M. Bergman, MSc, Christopher J. Thompson, DSc, James L. Robar, MSc, Robert Lisbona, MD, Antoine Loutfi, MD and Jean H. Gagnon, MD

1 From the Montreal Neurological Institute (K.M., C.J.T., J.L.R.), Medical Physics Unit (M.A., A.M.B., C.J.T.), and the Departments of Nuclear Medicine (R.L.), Surgical Oncology (A.L.), and Radiology (J.H.G.), the Royal Victoria Hospital, McGill University, 3801 University St, Rm 798, Montreal, Quebec H3A 2B4, Canada. Received September 29, 1998; revision requested December 23; final revision received August 4, 1999; accepted August 11. Supported in part by the National Cancer Institute of Canada, Canadian Breast Cancer Research Institute grant 6139. Address reprint requests to K.M. (e-mail: kavita@bic.mni.mcgill.ca).


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Performance characteristics of a positron emission mammographic (PEM) instrument were studied. This dedicated metabolic breast imaging system has spatial resolution of 2.8-mm full width at half maximum (FWHM), coincidence resolving time of 12-nsec FWHM, and absolute efficiency of 3%. Hot spots with diameter of 16 mm in a phantom with signal-to-background activity ratio of 6:1 were distinguishable with a scanning time of 5 minutes.

Index terms: Breast radiography, technology, 00.1219


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Findings in current research studies in breast cancer have underscored the importance of early detection in reducing the mortality rates due to this disease (1). The increased uptake of glucose caused by increased glycolysis in certain types of cancer cells has recently led to much interest in the use of positron emission tomography (PET) to detect cancer (2,3). A radioactive analogue of glucose, 2-[fluorine-18]fluoro-2-deoxy-D-glucose (FDG) is commonly used in oncologic applications of PET (4,5). Radiotracer imaging provides a noninvasive method in the screening of women for the presence of the disease, as well as for staging disease and tracking its response to therapy (6,7).

Early detection of cancer requires use of an instrument with very high spatial resolution and efficiency. The positron emission mammographic (PEM) instrument described herein was built in our laboratory on the basis of the basic principles of PET applied to the specific problem of breast cancer detection, and it produces high-spatial-resolution emission images of the breast with use of very low doses of FDG (8–11). In this study, we measured the spatial, timing, and contrast resolution characteristics of images obtained with the PEM instrument in the laboratory. Typical clinical scanning results are also presented.

In the case of PET and single photon emission computed tomography (SPECT) after image reconstructions and all corrections are applied, the image contrast equals the object contrast, provided the object size is more than twice the full width at half maximum (FWHM). These corrections cannot be applied to limited-angle back-projection images such as those obtained with our PEM instrument system. Therefore, scatter contribution cannot be eliminated from the reconstructed image. In previous work (12), we have shown that lesion detection ability is not compromised by performing limited-angle projection. However, the contrast detected will be lower than the true value, which can be obtained when the image is reconstructed from projections (13). In general, for object thickness (compressed breast) T and a tumor diameter D, true contrast (C, defined as the ratio of activity at the tumor to that at the background) at CT, the detector contrast CD is given with the following expression:

This expression applies to transmission imaging systems such as x-ray or planar nuclear medicine imaging systems and to emission images produced with the PEM instrument.

Ideally, if a point source is placed at the center of a cube of which two opposing faces can be considered detectors, then the probability that both annihilation photons would be directed toward the detectors is one-third or 33%. The probability (P) that an incident 511-keV photon will be attenuated (att) in a bismuth germanate (BGO) detector of thickness x and attenuation coefficient µBGO is given with

For pixilated BGO, the effective (eff) value of x must be further reduced to take into account the fraction (f) of mass lost during cutting and surface preparation. For a pair of planar detectors, with isotropic emission of annihilation photons, the probability of detection of both photons is then

