Notes from SNM Saint Louis Meeting, June 2000

Robert E. Zimmerman

Amended 28 July with GE photo.

Saint Louis is a nice place to have a meeting, if you can adjust to the fact that this is not quite a big league city. The dearth of hotels in the downtown area is very surprising. Always reluctant to use the high priced convention hotels my search for a suitable place for my wife and I to stay was turning up only hotels in very undesirable locations until I thought: bed and breakfast! A little Internet searching turned up three promising candidates. We finally settled on one a short walk from Union Station and the train/subway into downtown Saint Louis. That's right, Saint Louis now has a subway and it is clean, reliable and pleasurable to take. And it easily connects to Forest Park - the Museum Center of Saint Louis. The weather in Saint Louis is known to be very warm and humid in June but this year we were blessed with extremely pleasant weather making the daily walk from Lafayette Square (the location of our bed and breakfast) and Union Station very pleasant.

The bed and breakfast, Lehmann House, (http://www.bnbinns.com/lehmannhouse.inn) turned out even better than expected as the Lafayette Square (http://explorestlouis.com/gen09g.html) association was holding their annual house and garden weekend. Perfect for my wife to enjoy till I could get free of some meetings.

It was hard for me to appreciate why the meeting now starts on a Sunday, moving the Seminars to Saturday and committee meetings to Friday. I suppose someone liked it but I have not met that person, yet.

The first order of my business was to sniff around at the Friday evening Board of Directors meeting to see what was going to happen regarding a controversial "strategic alliance" that the SNM and GE announced in May. (Press Release). Apparently this was not important enough to make it onto the Board of Directors agenda. Did you know that the Board meets in private at a posh dinner meeting?

So, I figure, all the action must be happening at the House of Delegates meeting on Saturday. Better be there. Friends who are members of the House told me that there had been no items in their pre-meeting packet indicating this alliance business was to be discussed. Sure enough it was NOT on the agenda. But in the new business section of the meeting Michael O'Conner put it on the floor and a 45 minute discussion ensued. There was general agreement that the Board had erred in making the alliance with GE and GE had exploited the alliance. There was less agreement on what to do about it. The root of the problem seems to lie in a vote taken at last years House of Delegates meeting wherein the House instructed the Board to "make alliances". The House has now decided that is should formulate guidelines or make a "how to make alliances" kit for the Board to follow. The House also asked the Board to revoke the existing alliance with GE till the alliance kit could be completed. This could happen as early as the mid-winter meeting. It would behoove all chapters and anyone else who worries about the SNM selling out to big, powerful dominate companies to keep an eye out on this process.

Here is a (bad) photo from the GE booth. We can see how GE will exploit any special relationship they think that they might enjoy with SNM.

All this politicking was a distraction from the meeting and more importantly, on Friday and Saturday, a distraction from the city of Saint Louis. We did get to partake of the Lafayette Park festival and Forest Park: the Missouri History Museum (nice exhibit on Imperial Russia and its dealings with the US), The Jewel Box (a small flower conservatory) and the Saint Louis Zoo (certainly one of the better US zoos).

Sunday was the start of the scientific sessions and the equipment exhibition, the real meat of the SNM for most people.

The World Federation of Nuclear Medicine and Biology meeting will be held in Chile jointly with ALASBIMN Sept 29-Oct 4 2002 in Santiago. It promises to be a great meeting. Watch for more info. This group met in St. Louis to coordinate with the international group that is assisting with the meeting.

Sunday June 4

My notes on the scientific sessions will be significantly shortened from previous write-ups for two reasons: I lost my notes about the 3rd day of the meeting and my wife sort of pulled me away from a number of sessions I might have otherwise attended. So I am mostly going from pure memory, a very dangerous thing for me.

The opening plenary session was memorable mostly because two basic scientists were honored. Mathew Thakur received the Georg de Hevesy award for his many contributions to radiochemistry and radiopharmaceutical developments. Ronald Jaszczak was awarded the Paul C. Abersold award for contributions to basic science applications to nuclear medicine.

