Personal Notes from the SNM99


Los Angeles, 5-10 June 1999

Robert E. Zimmerman

Monday, 16 August, 1999

" Notes" added 27 Oct 1999

It has taken much too long to get around to organizing and preparing these notes. I suspect they will not be terribly useful to others. In fact the notes will be less useful to many, myself included, because I took a laid back approach to the meeting in Los Angeles this year.

In contrast to many persons who visit Los Angles, I think it is a pleasant city to visit. I have pleasant memories of past SNM meeting in this city. For this reason I convinced my wife to come to LA and we, indeed, had a very nice time. But I definitely spent less time at SNM than in previous years. And I probably enjoyed it more.


Saturday is mostly devoted to committee meetings, registration of various kinds and learning the layout of the meeting and the city. I have great difficulty participating in a lot of these committees, especially when the meeting is in the Western US. I hate to be away from the office for such a long time. There is always the problem of: "do I come on Friday so I can make the Saturday AM meetings or do I catch the first flight on Saturday AM and hope I make the meetings I am supposed to be at." Well, I always opt for the Saturday travel and usually my presence at meetings is compromised.

This year I did manage to make MOST of the Academic Council ( meeting but I feel like a real loser in these meetings. I got on the Academic Council about 3-4 years ago because our residency training program had developed a device we called the Resident Passport to aid us in helping our residents complete training in practical nuts-and-bolts issues related to running a nuclear medicine clinic. The Academic Council was initially very interested but became wrapped up in questions such as "Should we be training new residents?" and "Where are the nuclear medicine jobs?" and "Is it morally right to train residents when there are no jobs". So I just hung in the background and am still in the background. Meanwhile our own program continues to train 7 (now 6) residents per year. Our Passport project continues to be refined and perhaps is useful and may yet find a use outside our own institution.

The Computer and Instrumentation Council (CIC) was held Saturday afternoon and since my wife did not arrive till late in the evening I killed some time at this board meeting. The Council is concerned about declining membership and would like to know how to attract back the numbers they enjoyed earlier in the decade. If you have any ideas on this let Jerry Wallis ( ), the Chair know about it. The CIC sponsored two continuing education sessions and one Sunday Categorical Seminar at this year's SNM meeting. There was considerable sentiment expressed for computer projection for at least some of the scientific sessions for future SNM meetings. Also there was strong interest in allowing supporting data to be submitted with the abstracts to allow more complete review of the quality of the submission. Mark Wittry will be the incoming vice-chair and the next year's chair and new board members are Beth Harkness, John Votaw and Tim Turkington. Steve Glick and Jim Collum are planning the mid-winter 2000 meeting. Pass comments to them if you have suggestions. The program is nearing completion and likely subjects include quantitative cardiac imaging and new instrumentation.

Saturday evening was devoted to preparing talks and meeting my wife at the LA airport. When you opt for the cheap flights on you get little choice with regard to flight times.



Sunday is holiday time at SNM for me. The one day I can play hooky from the meeting and enjoy the city. And we did. Sort of. First there was sleeping late on a Sunday morning followed by church and by the obligatory breakfast at The Pantry ( ). Best breakfast in town and I think it is better than that Zagat Survey indicates. My thoughts run more along this review: . Since it is very near the convention center it is easy to hit this at least once for each visit to LA. For the afternoon adventure we decided to visit Griffith Park . I had not been there since the mid 70s. Certainly one of the best spots in all of Los Angeles. Over 4000 acres and 53 miles of hiking trails. We had a wonderful walk for that lasted the entire afternoon The number of flowers and flowering shrubs was a real treat to this visitor from a colder realm where spring is a passing thought. Weather was terrific, warm, dry and sunny. The high point of the visit!

Returned from the outing to attend the American Board of Science in Nuclear Medicine ( board meeting. The ABSNM continues to provide an alternative certifying body for those physical scientists who want to enhance their credentials in nuclear medicine. The exam is now under constant revision with an ever-growing database of questions and is supervised by a psychometrician. In coming president, Gary Sayed, is expected to continue and improve on the work of his recent predecessors. A major accomplishment that is expected to greatly benefit the nuclear medicine community is the 3rd Edition of the Nuclear Medicine Syllabus of the ABSNM. Watch for it.



Now the real work began. The physics breakfasts, the scientific sessions the paper presentations, the back room conferencing, the lunch meetings, etc., all the while trying to convince my wife that the trip was worthwhile for her, too.

At this meeting I was aware of two companies conducting physics breakfasts, Siemens and Picker. These breakfasts are almost always a cut above the user meetings, which I long ago stopped attending. There was just not enough information exchanged to make them worthwhile for me. The physicists' breakfasts do improve the match between my interests and what the companies are willing to share but the 6:30 AM meeting time is a real difficult. So they had better be good. Another point in their favor is that the companies do not seem to think that they HAVE to have one every year. If there is nothing of import to say, why have one. I can buy that.


Siemens Physicists' Breakfast

Siemens was first up with a 6:30 AM starting time. I was actually supposed to be speaking first on the original agenda about some work we have been doing with coincidence cameras (See Paper No. 482 on Wednesday). But I was moved to last for some reason. So I could have slept a bit longer. But there were some interesting reports. CTI continues to ramp up LSO production and the PET camera using ECAM geometry is sounding very attractive. (See Paper No. 306, for example.) David Townsend reported on his project at U Pittsburgh with the combination PET/CT machine. Perfectly registered images! This was further reported on in Paper Nos. 597 (Instrumentation Track) and Paper No. 1134 (Oncology Track Poster) in. I think Siemens is seriously thinking of producing this machine or something like it. We should find out soon. Ed Ficaro reported on their experience developing a gender-neutral normal database and comparing it to their modified STEP database at U Michigan. (This was discussed further in Papers No. 15 and No. 361 in the Cardiovascular Track.) Eric Frey gave us encouragement that simultaneous Tc/Tl heart scans using advanced scatter correction techniques would eventually be ready to use. The most interesting presentation was from John Engdahl of Siemens R&D. He presented a progress report on the CZT semiconductor detector in ECAM form factor. Very encouraging planar and tomographic images were presented. Better than half a camera is working in the lab now. Impressive! Should be in a position to give Digirad ( a run for their money.