In a perfect system with every attenuated photon detected and recorded, Peff represents the maximum achievable efficiency. In the PEM instrument, roughly 37% of BGO mass is lost during crystal preparation (12) (f = 0.37). For 2-cm-deep pure BGO with µBGO of 0.92 cm-1 (14), theoretic system efficiency with Equation (3) is 16%.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
PEM Instrument System Hardware
The PEM instrument design, geometry, and components have been discussed in previous publications by our group (10–12). Components relevant to the subject matter of this study will be described briefly for completeness. The system consists of two planar detectors operated in coincidence. The PEM instrument is mounted on a standard mammographic magnification table attached to the film cassette tray on the mammographic system (Diagnostic UC; Philips Medical Systems, Shelton, Conn). The entire system can be tilted with the mammographic gantry to allow mediolateral oblique views to be acquired if required. Figure 1 includes a photograph of the PEM instrument system installed on a mammographic unit and a cross-sectional diagram of the system and details of patient positioning during clinical scanning.



View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. PEM instrument (PEM-I). Left: Photograph in the mammographic clinic shows the instrument attached to the cassette holder of the x-ray mammographic system. Right: Cross-sectional diagram shows the layout of the detectors and the details of patient positioning during clinical scanning.

 
Detectors.—In the PEM instrument , a 2 x 2 matrix of 36 x 36 x 20 (deep)-mm blocks of BGO is used. Each crystal is cut in a matrix with 0.25-mm-wide slits spaced 2 mm apart on both 36 x 36-mm faces (ie, the faces coupled to and distal to the photomultiplier tube). The BGO blocks are etched and coated with a reflector and optically coupled to a position-sensitive photomultiplier tube (R3941-05; Bridgewater, NJ) that has a 75 x 75-mm entrance window. The effective number of sampling points is 2,452. Owing to the inherent spatial nonlinearities of the position-sensitive photomultiplier tube, elements situated at the periphery of the detector are not sufficiently resolved spatially. The useful field of view of the instrument is 2.5 cm away from its physical edge. Full details of the detectors can be found in reference 11.

Data acquisition system.—The energy thresholding utility of the PEM instrument software is used to window the processed events, apply spatial and energy distortion corrections, and perform limited-angle back projection to produce the final image. Coincidences between any two crystals are accepted, which gives 2,452 possible lines of response with the central plane sampled at a rate of two samples per millimeter. The space occupied by the compressed breast between the detectors is divided into seven imaging planes, and the intersection points of the lines of response with the planes are determined. Seven images are formed by adding a number N to the appropriate locations in a 128 x 128 imaging matrix (10). This number N is inversely proportional to the probability of the annihilation occurring in that plane. This technique is extremely fast, and it allows almost real-time display of the activity distribution (images are updated every 5 seconds). Additionally, it provides important depth information since imaging of the plane closest to an actual site of high activity results in acquisition of the best-focused image of the site. Full details of the acquisition system are in reference 9.

Intrinsic spatial resolution of detector pair.—The spatial resolution of the PEM instrument system was measured by optically coupling one block of prepared BGO crystal to each photo multiplier tube. Two germanium 68 point sources were placed 10 mm apart at the midpoint of the detectors, which were separated by 10 cm. Coincidence data were acquired with an energy window of approximately 300–700 keV. The experiment was repeated at each stage of crystal preparation starting with the uncut BGO block and ending with a block that was cut, acid etched, and filled with white epoxy.

System spatial resolution.—The system spatial resolution was measured by filling capillary tubes with 1-mm inner diameter with radioactive solution of 18F and imaging them with 10-, 8-, and 5-mm separations. The peak-to-valley ratio for each experiment was calculated by drawing profiles through the back-projected image and calculating the average ratio of the local maxima to the local minima.