Incidentally, the SNM included a FREE CD-ROM with searchable copies of all abstracts presented at the meeting in our registration packet. I like it!

Session 10: Instrumentation & Data Analysis Track

Computer & Instrumentation Young Investigator Symposium

64

PERFORMANCE CHARACTERISTICS AND EVALUATION OF THE TIERPET, A HIGH RESOLUTION ANIMAL PET SYSTEM.

S. Weber*, A. Bauer, H. R. Herzog, F. Kehren, K. Ziemons, H. H. Coenen, K. Zilles, H. Halling, Forschungszentrum Jülich, Jülich, Germany.

There was certainly a lot of interest at this meeting in imaging of small animals. The The Jülich TierPET was a good example of this. The group reported on NEMA-like measurements on their YAP crystal scanner. Resolution of 2-2.3 mm with scatter fractions of 12 to 15.4% were reported. (http://ime-web.ime.kfa-juelich.de/ime_www/pet/tierPET/engl.html)

 65

INPUT FUNCTION DERIVATION FROM BRAIN H215O PET IMAGE USING INDEPENDENT COMPONENT ANALYSIS.

 J. Y. Ahn*, J. S. Lee, S. K. Kim, K. W. Kang, D. S. Lee, J.-K. Chung, S. A. Shin, M. C. Lee, College of Medicine, Seoul National University, Seoul, South Korea; College of Natural Science, Ewha Womans University, Seoul, South Korea.

I was pretty lost on this paper not being at all familiar with ICA - Independent Component Analysis. The object is to get the input function on H215O brain perfusion studies. They tested it on 5 dogs and feel that it is better than factor analysis for this purpose.

66

ESTIMATING THE UNCERTAINTY IN THE LOCATION OF A LOCAL MAXIMUM IN A PET IMAGE.

L. D. Nickerson*, C. C. Martin, J. L. Lancaster, J. H. Gao, P. T. Fox, University of Texas Health Science Center, San Antonio, TX.

Based on the derivation of Huesman for ROI variance calculations and using only sinogram data the method presented here allows the variance of the maximum pixel location to be calculated. Agreement with directly measured variance on pixel location was excellent.

67

IMPROVED QUANTITATIVE IMAGING FOR 111In-PROSTASCINT USING CT/SPECT AND DUAL-ENERGY RECONSTRUCTION.

K. H. Wong*, H. R. Tang, A. J. Da Silva, M. C. Wu, K. Iwata, B. H. Hasegawa, University of California at San Francisco, San Francisco, CA.

11% accuracy was achieved in a phantom. A single "effective attenuation map" was within 10% of the dual map method and much faster to compute.

68

AN INNOVATIVE HIGH EFFICIENCY AND HIGH RESOLUTION PROBE FOR PROSTATE IMAGING.

L. Zhang*, N. H. Clinthorne, S. J. Wilderman, C. Hua, T. J. Kragh, W. L. Rogers, University of Michigan, Ann Arbor, MI.

This was a simulated study based on Compton camera technology. One part of the camera is in the form of a transrectal probe and used Si detector 1x4x1 cm3 with 1 mm crystals. The other detector(s) is a dual 40x40x2 cm3 planar detectors with. Phantom results were encouraging over 140-511 keV. Resolution was about 2.5 mm.

69

REGIONAL LEFT VENTRICULAR FUNCTION AND PERFUSION FROM GATED SPECT 201Tl SCANS: A 4D MODEL USING SPHERICAL HARMONICS.

P. M. Mansour*, M. F. Smith, V. Dilsizian, S. L. Bacharach, The National Institutes of Health, Bethesda, MD.

A simulation study. Authors attempted to draw as much regional wall thickening, perfusion and maximum regional width as possible from noisy, low resolution data. Their 4D model reduces variation in values compared to 2D model without hurting accuracy.

70

LESION DETECTABILITY OF MAP RECONSTRUCTION USING COMPUTER OBSERVER: A THEORETICAL STUDY.