Of course the breakfast ran late and made us rush to the convention center to catch the opening ceremonies. My boss, Dr. S. J. Adelstein, was being presented with the Georg de Hevesy award after the Henry Wagner Lectureship was to be presented by Michael E. Phelps. Plenty of time, I thought, as the lecture will go for 30 minutes or so. But the schedule was rearranged and the De Hevesy Award was presented first and the lecture was presented last. Imagine doing that major reshuffling with a scientific session. At least, I thought, I can buy one of those tapes they were pushing and hear the De Hevesy award presentation that way. But I struck out there, too. Only heard the Phelps talk on tape. The award to Dr. Adelstein was not recorded. But it did happen, I was assured by eye witnesses.

So my wife and I escaped for a bus tour of the city to learn more about Los Angeles. The usual: Beverly Hills homes of the rich and famous, Rodeo Drive, Hollywood theaters, famous nite clubs, etc., but it was entertaining and no one took it very seriously. An interesting sidelight was that the most exclusive club we were shown on our bus tour turned out to be the very place we attended that night courtesy of Siemens.

Back to the scientific sessions in the afternoon. The SNM still has two books that they issue us: a program book and an abstract book. Since neither is complete the studious among us have to carry around two books to take notes in the scientific session, poster sessions and technical exhibit. One such book for all notes would be wonderful. Some day technology will arrive to make this happen.


No. 133 COMPARISONS OF PREDICTED LESION DETECTION PERFORMANCE FOR COMPTON CAMERAS AND PARALLEL-HOLE COLLIMATED CAMERAS C. Hua*, N. H. Clinthorne, S. J. Wilderman, T. J. Kragh, J. W. LeBlanc, W. L. Rogers, University of Michigan Medical Center, Ann Arbor, MI

These authors used a very powerful technique to predict performance of their Compton camera compared to a conventional camera. If I understood things right the presented paper had much less encouraging results than were presented in the abstract. During the presentation Hua mentioned that for 5 mm lesions the Compton camera was a factor of 8 worse than the Compton camera at 140 keV and only 3x better at 364 keV. This difference narrowed for larger lesion sizes. I was disappointed. I wonder if I heard this right? Notes: The authors have pointed out to me that the results presented were on a per photon basis. The result is that, for I-131 the Compton camera is slightly better at 140 keV and up to 20x better at 364 keV. A paper will be forthcoming.


No. 134 A COMPARISON OF LIMITED- AND COMPLETE-ANGLE ROTATING SLANT-HOLE OSEM IMAGE RECONSTRUCTIONS FOR MYOCADIAL SPECT D. E. Wessell*, E. C. Frey, B.M. W. Tsui, D. S. Lalush, The University of North Carolina at Chapel Hill, Chapel Hill, NC

Rotating slant hole tomography (RSH) has sort of gone the way of the 7 pinhole collimator. Its major limitation was incomplete angular sampling and resulting low resolution in the direction normal to the collimator surface. This group investigated the improvements that result when the collimator is rotated to 2-3 camera positions (usually 90 relative to each other). They found, in noise free simulations and phantom acquisitions, that the depth artifact can be removed. An OSEM reconstruction algorithm was developed to perform the reconstructions.


No. 135 TRANSMISSION SCANNING FOR GAMMA CAMERA PET AND SPECT IMAGING OF THE HEART N. J. Bird*, T. D. Fryer, R. W. Barber, Addenbrooke's Hospital, Cambridge, United Kingdom

These authors set out to design an attenuation correction system that would work for SPECT and coincidence imaging of the heart. They employed a Cs-137 point source and either a fan or offset fan collimator or axial shields for coincidence imaging. Useful field is about 17 cm. Scatter correction is done with a compensation PHA window. This sounds a lot like the Picker Beacon system which uses a Ba-133 source.


No. 136 DESIGN AND PERFORMANCE OF A SELF-SHIELDED CS-137 MULTIPLE-POINT-SOURCE TRANSMISSION SCANNING SYSTEM FOR ROTATING GAMMA CAMERA. T. G. Turkington*, C. M. Laymon, N. Wainer, J. P. Bouhnik, R. E. Coleman, Duke University Medical Center, Durham, NC; Elgems, Ltd., Haifa, Israel

This system was for measurement of attenuation maps for use in coincidence imaging only. Sixteen Cs-137 sources were used in an off-set fan geometry. Could be used simultaneous with coincidence imaging but required some crossover correction based on F-18 image. Hines of ADAC questioned the amount of crosstalk correction needed and implied it can be as much as 30%. ADAC uses a sequential technique.

No. 137 HIGH ENERGY CONE-BEAM TRANSMISSION IMAGING WITH PARALLEL COLLIMATORS G. L. Zeng*, G. T. Gullberg, D. Gagnon, C. H. Tung, P. E. Christian, University of Utah, Salt Lake City, UT; Picker International, Inc., Cleveland, OH

This system is meant for use in SPECT or coincidence imaging. The source is a point of Ba-133. The geometry is cone-beam because the Ba-133 gammas penetrate the low energy collimator foil in SPECT mode. Slats are used to collimate in coincidence mode. This is like the BEACON system of Picker. The authors were pleased with it's potential but more measurements of its physical parameters are needed.

No. 138 ULTRA-HIGH RESOLUTION SPECT IMAGES WITH URANIUM PINHOLE COLLIMATORS C. R. Tenney*, M. F. Smith, M. P. Tornai, R. J. Jaszczak, Duke University Medical Center, Durham, NC; National Institutes of Health, Bethesda, MD

The Duke SPECT group has been working with pinhole SPECT of the brain for several years. The brains have therapeutic does of I-131 in them. This year they present an improved collimator that employed a depleted uranium pinhole. And of course it works very well compared to a tungsten pinhole. More contrast, better resolution resulted.