To compare performance of the PEM instrument system with that of a conventional PET brain scanner (Scanditronix PC2048; GE Medical Systems, Milwaukee, Wis) (15,16) used at the Brain Imaging Centre of the Montreal Neurological Institute, we performed identical experiments with the two instruments. Five capillary tubes with 1-mm inner diameter were filled with 18F solution and placed 5 mm apart; they were imaged with the conventional scanner for 10 minutes with use of a wobbled scanning protocol. The PET images were reconstructed by means of filtered back projection and ramp filtering into a 128 x 128-pixel image. Identical experiments were performed with the PEM instrument with one-fourth the activity and scanning time of 2 minutes.

Contrast resolution.—The contrast resolution of the PEM instrument system was measured by using a custom-made phantom and a spheric 16-mm-diameter FDG hot spot without exterior walls (17). The phantom consisted of a 150 x 150 x 100-mm (high) rectangular box with a fine needle for positioning the hot spot at different positions. The depth of background activity was 70 mm, and initial hot-spot activity was 1.6 MBq. Activity was added to the background to simulate different values of true contrast ratio (hot-spot activity divided by background activity). An energy window of 350–650 keV was used. All activities were measured in a well counter (model 802-3W; Canberra Packard, Mississauga, Ontario, Canada). With the energy window used to measure blood samples during conventional PET scanning, the efficiency of the counter was 0.24 cps/Bq.

System efficiency.—System efficiency was measured by placing a 3-mm-diameter, 33-kBq FDG planar source at the midpoint of the central plane between the two detectors and recording the number of counts detected in a 60-second period as a function of the detector separation with use of an energy window of 350–650 keV. The source activity was measured in the calibrated well counter and corrected for detection efficiency of the counter. The fraction of emitted annihilation photons detected by the PEM instrument detectors was plotted as a function of the detector separation.

System timing resolution.—Timing resolution of the PEM instrument system was measured by means of the fast-slow coincidence technique. A time-to-amplitude converter (model 1443A; Canberra Packard) was used to convert the difference between the arrival times of the timing signals from opposing detectors into a single signal. The amplitude of this signal was proportional to the difference between the arrival times. The time axis was calibrated by inserting a known delay into the stop channel of the time-to-amplitude converter. The output was viewed on a multichannel analyzer.

Patient Scanning
Fifteen female patients (age range, 34–75 years; mean age, 54.8 years) underwent scanning at the Cedars Breast Clinic at the Royal Victoria Hospital in Montreal. This study was approved by the institutional research ethics committee, and participants gave written informed consent. All patients were asked to urinate before the start of scanning to reduce the amount of scattered background radiation that entered the detectors and to minimize absorbed dose in the patient.

One of the patients (age, 75 years) had an abnormal x-ray mammogram that indicated the presence of an undiagnosed suspect dense mass in the right breast. A 75-MBq bolus of FDG was injected, and emission scanning was performed 40 minutes after injection. The patient's breast was positioned in the mammographic system with use of minimal compression. An x-ray mammographic image of the breast was obtained and subsequently digitized. The detectors in the PEM instrument system were positioned over the tumor site, and emission data were collected for 5 minutes. The PEM instrument produced a nearly real-time image of the breast as scanning progressed, with images that were updated every 5 seconds. The best focused of the seven PEM instrument images was scaled to match the size of the digitized x-ray mammographic image, and the two images were coregistered to result in a composite PEM instrument and x-ray mammographic image.


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
PEM Instrument
Intrinsic spatial resolution of detector pair.—The Table summarizes the results of the spatial resolution measurements at each stage of crystal preparation. The spatial resolution of the BGO block improved by 63% after cutting, etching, and filling with white epoxy. When only nearly normal lines of response and a narrow energy window were used, the FWHM of the system changed to 2.00 mm. When the energy acceptance window was opened and all lines of response were accepted, the FWHM degraded to 2.75 mm. These results were obtained without distortion correction for the true geometric position of the event on the face of the photomultiplier tube. The completed system had a spatial resolution of 2.8-mm FWHM.