 J. Qi*, R. H. Huesman, Lawrence Berkeley National Laboratory, Berkeley, CA.

Theoretical results for PET systems indicate that detectability is better for MAP compared to FBP because of noise modeling. Lesion detectability in PET also profits from attenuation correction.

Session 11: Instrumentation & Data Analysis Track

Small Animal Imaging

Session 12: Instrumentation & Data Analysis Track

Instrumentation in Emission Tomography

This promised to be a very interesting sessions but I missed the whole afternoon because:

The American Board of Science in Nuclear Medicine (http://www.absnm.org/)held its meeting in direct competition with the scientific program on Sunday afternoon. One of the many indications that not everybody had adjusted to the change in the starting day for the meeting. The ABSNM offers certification to physicists, chemists and other scientists in nuclear medicine. I have been serving on the board for two terms and have now rotated off, so there will be a new representative from the Computer and Instrumentation Council. One less chore at next SNM.

Monday June 5

The Siemens Physicist Breakfast was held from 6:30 to 8:00 AM. It was pretty good as these things go. Vilim Simcic told us about Duet, the next generation of coincidence camera based on the e.cam. It has 1" thick Bicron StarBrite crystals, higher counting rates and better performance than cameras with thinner detectors. To be released soon at a store near you. Mark Groch showed a few measurements he was able to take at Northwestern before the meeting. Conclusion was that it is a pretty good coincidence camera.

The always entertaining, ever optimistic Ron Nutt convinced us that the best use of the limited but ever increasing supply of LSO was to make high-end PET scanners. Gone is the idea of completing and bringing to market a LSO/NaI scanner for hybrid PET and SPECT applications. And YSO does not rate a mention.

The syngo computer platform that cuts across all Siemens medical imaging platforms was expounded on by some suits from Germany. e.soft is the nuclear medicine contribution to this corporate effort. Platform is NT all the way. Cardiac software is the first clinical e.soft application with more to follow. I wish it was UNIX or even Linux.

Dennis Nelson described the software package he had developed for image fusion MIM and MIM2. Definitely worth a try. See the Website: http://www.nuclear.uhrad.com/software.htm

The Academic Council meeting took up my lunch hour. There is still great concern about the quality and number of nuclear medicine residents coming into programs. The number of training programs is still dropping. Mike Graham is starting an organization for Nuclear Medicine Program directors. The Academic Council has a web page now: http://www.snm.org/about/new_councils_1.html. Note the links to our Harvard Passport project. Comments welcome.

There was a PACS presentation in the continuing ed program. You can catch the essence at: http://www.med.harvard.edu/JPNM/Lectures/SNM00/PACS.html.

The President's Reception was held at the Missouri Athletic Club and was very nice. Those Missouri athletes have a great place.

Tuesday June 6

ADAC Physicist breakfast was 7-8 AM but ran over. It was a very exciting presentation, especially the one by Roger Lecomte who described his recent work with LSO and GSO. GSO has certainly captured ADAC's thoughts recently and they are building prototypes (with Penn). Lecomte's presentation included a fascinating discussion of the light centers in GSO and the various times to produce the scintillation and how dependent it is on material effects You can find more at: http://www.usherb.ca/bleu/anglais/industrial/biotechnology/lecomte.html

Joel Karp from Penn told us about some of the simulation results and prototype construction. It certainly looks promising. Hard to get a sense of the time scale to start production but it looks like clinical work for the masses will be 2-3 years away.

Wednesday June 7

Missed the Wagner Highlights Lecture but last time I checked it was available by streaming video from the SNM web site: www.snm.org. Have not looked at it but if it is like most streaming video I'll be hard put to watch it all.

POSTERS

802

OPTIMIZATION OF CLINICAL PET-FDG ACQUISITIONS ON THE ADAC C-PET SCANNER: A STUDY BASED ON A WHOLE-BODY PHANTOM AND PATIENTS DATA.