No. 139 HIGH-RESOLUTION IN VIVO SPECT IMAGING OF I-131 BIODISTRIBUTIONS USING A ROTATING PARALLEL-HOLE COLLIMATOR D. E. Gonzalez Trotter*, J. E. Bowsher, C. D. Stone, R. J. Jaszczak, Duke University Medical Center, Durham, NC

The authors from Duke again attempt to devise a way to image their I-131 brain therapy patients. A new high resolution medium energy parallel hole collimator was designed that gave about 1/2 the resolution of the conventional ME collimator. The sensitivity was about 25 time less. The new collimator was rotated while the conventional ME collimator was used in the usual SPECT mode. But the patient had 40 mCi of I-131 on board. The images with the new rotating collimator were superior to those of the non rotating ME collimator.


No. 140 EFFECT OF THE COLLIMATOR CHOICE UPON LESION DETECTABILITY IN PLANAR SCINTIMAMMOGRAPHY (SMM) I. Buvat*, S. Laffont, J. Le Cloirec, H. Benali, P. Bourguet, A. Todd-Pokropek, R. Di Paola, U494 INSERM, Paris, France; Centre Eugene Marquis, Rennes, France; University College London, London, United Kingdom

The authors showed that an Ultra High Resolution Fan beam collimator was significantly superior to a HRFB, LEHR and LEUHR collimators when the task is to detect small lesions in high background. Conclusions were based on a rather thorough ROC study of breast phantom images.


Picker's Physicists' breakfast started at 6:30 AM. Luckily it was right across the street from my hotel. First up was Frank Valentino of the engineering department. He skimmed over the new features in the cameras. Here is a list:

Acquisition at 90░, 102░, 120░ and 180░ detector orientations.
Variable tangential configuration
Big detectors
"Irising" for good brain scans
Improved low attenuation pallet
Precision detector technology - all digital (finally) and fully field programmable
New optimized collimator cores (the old ones were not?)
Advanced technology options:
g PET3 - 3 detectors in coincidence (Alpha test?)
g PET3 - 2 detectors in coincidence
BEACON - Attenuation correction for PET and SPECT acquisitions
DOI - Correction for depth of interaction (When?)
Sales are great!

The meeting concluded with a looong talk by Daniel Gagnon, Chief Physicist and Manager of Advanced Technology and father of BEACON. BEACON uses 2- 10 mCi offset, moving Ba-133 sources with photons at 322, 353 and 383 keV. Shielding is much easier than with Cs-137. It readily penetrated LE collimators. Interleaved scans are used and add 2-4 minutes per scan. The angle of rotation has to be more than 180░ to avoid truncation. There is still a lot of work to do on optimizing protocols for scatter correction in both PET and SPECT mode. Core sight for PET is U Mass and for SPECT is VA Decatur.

The rewards for coming were a lightweight white paper on BEACON and a very nice laser pointer/pen. There was considerable information transmitted but also lots of fluff.


No. 303 MICROPET II - DESIGN OF A 1 MM RESOLUTION PET SCANNER FOR SMALL ANIMAL IMAGING S. R. Cherry*, Y. Shao, R. B. Slates, E. Wilcut, A. F. Chatziioannou, M. Dahlbom, University of California at Los Angeles School of Medicine, Los Angeles, CA

The UCLA group continues to amaze and thrill with their relentless pursuit of 1 mm animal PET machines. This talk was an update of the machine that uses optical fibers to readout the LSO detectors. The prototype detector performance is consistent with simulations that should result in 1.2 mm FWHM reconstructed spatial resolution.

No. 304 A COMPACT HIGH RESOLUTION PET DETECTOR USING AN APD ARRAY Y. Shao*, R. W. Silverman, R. Farrell, L. Cirignano, R. Grazioso, K. S. Shah, S. R. Cherry, University of California at Los Angeles School of Medicine, Los Angeles, CA; Radiation Monitoring Devices, Inc., Watertown, MA

This UCLA PET scanner, intended for breasts and small animals uses avalanche photodiodes instead of bulky PMTs. 14% energy resolution was achieved with LSO, time resolution of 3.5 ns. Depth of interaction information can also be obtained. They think feasibility is confirmed and a full sized system should be built.

No. 305 EFFECTS OF LU-176 BACKGROUND RADIATION ON SINGLES TRANSMISSION FOR LSO-BASED PET CAMERAS J. S. Huber*, W. W. Moses, W. F. Jones, C. C. Watson, Lawrence Berkeley National Laboratory, Berkeley, CA; CTI, Inc., Knoxville, TN

Lu-176 in LSO causes a significant background counting rate. This can contribute to the singles rate. These authors estimate that the effect on the transmission scan will cause a 60% increase in scan time to maintain statistics.

No. 306 HRRT A NEW HIGH RESOLUTION LSO-PET RESEARCH TOMOGRAPH M. Schmand*, M. E. Casey, K. Wienhard, L. Eriksson, W. F. Jones, M. Lenox, J. W. Young, K. Baker, S. D. Miller, J. H. Reed, W. D. Heiss, R. Nutt, CTI PET Systems, Inc., Knoxville, TN; Max-Planck-Institute, Cologne, Germany; CTI, Inc., Knoxville, TN

Finally we get to hear about a real LSO scanner. For the past 3-4 years we have been teased with simulation, prototype detectors, partial rings, animal scanners, etc. Now we can see some real results. This machine was installed very recently and has only begun scanning since May 1999, I think. This machine uses two layers of LSO with different decay times and can get DOI by pulse shape discrimination. The specs are staggering: 120K crystals in an octagon configuration surrounding the patient, 4 Mcps cps data storage rate, 10 ns timing window, NEC = 250 kcps @ 1 uCi/cc, 37% SF at 350 keV threshold. I am sure we will be hearing a lot more about this scanner.

No. 307 IMAGING PRIMARY TUMORS AND AXILLARY METASTASES WITH A DEDICATED BREAST PET CAMERA W.-H. Wong*, J. Uribe, H. Baghaei, H. Li, N. Zhang, J. Wang, University of Texas M D Anderson Cancer Center, Houston, TX

This scanner has a breast mode with breast FOV and a WB mode for staging. The gantry tilts 90░. 2-3 mm resolution in breast mode and 6-7 mm in WB mode. A phantom study was encouraging for early lesion detection. See also Poster 1233.