View this table:
[in this window]
[in a new window]

  		Effect of Surface Treatment on Spatial Resolution of Pixilated BGO Crystal Blocks Used in the PEM Instrument Detectors
 
System spatial resolution.—With profiles drawn through the same section of each image set, the peak-to-valley ratios for PEM instrument images of 1-mm-inner-diameter capillary tubes placed 10, 8, and 5 mm apart were calculated as 10.6, 2.5, and 1.5, respectively. Figure 2 is a comparison of the images obtained when five capillary tubes were placed 5 mm apart and imaged with the conventional PET scanner and then with the PEM instrument. The PET image data were collected during 10 minutes and reconstructed by means of standard filtered back-projection algorithms. The PEM instrument images were obtained after imaging for 2 minutes with four times lower activity concentration. The highest peak-to-valley ratios for the PEM instrument and the PET images were 2.5 and 1.2, respectively.



View larger version (77K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Comparison images of five 1-mm-inner-diameter capillary tubes filled with 18F and placed 5 mm apart. Left: Conventional FDG PET scan obtained with scanning time of 10 minutes. Right: PEM instrument image obtained with scanning time of 2 minutes with one-fourth of the activity has superior spatial resolution.

 
Contrast resolution.—True contrast ratios (hot spot to background) that ranged from 3.4:1.0 to 55:1 were studied for a 16-mm-diameter hot spot. A profile was drawn through each image, and image contrast was calculated for comparison with the values obtained with Equation (1). The thickness of the background T was kept constant at 70 mm. The results are summarized in Figure 3. The theoretic value calculated with Equation (1) was plotted as a straight line. The image contrast values determined experimentally follow the line closely, especially in the region of low values of true contrast such as those seen in clinical studies.



View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Graph depicts contrast on PEM instrument images for a 16-mm-diameter hot spot as a function of the true contrast, calculated as the ratio of the hot-spot activity to background activity. The theoretic values obtained with Equation (1) are drawn as a solid line. Background thickness, 70 mm; energy window, 350-650 keV.

 
We studied the effect of increasing the low-energy threshold, which rejects low-energy events, on the image contrast and overall efficiency by replaying the list mode data with different values of the low-energy threshold. As a consequence of increasing the low-energy threshold, the number of scattered events decreased, which resulted in improved image contrast (Fig 4). However, this improvement was achieved at the cost of system count efficiency as indicated in Figure 5 by the decrease in the fraction of the total number of counts used to produce a useful image.



View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Graph depicts effect of low-energy threshold on contrast on PEM instrument images. As the low-energy threshold gets higher, the number of scattered events accepted decreases, which results in better contrast.

 


View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Graph depicts effect of low-energy threshold on the relative number of detected counts used to form images. The rejection of low-energy events results in images that have better contrast but are relatively count starved.

 
System efficiency.—Figure 6 is a plot of the percentage of emitted annihilation photons detected by PEM instrument detectors as a function of the detector separation. The data points were fitted to a polynomial of the third degree. As the source-to-detector solid-angle increased, detection efficiency decreased at a rate faster than that predicted with the inverse square law. At a detector separation of 55 mm (face to face), with the source at the geometric center of the midplane, the PEM instrument detectors demonstrated an efficiency of 3% with a wide-open energy window.



View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6. Graph depicts absolute system efficiency as a function of detector separation, which was measured by placing a calibrated source at the center of the midplane between the detectors and calculating the ratio of the number of counts detected to the number of counts emitted.

 
Timing resolution.—The spectrum recorded from the multichannel analyzer was analyzed, and the FWHM of the timing resolution curves was determined. The timing resolution was calculated from the spectrum as 12-nsec FWHM. More than 75% of the counts fell within a window width of 12 nsec.