O. de Dreuille*, C. Lartizien, H. Foehrenbach, P. Maszelin, D. Brasse, J. F. Gaillard, HIA Val de Grace, Paris, France; Service Hospitalier Frederic Joliot-Commisariat A L'Energie Atomique (CEA), Orsay, France.

Authors found that max NEC was 12.8 kcps in a realistic phantom. Injected should be in the range 1.4-3.4 MBq/kg and images taken at 90 min.

803

OPTIMIZING GEOMETRY FOR HYBRID PET: DUAL-, TRIPLE-, AND FOUR-HEAD CAMERAS.

R. Z. Stodilka*, S. J. Glick, University of Massachusetts Medical School, Worcester, MA.

Sensitivity was studied using computer simulations of up to 4 heads. No collimation used. 4 heads were the best , of course, but 3 heads approached the sensitivity of 4 heads with certain radial positions.

806

DIGIRAD 2020TC SEGMENTED CAMERA - CALIBRATION METHODS.

N. E. Hartsough*, R. Yao, B. Pi, Digirad Corporation, San Diego, CA.

The calibration method for uniformity and energy were described in this poster.

More at http://www.digirad.com/default.htm

807

PERFORMANCE AND CLINICAL TRIALS OF A MULTI-PINHOLE NONROTATIONAL CARDIAC SPECT SYSTEM.

D. L. Kirch*, J. E. Koss, P. P. Steele, T. K. Johnson, Nuclear Cardiology Research, Englewood, CO; University of Colorado Health Sciences Center, Denver, CO.

Three seven pinhole cameras surround the patient. No rotation required. Energy information is acquired for correction of simultaneous Tc/Tl imaging. Provides an inexpensive SPECT system, say the authors.

808

ACCURACY OF 131I TUMOR QUANTIFICATION IN RADIOIMMUNOTHERAPY (RIT) USING SPECT IMAGING WITH AN ULTRA HIGH ENERGY COLLIMATOR.

Y. K. Dewaraja*, M. Ljungberg, K. F. Koral, University of Michigan, Ann Arbor, MI; University of Lund, Lund, Sweden.

The authors found the UHE collimator much better for quantitation.

 809

USE OF A PRE-STUDY INFORMATION DENSITY SCAN TO IMPROVE TRANSMISSION MAP ACQUISITION FOR ATTENUATION CORRECTION.

L. Shao*, J. Ye, M. K. Durbin, ADAC Laboratories, Milpitas, CA.

Authors found that statistics could be optimized by doing a prescan to determine heart count rates then adjusting emission and transmission scans accordingly.

810

PERFORMANCE EVALUATION OF THREE DIFFERENT SCINTILLATOR ARRAYS FOR BREAST IMAGING WITH POSITION SENSITIVE PMT.

J. H. Kim*, Y. Choi, K. C. Im, S. K. Woo, Y. S. Choe, K.-H. Lee, S. E. Kim, B.-T. Kim, Samsung Medical Center, Sungkyunkwan University, Seoul, South Korea.

The conclusions were that CsI(Tl) was better than NaI(Tl) and CsI(Na). This is different than the published abstract where NaI was favored.

811

CARDIAC IMAGING WITH GAMMA CAMERA COINCIDENCE (HYBRID) PET.

M. F. Smith*, B. P. Brown, D. R. Driver, V. Dilsizian, F. V. Schraml, J. M. Carson, S. L. Bacharach, The National Institutes of Health, Bethesda, MD; National Naval Medical Center, Bethesda, MD.

Coincidence camera was not very accurate in absolute quantitation due to dead time correction inadequacies. Relative quantitation in the heart was OK.

818

RELATIONSHIP BETWEEN CAMERA PHYSICAL CHARACTERISTICS AND CLINICAL PERFORMANCE OF PATIENT STUDIES: AN EVALUATION OF TWO DEDICATED PET SYSTEMS.

 F. Benard*, R. Smith, A. A. Alavi, J. Verreault, R. Hustinx, A. Matthies, Centre Universitaire de Sante de l'Estrie, Sherbrooke, PQ, Canada; University of Pennsylvania, Philadelphia, PA; Centre Hospitalier Universitaire de Liege, Liege, Belgium.