No. 308 PET CAMERA PERFORMANCE MEASUREMENTS: A COMPARISON BETWEEN THREE PET CAMERAS. L. E. Adam*, J. S. Karp, R. J. Smith, University of Pennsylvania, Philadelphia, PA

The draft NEMA and IEC protocols for PET measurements were used to determine parameters for HEAD PENN-PET, UGM QUEST and UGM CPET. The abstract contains all the numbers. Last year the group reported on image quality measurements for these PET machines. The comparisons are quite thorough and the methods used here should go a long way in making PET parameters standardized. Now we have to hope that all PET equipment manufacturers will adopt these methods for reporting performance parameters.


No. 456 REVIEW OF PHYSIOLOGICAL AND ANATOMICAL FACTORS IMPACTING COUNT UNIFORMITY IN ATTENUATION AND SCATTER CORRECTED TC-99M SESTAMIBI SPECT SLICES. P. H. Pretorius*, M. V. Narayanan, S. T. Dahlberg, J. A. Leppo, B. A. McSherry, M. A. King, University of Massachusetts Medical School, Worcester, MA

Many authors have reported lower apex counts and increased inferior wall count in attenuation corrected cardiac SPECT scans. The reasons have not been so clear, however. These authors studied 174 patient MIBI 3-head scans with FBP uncorrected reconstructions and with OSEM scatter and attenuation corrected reconstructions. They conclude that respiratory motion may be a major culprit as well as extra cardiac activity.

No. 457 DATA-DRIVEN RESPIRATORY GATING IN LIST MODE CARDIAC PET G. J. Klein*, B. W. Reutter, R. H. Huesman, Lawrence Berkeley National Laboratory, Berkeley, CA

This group investigated the center of mass of FDG heart scans obtained from rebinned list mode data as a method to reduce respiratory motion. The center of mass correlated well to a bellows signal in 3 patient studies.

No. 458 ECT ATTENUATION MAPS FROM X-RAY CT IMAGES H. R. Tang*, C. E. Schreck, B. H. Hasegawa, R. A. Hawkins, University of California at San Francisco, San Francisco, CA

This method of obtaining attenuation coefficients from CT includes the effect of iodinated contrast agents as well as bone and tissue. The maps derived in this way can be used for PET or SPECT corrections. Poster No. 820 describes the results in pigs and Paper No. 596 gives more details.

No. 459 EFFECT OF ACQUISITION TIMES ON SEGMENTATION ALGORITHMS FOR TRANSMISSION SCANS WITH GATED ROD SOURCES. T. K. Lewellen*, S. G. Kohlmyer, R. L. Harrison, V. Bettinardi, E. Pagini, P. S. Crandell, S. G. Ross, University of Washington Medical Center, Seattle, WA; Istituto H S Raffaele, Milano, Italy; General Electric Medical Systems, Waukesha, WI

The conclusion seems to be that transmission acquisition times of 3 minutes with segmentation are about equal in quality and noise to scans of 15 minutes. OSEM reconstruction is used. GE Advance is the machine. Bone may still present difficulties.

No. 460 ATTENUATION CORRECTION IN WHOLE BODY PET: ARE TRANSMISSION SCANS POSSIBLE IN LESS THAN ONE MINUTE? K. Bilger*, J. Kupferschläger, W. Müller-Schauenburg, M. Reimold, B. M. Dohmen, F. Nüsslin, R. Bares, Department of Nuclear Medicine, Tuebingen, Germany; Department of Medical Physics, Tuebingen, Germany

Also using the GE Advance these authors pushed the time scan time to under 1 minute. Perhaps they did better than Lewellen et. al., above, because their segmentation model allowed for 5 tissue types. Bone is a problem here, too. Bigger patients required up to 90s.

No. 461 ATTENUATION CORRECTION WITHOUT TRANSMISSION SCANNING FOR FDG LUNG STUDIES ON A COINCIDENCE CAMERA SYSTEM D. S. Lalush*, B.M. W. Tsui, C. Tocharoenchai, X. D. Zhao, The University of North Carolina at Chapel Hill, Chapel Hill, NC

These authors have attempted to reduce the transmission scan time to 0 minutes by using the consistency condition that the total counts in the attenuation corrected projections must vary minimally with projection angle. It was necessary to correct for scatter. Attenuation coefficient was within 20% of truth for a realistic phantom.


This group described their effort to find an efficient and accurate scatter compensation method. The method used an effective scatter source image for slices parallel to the collimator obtained by using a kernels based on the Klein-Nishina formula. These were then blurred with a small kernel related to the collimator to get multiple slices. They claimed efficiency and accuracy.

No. 477 COMPARISON OF DIAGNOSTIC PERFORMANCE IN BRAIN PET IMAGING USING 3D AND 2D ACQUISITION MODES R. J. Hagge*, T. G. Turkington, S. Borges-Neto, M. W. Hanson, R. E. Coleman, Duke University Medical Center, Durham, NC

This is a ROC study using real PET images with small (<=0.06 ml), low contrast (1.1 to 3:1) lesions placed in a Hoffman 3D phantom. Conditions compared were 2Dx20 min, 3Dx20 min and 3Dx6 min. There was no difference between the 20 min scans and the 6 min 3D scan was only slightly worse.


Some Tc-99m peak drift was noted after coincidence imaging. Measurable offset existed for up to 36 hr after the FDG insult.


NEC was obtained and compared with contrast of hot spherical lesions (5:1) in an elliptical phantom. Scatter fraction was obtained on low count rate acquisitions and used to obtain the randoms rate at high counts, thereby allowing the calculation of NEC. Sphere contrast correlated with NEC with max at about 0.8 -1.2 Mcps/detector singles rate.

No. 480 SIMULTANEOUS COINCIDENCE AND TC99M SPECT IMAGING USING A DUAL PURPOSE CAMERA WITH COLLIMATORS AND A CROSS-TALK CORRECTION ALGORITHM J. M. Baron*, I. G. Zubal, L. Daley, J. Seibyl, Tel Aviv Medical Center, Tel Aviv, Israel; Yale University Medical school, New Haven, CT; Veterans Administration Hospital, West Haven, CT

These authors concluded that it is feasible to perform simultaneous studies from their phantom experiment. They found that a correction for the F-18 downscatter to Tc-99m window was mandatory.