Patient Scanning
Figure 7 shows typical clinical images obtained with the PEM instrument. In this patient, the abnormality was located almost centrally, slightly above the midplane. The seven emission images in the upper row represent seven horizontal sections through the left breast obtained after 300 seconds of data acquisition. The section closest to the tumor site (section 4 [from left to right]) is the best-focused image of the tumor, which allows an estimate of the depth of the tumor within the breast (15–22 mm from the top of the breast). Figure 8 shows the coregistered composite PEM instrument and digitized x-ray mammographic image with the coregistration tool outlined in green. The location of increased glucose uptake corresponded to the location of the tumor. There was a fourfold increase in the relative number of counts detected at the tumor site compared with that in the rest of the breast. From the emission images, the size of the hot spot was estimated as 10 x 15 mm (horizontal x vertical dimensions). This was consistent with the clinical findings. Disease in this patient was subsequently diagnosed as intraductal and infiltrating ductal carcinoma with histologic grade of 3/3, measurements of 15 x 15 x 13 mm, and location in the laterosuperior aspect of the breast.



View larger version (48K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7. PEM instrument images of the breast in a patient with undiagnosed mass in the right breast. Top row: Seven sections through the right breast with a region of high glucose uptake. Bottom row: Emission images from the left (control) breast. Imaging time was 5 minutes, and injected activity was 75 MBq. At biopsy, the lesion was diagnosed as an invasive carcinoma not otherwise specified, or NOS, histologic grade 3 on the scale of Bloom and Richardson.

 


View larger version (200K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8. In the same patient as in Figure 7, best-focused PEM instrument image (full color) smoothed and scaled to match the size of the x-ray mammographic image, which is shown in gray scale, and automatically coregistered with it. Green lines correspond to the software tool, with size adjusted to match that of a similar wire tool on the mammogram. Agreement is good between the location of the abnormality on the mammogram and the region of focal uptake on the PEM instrument image.

 

   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The PEM instrument was found to be a high-spatial-resolution, high-efficiency system for coincidence detection. We tried to understand the problems of light collection from individual elements of the block. The aspect ratio of the crystal elements is much higher than that for a conventional PET detector; therefore, scintillation light must reflect off the crystal faces more times before it reaches the photomultiplier tube, which results in greater light loss. This degraded the energy resolution of the BGO block considerably (from 23% for the uncut block to 53% for the cut, etched, and filled block). The intrinsic spatial resolution of the system at 2.8-mm FWHM is better than that for other PET scanners used in human imaging because of smaller element size and higher sampling frequency. This is evident from the results of identical sets of experiments performed with the PET scanner and the PEM instrument (Fig 2).

Results in the contrast-resolution experiments show the close agreement of the detector contrast expected theoretically and the actual image contrast, which demonstrates that the scatter contribution to the image is minimal and can be effectively eliminated by means of appropriate choice of the low-energy threshold. With our PEM instrument system, a hot spot with a true contrast ratio of 6:1 compared with background was well resolved spatially.

The system efficiency (the fraction of total annihilation events detected with the PEM instrument detectors expressed as a percentage) varied from 4.7% (40-mm detector separation) to 0.7% (140-mm separation). The experimental value of the efficiency at 55-mm detector separation (3%) was a factor of more than five times lower than the theoretic prediction of 16% for a perfect system. Several factors contribute to the loss in efficiency. There was a loss of efficiency owing to the finite width of the coincidence timing window. Approximately 75% of events fell within the coincidence window of 12 nsec. The remaining events were rejected. This led to a further reduction in efficiency from 16% to 12%.