The authors were genuinely surprised that C-PET compared very favorably in clinical imaging tests compared to a 2D HR+ high end machine. They say they would get C-PET in spite of the better numbers for the more expensive machine. Was this clinical test capable of discriminating between machines? I suspect not.

819

EVALUATION OF SPECT IMAGING SYSTEMS BASED ON ACTIVITY ESTIMATION IN SMALL BRAIN STRUCTURES.

H. Jadvar*, S. C. Moore, M. F. Kijewski, S. Mueller, A. Bonab, R. E. Zimmerman, A. J. Fischman, University of Southern California, Los Angeles, CA; Harvard Medical School, Boston, MA.

It was found that dual head fan-beam collimators easily beat out dual HR collimators and was slightly superior to the Ceraspect dedicated brain imager.

823

DESIGN OF A PINHOLE COLLIMATOR FOR HIGH RESOLUTION 99mTc IMAGING.

E. C. Frey*, B. A. Chalker, B. M. W. Tsui, The University of North Carolina at Chapel Hill, Chapel Hill, NC.

A Monte Carlo study that looked at optimizing the material and the shape of the pinhole. W and Au are best and a pinhole keel thickness of 0.5 mm was found to be optimum.

864

MULTIDIMENSIONAL ANALYSIS AND VISUALIZATION SOFTWARE FOR DYNAMIC SPECT.

N. Roeber*, T. Moeller, A. M. Celler, T. Strothotte, T. Farncombe, Vancouver Hospital and Health Sciences Centre, Vancouver, BC, Canada; Simon Fraser University, Burnaby, BC, Canada; University of Magdeburg, Magdeburg, Germany.

Tools for dynamic SPECT were created in IDL.

 869

QUALITY CONTROL FOR SPECT ATTENUATION CORRECTION.

A. M. Celler*, L. Hook, D. W. Currie, R. D. Steuart, D. M. Lyster, Vancouver Hospital and Health Sciences Centre, Vancouver, BC, Canada.

This looked useful. I was sorry I could not take it home.

 870

THE IMPORTANCE OF THE TRANSMISSION RECONSTRUCTION ALGORITHM IN ATTENUATION CORRECTED TOMOGRAPHY: A PHANTOM STUDY.

E. P. Ficaro*, J. A. Fessler, University of Michigan, Ann Arbor, MI.

In a phantom study it was found that an iterative penalized likelihood reconstruction gave lower transmission crosstalk ratios over the 6 mo. life of the source than did FBP reconstruction.

872

THE INFLUENCE OF SCATTER COMPENSATION ON ATTENUATION COMPENSATED 99mTc SESTAMIBI SPECT CARDIAC SLICES.

P. H. Pretorius*, M. V. Narayanan, S. T. Dahlberg, J. A. Leppo, M. A. King, University of Massachusetts Medical School, Worcester, MA.

Scatter correction resulted in improved uniformity over the heart.

878

EVALUATION OF FOURIER REBIN METHOD FOR DUAL-HEAD HYBRID PET CAMERA.

J. Ye*, E. Gualtieri, L. Shao, D. Coles, B. Khanvali, J. Koster, ADAC Laboratories, Milpitas, CA.

FORE rebinning was good for contrast, resolution and reduces object shape distortion. When will it be clinical?

885

INFLUENCE OF THE RECONSTRUCTION TECHNIQUE (OSEM AND FBP) ON THE MAGNITUDE OF RING ARTIFACTS IN SPECT.

L. K. Leong*, R. L. Kruger, M. K. O'Connor, Mayo Clinic, Rochester, MN.

Still need the same good uniformity . Algorithm does not make any difference.

886

FULLY-3D BINNING AND FOURIER-SPACE REBINNING OF PET DATA ACQUIRED WITH ROTATING PLANAR DETECTORS.

D. J. Kadrmas*, E. V. Di Bella, University of Utah, Salt Lake City, UT.

I suspect this was an important paper. Authors found that angular bin size can cause unusual sensitivity functions. They went on to implement a version of FORE more suitable to large area coincidence detectors.