No. 481 CONTRAST AND SNR STUDIES ON THE ADAC MCD WITH ATTENUATION CORRECTION. V. Sossi*, L. Shao, J. Karp, H. Hines, University of British Columbia/Triumf, Vancouver, BC, Canada; ADAC Laboratories, Milpitas, CA; University of Pennsylvania, Philadelphia, PA; ADAC, Milpitas, CA

Authors used a protocol derived from draft update NEMA NU-2 which defines protocols for measuring image quality, among other things. At 8:1 contrast a sphere with 10 mm ID dia. Was just barely seen at scan time of ? min. (20 to 40 min, I am not sure). Activity at the edge of the field reduced contrast markedly. Realistic tests are important in characterizing these machines.

No. 482 COMPARISON OF COINCIDENCE IMAGING SYSTEMS USING MEASURED LESION CONTRAST AND SIGNAL TO NOISE RATIO (SNR) R. E. Zimmerman*, S. C. Moore, L. S. Graham, R. E. Malmin, Harvard Medical School, Boston, MA; University of California at Los Angeles School of Medicine, Los Angeles, CA; Siemens Medical Systems, Inc., Hoffman Estates, IL

Using spheres at 8:1 contrast the SNR (Detectability index) and measured image contrast were determined for ECAM+, MCD and PCD. The effect of attenuation correction, randoms correction, scatter correction and activity were determined. Differences among machines and imaging conditions could be determined quantitatively. Measured attenuation correction was good. Thicker crystals improved lesion detection and measured random correction improved the images.

No. 483 THE ESTIMABILITY OF ACTIVITY CONCENTRATIONS WTH COINCIDENCE IMAGING SYSTEMS S. C. Moore*, R. E. Zimmerman, J. A. Parker, S. P. Müller, Harvard Medical School, Boston, MA; Beth Israel Deaconess Medical Center, Boston, MA

How well can a gamma camera coincidence system estimate lesion and background activity? And what conditions affect the accuracy of this estimate? It was found that bias was under 12% at low activity but rapidly increased when singles rate went above 150 kcps, reaching 100% bias at 450 kcps. Out of field activity affected bias only 10%. These results were obtained using rebinned projection data only so are independent of reconstruction algorithm.



Tc-99m teboroxime tracer in the heart was simulated. Acquisitions with six scanning protocols using from 180 to 10 projections, all for 900 seconds was also simulated. OSMAP with a compartmental model-based prior performed with significantly less bias than OSEM followed by compartmental modeling. Bias was best for the shorter projection times and increased with projection time. Noise performance was also better with OSMAP.

No. 590 COMPARING FBP TO OSEM SPECT RECONSTRUCTION USING HUMAN OBSERVERS R. G. Wells*, M. A. King, P. H. Simkin, P. F. Judy, A. B. Brill, H. C. Gifford, R. Licho, P. H. Pretorius, P. Schneider, D. W. Seldin, University of Massachusetts Medical School, Worcester, MA; Brigham and Women's Hospital, Boston, MA; Lahey Clinic Medical Center, Burlington, MA

The group compared FBP w/wo attenuation correction vs. OSEM w/wo attenuation correction vs. OSEM with low noise correlation (not physically possible) using ROC techniques on low contrast lesions. This was all done with simulated data. OSEM with attenuation correction was best overall for the physically realizable reconstructions, but OSEM with low noise correlation was far better than the other reconstruction methods. Shame it cannot be used, or can it?


The author described a simulation experiment to prove that their own Smoothed OSEM reconstruction method was superior to FBP and standard OSEM. Tsui posed the question: Did they control for the number of OSEM iterations?


The authors compared FBP (PROMIS) vs. OSEM for parameter extraction from brain studies with C-11 alpha-methyl-tryptophan. They said that OSEM gave lower noise and better contrast with the same parameter values as PROMIS.

No. 596 A COMBINED CT/SPECT SYSTEM CAN IMPROVE THERAPEUTIC RADIONUCLIDE DOSIMETRY K. H. Wong*, A. J. Da Silva, H. R. Tang, M. C. Wu, B. H. Hasegawa, University of California at San Francisco, San Francisco, CA

This was a simulated study. CT was used to define the ROIs which were then projected onto the reconstructed unfiltered SPECT images. There was a dramatic improvement in quantitation noise and accuracy. See also Poster No. 820 and Paper No. 458.

No. 597 FUSION IMAGING FOR WHOLE-BODY ONCOLOGY WITH A COMBINED PET AND CT SCANNER D. W. Townsend*, T. Beyer, P. E. Kinahan, M. Charron, C. C. Meltzer, M. A. Dachille, J. Jerin, T. Brun, R. Roddy, R. Nutt, L. G. Byars, University of Pittsburgh, Pittsburgh, PA; CTI PET Systems, Inc., Knoxville, TN; Byars Consulting, Oak Ridge, TN

Dave updated us on his project that involves an ART on the same axis as a Somatom spiral CT scanner. Very promising. Over 50 scans have been done. This project was also discussed at the Siemens physicist's breakfast and in Poster 1134 and 1545. One of his images was selected by Dr. Wagner as Image of the Year

No. 598 A HIGH SPEED GAMMA CAMERA USING THE HIGH YIELD PILEUP-EVENT RECOVERY (HYPER) METHOD W.-H. Wong*, H. Li, J. Uribe, H. Baghaei, N. Zhang, J. Wang, University of Texas M D Anderson Cancer Center, Houston, TX

Dr. Wong updated us this year with performance data on a four PMT version of his very fast gamma camera circuitry. 2.2 Mcps was reached with good results. How do we compare this with existing commercial circuits?

No. 599 SOLID STATE ANGER CAMERA B. E. Patt*, J. S. Iwanczyk, L. R. MacDonald, C. R. Tull, Y. Shao, E. J. Hoffman, Photon Imaging, Inc., Northridge, CA; University of California at Los Angeles School of Medicine, Los Angeles, CA

Silicon drift detectors replace the PMTs in this camera. CCD technology allows for inexpensive readout electronics. These light detectors have higher quantum efficiency than PMTs permitting better energy resolution. Small size (0.5 to 1 cm2) may lead to better intrinsic resolution. 