The single greatest source of loss of efficiency is dead time. Dead time for the PEM instrument had four distinct components: (a) the dead time associated with the decay time of the scintillation light (300 nsec for BGO) and that associated with the timing electronics for each detector unit; (b) the dead time associated with the coincidence detection system; (c) the finite time associated with the data interrogation between the analog-to-digital converter (model 14; Jorway Aurora, Westbury, NY) and computer (prior to the transfer of data between the converter and the computer), which is of the order of 0.3–0.4 µsec; and, most important, (d) the time associated with data readout. The converter in the present system cannot send an interrupt command when the memory buffers are nearly full. Instead, the computer polls the converter buffer every 0.1 second and initiates a data transfer if at least 1,024 events are in the buffer. Four milliseconds are required to transfer data to the computer. Data acquisition is disabled completely during the period when the converter is busy transferring information from its buffer to the computer memory. For a given detector separation, use of a higher input event rate would require more frequent data transfers, which would lead to lower efficiency. This is a shortcoming of the present data acquisition system that can be easily overcome by using two separate converters and alternately polling each, which would eliminate the dead time due to data transfers.

Results of the clinical study compared well with the histopathologic findings. The tumor location on the PEM instrument image correlated well with its position on the x-ray mammographic image. Wahl et al (4) report a median increase in FDG uptake of 8:1 in cancer cells compared with that in normal cells. In clinical application of the PEM instrument system, we believe that sites with high glucose uptake will be easily detected.

In summary, we designed and tested an instrument for PEM that makes use of pixilated BGO detectors and position-sensitive photomultiplier tubes to achieve high spatial resolution. In human subjects, the intrinsic spatial resolution of PEM instrument images is up to a factor of two better than that of conventional PET scans. Clinical PEM instrument images with good spatial resolution can be obtained with 75 MBq of FDG and maximum imaging time of 5 minutes. FDG PET studies on whole-body PET scanners for detection of breast cancer are typically performed with a dose at least five times as high and in as long as 1 hour (8). Our PEM instrument system performs reliably in a clinical setting and could potentially be a useful complementary imaging modality to x-ray mammography.


   Acknowledgments
 
The authors thank the following people for technical help: Patrick Sciascia and Dean Jolly, BSc, of the Medical Cyclotron Facility of the Montreal Neurological Institute, and Lisle Proulx of the Department of Nuclear Medicine and Bonnie Courte, Joanne Ceccarelli, and Linda Matueszswska of the Cedars Breast Clinic at the Royal Victoria Hospital.


   Footnotes
 
Abbreviations: BGO = bismuth germanate FDG = 2-[fluorine-18]fluoro-2-deoxy-D-glucose FWHM = full width at half maximum PEM = positron emission mammography