894

THE APPLICATION OF DYNAMIC SPECT IMAGING (DSPECT) TO RENAL IMAGING.

T. H. Farncombe*, A. Celler, R. Harrop, D. Lyster, D. Noll, J. Maeght, Vancouver General Hospital, Vancouver, BC, Canada; Simon Fraser University, Burnaby, BC, Canada; Universite Paul Sabatier, Toulouse, France.

Regional time activity curves were obtained from SPECT study. Sounds like it could be useful.

896

COMPARISON OF TWO CORRECTION METHODS FOR 18F FDG COLLIMATED SPECT IMAGING.

K. Knešaurek*, Mount Sinai Medical Center, New York, NY.

Restoration (resolution recovery?) was found to significantly improve the F-18 images. Scatter correction was little help.

1052

RADIOPHARMACEUTICAL PACKAGE INSERT DOSIMETRY TABLES: JUST HOW ACCURATE IS THE DATA?

J. C. Hung*, K. L. Classic, T. L. Mays, Mayo Clinic, Rochester, MN.

Package inserts are rarely updated with refined dose calculations as new information becomes available. It is off by as much as a factor of 2.

816

LOCALIZATION OF RECURRENT COLORECTAL CANCER USING FDG: CORRELATION BETWEEN PET AND INTRAOPERATIVE BETA AND GAMMA PROBES.

E. E. Zervos*, L. R. DePalatis, D. Desai, D. Soble, J. Frye, E. W. Martin, Ohio State University, Columbus, OH; Wendt-Bristol Radiology/PET, Columbus, OH

Both types of probes worked about the same.

BOOKS

Sorenson and Phelps 2nd edition of Physics in Nuclear Medicine remains my favorite book for resident teaching. I understand there will be a 3rd edition soon. Meanwhile Gopal Saha has a 2nd edition of Physics and Radiobiology of Nuclear Medicine on its way to bookstores in October. It may be worth a look.

Technical Exhibitors

BICRON

Bicron StarBrite crystals are a slotted crystal specially made for each application. Siemens is using them in the new Duet coincidence camera. The slots control light distribution in the thick crystals and help maintain spatial resolution across the energy spectrum. Bicron was not handing out technical data on this.

DIGIRAD (www.digirad.com)

Digirad has dropped the CZT detector for the time being. Too expensive, I guess. Their current camera is a small CsI(Tl) matrix with one-on-one Si photodiodes. About 3 mm intrinsic resolution with 15% energy resolution. 20 cm square FOV. May have some utility. Is SPECT in a spinning chair one of them?

GE

GE featured Hawkeye the x-ray tube on a SPECT gantry to do registered CT and SPECT. They also featured a number of strategic alliances, one of which was with the SNM. Hmmm. Sounds to me like they are exploiting the SNM to the detriment of other vendors.

Siemens

Duet (see above), they were taking orders for the CT/PET system. Consists of a HR with a spiral CT with gantries on a common axis. Caused a lot of interest. Also featured the corporate computer platform syngo with nuc med application e.soft.

SOFT Medical (www.softmedical.com)

A new (to me) PACS vendor playing in the nuclear medicine field. Seems reasonably comprehensive. Did not see a user list.

segami

This computer vendor had a lot of action, it seemed. In addition to selling direct, IS2, Digirad and GammaMedica (imaging probes) were using this computer on their own systems.

DAXOR (www.daxor.com)

First time I saw this company at SNM. They make a system for blood volume determinations. Pre-loaded syringe with I-131 albumin, small sample changer/counting system and computer for rapid calculation

Diagnostic Technologies (division of US Surgical Corp.)

Selling the Navigator GPS surgical probe for lymph node mapping. I wish there were NEMA specs for surgical probes.

 

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

zimmer@bwh.harvard.edu

Tel: 617 732 7196

FAX: 617 732 6336

http://www.med.harvard.edu/JPNM/physics/physics.html