No. 600 A MINIATURE BETA PARTICLE-SENSITIVE PROBE FOR ENDOSCOPY R. R. Raylman*, S. Majewski, R. Wojcik, V. Popov, West Virginia University, Morgantown, WV; Jefferson Lab, Newport News, VA

This 5 mm dia., 200 m m thick detector with a 2m long liquid light guide works but the authors feel that a background rejection technique is necessary before clinical trials can be performed. Stacked, coaxial and phoswich approaches are proposed.

No. 601 COINCIDENCE DETECTION OF TUMORS USING LINEAR DISCRIMINANTS WITH 111IN K. J. Matherson*, H. B. Barber, H. H. Barrett, J. M. Woolfenden, University of Arizona, Tucson, AZ

This group reported the results of a phantom experiment that justifies further work on a coincidence probe for in-vivo tumor detection. The phantom study used a NaI probe counting the two photopeaks and the sum peak. Only the sum peak produced desirable performance. Indicating that coincidence counting had great promise.

No. 602 AN INTRA-OPERATIVE COMPACT GAMMA IMAGER FOR RADIO-GUIDED CANCER SURGERY L. Menard*, Y. Charon, M. Ricard, M. Solal, P. Laniece, R. Mastrippolito, L. Pinot, L. Valentin, Institut de Physique Nucleaire, Orsay, France; Institut Gustave-Roussy, Villejuif, France

This imager uses CsI with position-sensitive silicon photodiode. Diameter is 24 mm. And a tungsten etched collimator with 1 mm hex holes and thickness of 3 cm is used. Spatial FWHM is 2 mm at 3 cm. A 6 cm dia. model is being built.

Henry Wagner, Jr. MD: Presentation of Annual Meeting Highlights

Dr. Wagner gave his 23rd SNM highlights lecture. You can read it in the August issue of the JNM. The image of the year was a fused head PET/CT image from the Pittsburgh group. The Research Image of the Year was by Dr. Simon Cherry and his UCLA group acquired with MicroPET showing the dopamine system in a rat brain.

Dr. Wagner is always so upbeat in his highlights lecture it is refreshing to listen. While the much of our profession is wondering where NM is headed and how it will continue to function in the face of declining reimbursements he paints a bright picture for the future .


No. 820 ABSOLUTE IN VIVO QUANTITATION OF MYOCARDIAL ACTIVITY USING A COMBINED X-RAY CT AND SPECT SYSTEM A. J. Da Silva*, H. R. Tang, K. H. Wong, M. C. Wu, M. W. Dae, B. H. Hasegawa, University of California at San Francisco, San Francisco, CA

This pig study confirmed the simulated study in Paper No. 596 and the method described in Poster No. 458. A GE XRT SPECT camera and GE 9800 CT were used. The CT derived ROIs are use to correct for partial volume effects and to define the ROIs for SPECT counting.

No. 1134 WHOLE-BODY FDG PET AND CT IMAGING OF MALIGNANCIES USING A COMBINED PET/CT SCANNER M. Charron*, T. Beyer, P. E. Kinahan, C. C. Meltzer, M. A. Dachille, D. W. Townsend, University of Pittsburgh PET Facility, Pittsburgh, PA

This contains details of 30 tumor patients studied on the new machine. Breathing caused problems with location of lung tumors.

No. 1223 NOVEL DESIGN OF A MODULAR DETECTOR FOR THE MAXPET BREAST CANCER IMAGING SYSTEM N. K. Doshi*, S. R. Cherry, R. W. Silverman, University of California at Los Angeles School of Medicine, Los Angeles, CA

The best way to get the light out of the LSO was with a tapered light fiber bundle.

No. 1224 A PHYSICAL EVALUTION OF THE LARGE AREA POSITRON CAMERA (PETTRA) R. J. Ott*, D. M. Duxbury, M. A. Flower, K. Erlandsson, A. J. Reader, J. E. Bateman, R. Stephenson, E. J. Spill, Institute of Cancer Research, Sutton, Surrey, United Kingdom; Rutherford-Appleton Laboratories, Chilton, Oxon, United Kingdom

A pair of barium fluoride crystals are read out with a MWPC filed with TAME. The sensitivity and time resolution indicates that it performs better than a gamma camera in coincidence mode. New, faster electronics are being built. They have plans to use it on patients.

No. 1225 PERFORMANCE CHARACTERISTICS OF A COMPACT, QUANTIZED GAMMA CAMERA N. Hartsough*, B. Pi, J. Gormley, R. Conwell, W. Ashburn, Digirad Corporation, San Diego, CA

The 4096 element CZT detector is described. Specs are not particularly exciting. The most attractive feature is the low amount of dead space at the edges. Notes: I missed a major point here. This paper described a pixelated CsI/photodiode camera not a CZT camera.

No. 1226 DEVELOPMENT OF A SMALL GAMMA CAMERA USING PARALLEL-HOLE AND DIVERGING-HOLE COLLIMATORS (PHC AND DHC) FOR BREAST IMAGING J. H. Kim*, Y. Choi, J. Y. Kim, M. Y. Lee, K. C. Im, Y. S. Choe, S. E. Kim, K. H. Lee, K. S. Joo, B. T. Kim, Samsung Medical Center, Seoul, South Korea; Myongji University, Yongin, South Korea

Uses as PSPMT. 3.1 mm intrinsic resolution shown. Images demonstrate a need for linearity correction.

No. 1227 COUNTING PROPERTIES OF GAMMA PROBES USED IN THE PERIOPERATIVE MANAGEMENT OF MALIGNANT MELANOMA AND BREAST CANCER V. Ivancevic, M. Voth*, S. Fischer, D. L. Munz, Clinic for Nuclear Medicine, Charite, Berlin, Germany; University Hospital Charite, Berlin, Germany

Three systems were compared: Tecprobe 2000 from STRATEC in Germany, C-Trak from CareWise in USA and CSX-OP from CRYSTAL in Germany. Various combinations of collimators were tested. The authors liked the German probes the best. No mention of pulse height analysis was made. This is an important parameter to reduce scatter counts.


Again no mention was made of energy resolution in comparing of these probes. Four systems were compared.


No. 1229 STARBRITE: OPTIMIZATION OF CRYSTAL THICKNESS TO LSF FOR NAI (TL) M. A. Lopez*, Bicron, St. Gobain Industrial Ceramics, Newbury, OH

This poster was missing. What the heck is StarBrite?

No. 1230 EVALUATION OF A 3RD GENERATION FAN-BEAM BONE DENSITOMETER THAT UTILIZES A CADMIUM ZINC TELLURIDE DETECTOR M. Wilson*, R. Hovey-Andersen, B. Herrington, University of Wisconsin/Madison, Madison, WI; LUNAR Corp., Madison, W

CZT has found its way into bone mineral scanners.

No. 1231 SIMULTANEOUS EMISSION AND TRANSMISSION WB-PET SCANNING USING SHORT ROD SOURCES. M. Dahlbom*, University of California at Los Angeles School of Medicine, Los Angeles, CA

The rods were extended to cover only about 4 cm of the FOV. Continuous bed scanning was used to cover the area of interest giving continuous axial sampling. Better Tx resulted. The EM were suffered some drop in NEC.

No. 1232 SPIRAL WHOLE-BODY PET - IMPLEMENTATION AND INITIAL RESULTS M. A. Flower*, K. Erlandsson, P. Collins, A. J. Reader, J. R. Symonds-Tayler, L. L. White, R. J. Ott, Institute of Cancer Research, Sutton, Surrey, United Kingdom

The authors claim that spiral PET is superior to step-and-shoot couch motion. They will apply this to PETTRA.

No. 1236 IMPLEMENTATION OF AN ANALYTIC SCATTER CORRECTION FOR THE GE ADVANCE PET SYSTEM S. G. Kohlmyer*, T. K. Lewellen, R. D. Badawi, R. L. Harrison, J. M. Ollinger, C. W. Stearns, University of Washington, Seattle, WA; Washington University, St. Louis, MO; General Electric Medical Systems, Waukesha, WI

The GE Advance needs a better scatter correction for body imaging. This is a method that seems to work on phantoms.


Poster missing in action.

No. 1241 REGIONAL FDG UPTAKE, ATTENUATION, AND GEOMETRIES MEASUREMENTS FOR WHOLE BODY PHANTOM DESIGN. T. G. Turkington*, N. E. Williams, S. M. Hamblen, R. E. Coleman, Duke University Medical Center, Durham, NC; Duke University, Durham, NC

This would be a good place to start if you were going to build a WB phantom for PET comparisons. Lots of data to get the activity ratios right.

No. 1242 PERFORMANCE CHARACTERISTICS OF A HEAD IMMOBILIZATION DEVICE FOR PET IMAGING O. Mawlawi*, R. Weiss, A. K. Shinn, J. Pidcock, M. Slifstein, M. Laruelle, New York State Psychiatric Institute and Columbia University, New York, NY; New York State Psychiatric Institute, New York, NY

The device is made with a 2 component mixture yielding a polyurethane mold (Solve Medical, Tampa, FL). It was found superior to the plastic mask.

No. 1243 EVALUATION OF A COINCIDENCE GAMMA CAMERA SYSTEM FOR SINGLE PHOTON SCINTIGRAPHY AND POSITRON EMISSION TOMOGRAPHY S. Dresel*, K. Brinkbarumer, P. Escher, R. Kantlehner, A. Enterrottacher, S. Ziegler, K. Hahn, University of Munich, Munich, Germany; Technical University of Munich, Munich, Germany

A Picker 3/8" system was compared to a 3/4" for loss of resolution. Findings: 3.8 mm FWHM goes to 4.7 mm. UHR collimators made up for it when system resolution was measured.

No. 1244 PERFORMANCE CHARACTERISTICS MEASUREMENTS OF A SPECT GAMMA CAMERA SYSTEM EQUIPPED WITH A PET OPTION A. Gagnon*, G. Bavaria, V. Rappoport, D. D. Burckhardt, R. Taillefer, Centre Hospitalier Universitaire de Montréal, Campus Hôtel-Dieu, Montréal, PQ, Canada; SMV America, Twinsburg, OH

DST-XL from Sophy/SMV uses 13 mm crystals. Coincidence option is called VCR. Uses Fourier rebinning with FBP reconstruction.

No. 1253 TRUNCATION EFFECTS IN DUAL-PLANAR CIRCULAR-ORBIT CONE-BEAM SPECT D. S. Lalush*, The University of North Carolina at Chapel Hill, Chapel Hill, NC

Found to yield better images than LEHR collimation. All areas of the brain were seen by at least one head. A simulation study.

No. 1254 SCINTIMAMMOGRAPHY VERSUS SPECT MAMMOGRAPHY G. L. Zeng*, G. T. Gullberg, University of Utah, Salt Lake City, UT

The authors compared planar vs. cone-beam SPECT of the breast using a phantom study on a 3-head camera. SPECT had better lesion detection.

No. 1257 FULL FOV HOLE ANGULATION ERROR MAPPING OF PARALLEL HOLE COLLIMATORS USING A PRETILTED COLLIMATED FLOOD SOURCE A. J. Arends, M.J. A. de Voigt, A. van Lingen*, Medical Physics Department, Catharina Regional Hospital, Eindhoven, The Netherlands; Eindhoven University of Technology, Accelerator Laboratory, Eindhoven, The Netherlands; Department of Nuclear Medicine, Free University Hospital, Amsterdam, The Netherlands

The trick employed by the authors is to use a second collimator that is error free to collimate the flood source. Now how do I get that error free collimator? Suggestion: Use a cast collimator. Or better yet throw out the foil collimator and use the cast collimator. Handout available: ( Notes: The authors have informed me that, in fact, cast collimators are not necessarily superior with regard to angulation errors. Furthermore, there are no error free collimators, so this method works to identify the region of the collimator that may be in error. Further test would be necessary to quantify the error.

No. 1258 DEVELOPMENT OF A SMALL ANIMAL PET IMAGING DEVICE WITH RESOLUTION APPROACHING 1MM J. A. Correia*, C. A. Burnham, D. Kaufman, A. J. Fischman, Massachusetts General Hospital, Boston, MA

LSO single ring system with 360 crystals, 12 crystals/block and 2 PMTs per block. 1.2 mm resolution at center to 1.65 mm at edge. System is in use. There are some problems with light collection at the ends of a block of 12 crystals.

No. 1259 THE EVALUATION AND CALIBRATION OF FAN-BEAM COLLIMATORS J. L. Mahowald, P. D. Robins, M. K. O'Connor*, Mayo Clinic, Rochester, MN

Published focal length is not the real focal length leading to unknown pixel size. This causes problems in image registration and image quality. The authors describe how to measure the focal length using Co-57 markers and tomographic images at 3 radii.

No. 1260 EFFECT AND CORRECTION OF DETECTOR SAG IN SPECT SYSTEMS J. Peter*, D. R. Gilland, J. E. Bowsher, M. P. Tornai, R. J. Jaszczak, Duke University Medical Center, Durham, NC

If your detectors sag you can learn how to correct for it with help from these fellows. But I need some tips on how to determine if I have detector sag.

No. 1261 METHODS TO CORRECT CAMERA SYSTEM SENSITIVITY FOR COLLIMATED AND PENETRATING RADIATION. D. Lange*, U. B. Noelpp, E. Klose, U. Haberkorn, Department of Nuclear Medicine, University of Heidelberg, Heidelberg, Germany; Department of Nuclear Medicine, University of Berne, Insel Hospital, Berne, Switzerland; Clinic of Nuclear Medicine, Clinical Center, Erfurt, Germany; Department of Nuclear Medicine, Heidelberg, Germany

Two methods for measuring penetration effects are given. Full details in the handout: ( )

No. 1263 CORE LABORATORY SPECT QC VALIDATION PROGRAM FOR MULTI-CENTER CLINICAL TRIAL S. M. Dyer*, J. Clulow, T. L. Frank, L. C. Becker, Institute for Radiological Image Sciences, Inc., Frederick, MD; The Johns Hopkins University, Baltimore, MD

A heart phantom (SpecPhan QC phantom, The Phantom Laboratory, Salem NY) is being used with a standardized acquisition protocol to validate cameras for inclusion in a multi-center trial. Resolution, scatter, size and uniformity is indicated by the phantom, claim the authors.

No. 1268 COULD STATISTICAL PARAMETRIC MAPPING (SPM) BECOME THE GOLD STANDARD FOR THE EVALUATION OF INTRA-SUBJECT CHANGES ON SEQUENTIAL BRAIN SPECT? D. G. Pavel*, P. Saxena, J. C. Quintana, V. V. Kumar, R. Chuang, K. Q. Lin, L. L. Sadler, University of Ilinois Medical Center, Chicago, IL; Catholic University, Santiago, Chile


No. 1281 TUMOR LOCALIZATION AND IMAGE REGISTRATION OF 18-FDG SPECT SCANS WITH CT SCANS C. L. Aitken*, G. McGuinness, F. Siddiqui, A. Ton, E. L. Kramer, G. Q. Maguire, M. E. Noz, New York University, New York, NY

Registration can be a problem because anatomic markers are extremely limited. Perhaps fiducial markers need to be added. In spite of the difficulties the authors had success with their program, qsh.

No. 1286 HOTTEST PIXEL ANALYSIS: USEFUL VALUE OR STATISTICAL ARTIFACT? P. Duarte*, R. Hustinx, O. Couturier, R. J. Smith, A. Alavi, University of Pennsylvania, Philadelphia, PA

Must use them critically. Email for reprint:

No. 1295 EVALUATION OF COSEM, A NEW ITERATIVE RECONSTRUCTION ALGORITHM FOR COINCIDENCE IMAGING J. A. Patton*, E. Shai, M. Wilk, G. Harden, M. P. Sandler, Vanderbilt University Medical Center, Nashville, TN; Elgems, Ltd., Haifa, Israel; General Electric Medical Systems, Milwaukee, WI

It is clearly better and we have been hearing about it for a while. When will it be released to the field?


The apriori information must be matched to the data from a specific patient. They found a way to do this.

No. 1298 TECHNICAL ASPECTS OF IMAGING RUBIDIUM-82 ON COINCIDENCE-DETECTION GAMMA CAMERAS D. J. Kadrmas*, P. E. Christian, E.V. R. DiBella, W. D. Adams, University of Utah, Salt Lake City, UT

Does not work well because of low sensitivity system and short half life of Rb-82 (75 s).

No. 1300 DOWNSCATTERED COMPONENT COMPENSATION METHOD DUE TO HIGH ENERGY PHOTON N. Motomura*, T. Ichihara, T. Takayama, S. Aoki, H. Kubo, K. Takeda, Toshiba, Tochigi, Japan; Suzuka Central Hospital, Mie, Japan; Mie University, Mie, Japan

Use a single window above the I-123 photopeak to subtract off the scattered radiation. Improved contrast in the heart by 12%. Not a very complete study as they should report noise and bias changes.


No. 1303 DIAGNOSTIC VALUE OF ATTENUATION CORRECTED (AC) VS. NON-ATTENUATED CORRECTED (NON-AC) IMAGES IN DUAL DETECTOR COINCIDENCE IMAGING (DDCI) L. Kostakoglu*, K. Goldner, G. Shevlak, N. Altorki, S. Vallabhajosula, S. J. Goldsmith, New York Presbyterian Hospital, Cornell Medical Center, New York, NY

Very small positive effect found in 29 patients.


"ůmay provide a real breakthrough in SPECT imaging of small objects." I hope so but we have been disappointed so often with coded apertures.


The authors found that the ability to see smaller lesion was not helped but image quality and lesion detectability was improved. May prove useful for following tumors.

Technical Exhibits

I was able to spend very little time in the technical exhibits this year.

I did note that the SNM was selling a CD-ROM with all the meetings abstracts again this year. It is better than that available in past years. Different vendor prepared it, obviously. There are a number of errors when one is searching for authors, however.

I am not sure it is worth the $50 unless you really want to know details about what went on. The AAPM gives things like this free to its members. Why not SNM?

Bob Z

Robert E. Zimmerman

Medical Physicist

Joint Program in Nuclear Medicine

Harvard Medical School

Brigham & Women's Hospital

Department of Radiology

75 Francis St.

Boston MA 02115

Tel: 617 732 7196

FAX: 617 732 6336