Author contributions: Guarantor of integrity of entire study, C.J.T.; study concepts, C.J.T., R.L., A.L., J.H.G.; study design, C.J.T., R.L., A.L.; definition of intellectual content, K.M., C.J.T.; literature research, K.M.; clinical studies, K.M., C.J.T., R.L., A.L., J.H.G.; experimental studies, K.M., M.A., C.J.T., A.M.B., J.L.R.; data acquisition and analysis, K.M., M.A., C.J.T., A.M.B., J.L.R.; manuscript preparation, K.M., M.A.; manuscript editing, K.M., M.A., C.J.T.; manuscript review, K.M., M.A., C.J.T., R.L., A.L.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Henson DE, Riess LA. Progress in early breast cancer detection. Cancer 1990; 65:2155-2158.[Medline]
  2. Schneidhauer K, Scharl A, Pietrzyk U, et al. Qualitative [18F]FDG positron emission tomography in primary breast cancer: clinical relevance and practicability. Eur J Nucl Med 1996; 23:618-623.[Medline]
  3. Adler DD, Wahl RL. New methods for imaging the breast: techniques, findings and potential. AJR Am J Roentgenol 1995; 164:19-30.[Abstract/Free Full Text]
  4. Wahl RL, Cody RL, Hutchins GD, Mudgett EE. Primary and metastatic breast carcinoma: initial evaluation with PET with the radiolabeled glucose analogue 2-[F-18]-fluoro-2-deoxy-D-glucose. Radiology 1991; 179:765-770.[Abstract/Free Full Text]
  5. Nieweg OE, Kim EE, Wong WH, et al. Positron emission tomography with fluorine-18-deoxyglucose in the detection and staging of breast cancer. Cancer 1993; 71:3920-3925.[Medline]
  6. Jansson T, Westlin JE, Ahlström H, Lilja A, Langstrom B, Bergh J. Positron emission tomography studies in patients with locally advanced and/or metastatic breast cancer: a method for early therapy evaluation?. J Clin Oncol 1995; 13:1470-1477.[Abstract]
  7. Nieweg OE, Wong WH, Singletary SE, Horotobagyi GN, Kim EE. Positron emission tomography of glucose metabolism in breast cancer: potential for tumour detection, staging and evaluation of chemotherapy. Ann NY Acad Sci 1994; 698:423-428.
  8. Thompson CJ, Murthy K, Picard Y, et al. Positron emission mammography (PEM): a promising technique for detecting breast cancer. IEEE Trans Nucl Sci 1995; 42:1012-1017.
  9. Bergman AM, Thompson CJ, Murthy K, et al. Technique to obtain positron emission mammography images in registration with x-ray mammograms. Med Phys 1998; 25:2119-2129.[Medline]
  10. Robar JL, Thompson CJ, Murthy K, Clancy R, Bergman AM. Construction and calibration of detectors for high resolution metabolic breast cancer imaging. Nucl Instruments Methods Phys Res-A 1997; 392:402-406.
  11. Clancy RL, Thompson CJ, Robar JL, Bergman AM. A simple technique to increase the linearity and field-of-view in position sensitive photomultiplier tubes. IEEE Trans Nucl Sci 1997; 44:494-498.
  12. Thompson CJ, Murthy K, Weinberg IN, Mako F. Feasibility study for positron emission mammography. Med Phys 1994; 21:529-538.[Medline]
  13. Adam LE, Zaers J, Ostertag H, et al. Performance evaluation of the whole-body PET scanner Ecat Exact HR plus In: IEEE Nuclear Science Symposium Conference Record. Vol 2. Piscataway, NJ: Institute of Electrical and Electronics Engineers, 1996; 1270-1274.
  14. Hoffman EJ, et al. Examination of the role of detection systems in quantitation and image quality in PET. IEEE Trans Nucl Sci 1986; 33:420-424.
  15. Evans AC, Thompson CJ, Marett S, Meyer E, Mazza M. Performance evaluation of the PC2048, a new 15-slice encoded crystal PET scanner for neurological studies. IEEE Trans Med Imaging 1991; 10:90-98.[Medline]
  16. Picard Y, Thompson CJ. Determination of the centroid of interaction of crystals in block detectors for PET. IEEE Trans Nucl Sci 1994; 41:1464-1468.
  17. Murthy K, Jolly D, Thompson CJ, Loutfi A, Lisbona R, Gagnon JH. Quantification in positron emission mammography (PEM) with planar detectors: contrast resolution measurements using a custom breast phantom and novel spherical hot-spots. IEEE Trans Nucl Sci Med Imaging 1998; (in press).



This article has been cited by other articles:


Home page
 RadioGraphicsHome page
E. L. Rosen, W. B. Eubank, and D. A. Mankoff
FDG PET, PET/CT, and Breast Cancer Imaging
RadioGraphics, October 1, 2007; 27(suppl_1): S215 - S229.
[Abstract] [Full Text] [PDF]

Home page
 JNMHome page
K. Murthy, M. Aznar, C. J. Thompson, A. Loutfi, R. Lisbona, and J. H. Gagnon
Results of Preliminary Clinical Trials of the Positron Emission Mammography System PEM-I: A Dedicated Breast Imaging System Producing Glucose Metabolic Images Using FDG
J. Nucl. Med., November 1, 2000; 41(11): 1851 - 1858.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Murthy, K.
Right arrow Articles by Gagnon, J. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Murthy, K.
Right arrow Articles by Gagnon, J. H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE