This article is reprinted in its entirety, with corrections, from J Nuci Med 21:971-977, 1980.Theoriginal manuscript and proofread galleys submittedby theauthorwerecorrect.Errorsintroduced at a later stage resulted from the implementation of new software, for which the printeraccepts full responsibility and deeply regretsany inconvenience caused to the author and thereaders. This unusual problem has been fullycorrected and will not occur again. Please placethis insert in the October 1980 issue.

BASIC SCIENCES

PHYSK@S AND RADIATiON BIOLOGY

Method for OptimizingSide Shieldingin Positron-EmissionTomographsand for

Comparing Detector Materials

Stephen E. Derenzo

University of California, Berkeley, Cailfornia

This report presents analytical formulas for the Image-forming and backgroundevent rates seen by circular posltron-emlssion tomographs with parallel sideshielding. These formulas include deadtime losses, detector effIciency, coincidence resolving time, amount of activity, patient port diameter, shielding gap, andshielding depth. A figure of merit, defined In terms of these quantities, describsthe signal-to-noise ratio in the reconstructed Image of a 20-cm cylinder of waterwith uniformly dIspersed actlvfty. For 1-cm-wide NaI(Tl) detectors, a 50-cm patient port, an actIvity of 200 iCi per axial centimeter, and a Shieldinggap of 2 cm,the optimum shielding depth Is 20 cm, which requires a detector circle diameterof 90 cm. For a 25-cm patient port and other conditionsas above, the optimumshielding depth Is 14 cm. Resufts are presented for the scintiliators Nai(Tl), bismuthgermanate(BGO), CsF, and plastic;and for Ge(Li) and wire chamberswfthconverters. in these examples, BGO provided the best signal-to-noise for activftylevels below 1000 iCi per cm, and CsF had the advantage for higher activftylevels.

.JNuciMed 21: 971—977,1980

In positron-emission, transverse-section tomography, the image is derived from the detection of unscattered coincident annihilation pairs. Almost all positron-emission tomographs use shielding on either sideof the detector plane to block activity external to thetransversesectionbeingimaged(1—8).Shieldingis alsoused between detector planes in multiple-section devices(9—15).In spite of this shielding, image contrast is degraded by true coincidencesof scattered annihilationpairs and by accidental coincidences of unrelated annihilation photons (Fig. 1). Most positron-imaging systemsoperate with scatteredand accidentalbackgroundsthatare each typically 20%of the detected coincidences. Evenif these backgrounds can be perfectly estimated andsubtracted from the detected coincidences, the random

Received Nov. 5, 1979;revisionaccepted June 18, 1980.For reprints contact: Stephen E. Derenzo, PhD, Donner Laboratory,

University ofCalifornia, Berkeley, CA 94720.

Volume 21, Number 10

fluctuations in the result are greater than if the backgrounds did not exist. In the following sections we cxamine the trade-off between sensitivity and backgroundsand describe a procedure for determining the optimumshielding depth for single-ring circular positron-emissiontomographs. This treatment does not consider, but canbe extended to include, nonparallel shields and multisliceconfigurations.

The procedure consists of (a) measuring image andbackground event rates from a 20-cm phantom using thetype of detector system being considered; (b) fittinganalytical expressions to these rates; and (c) varying theshielding depth in those expressions to maximize thestatistical accuracy in the reconstructed image.

As an example, this procedure is carried out for circular detector arrays using the Donner 280-Crystalpositron tomograph with NaI(TI) and bismuth germanate (BGO) detecturs. Since the result. depend significantly on detector characteristics, addtional exampies are provided for a variety of detectors, using typical

Thiea One

@ I@@ IIIII@III@I@II@IIIIl@II@III@@ @IIIID@II@IIII@ FNN4FNOGK6A

971

ERRATUM

by on January 11, 2020. For personal use only. jnm.snmjournals.org Downloaded from

STEPHENE. DERENZO

FATES OF ANNIHILATION PHOTONS

values of detection efficiency and of time and energyresolution.

EVENT RATES

As has been shown analytically for circular positron-emission tomographs (16), the overall rate of unscattered coincidentevents (C1)is givenby:

C1= B1f2pG@/(H+ @P)

UNSCATTERED TRUE COINODEP4CE SINGLES

FIG. 1. Types of coincident events dotected by positron tomographs. Image isformed from unscattered coincident annihilation pairs. Coincident scaflered pairsand accidental coincidences of unrelatedphotons resuft In broad backgrounds.

where H is the shielding depth in centimeters and B5 isa constant that incorporates pulse-height thresholds, theangular distribution of accepted Compton scatters, andnumerical factors.

FIG. 2. Detector, shiedng, and phantom geometry for op@mIzation(3) calculations.ScatterbackgroundcanbemeasuredInInn&cyikider

by imaging with activity in outer cylindricalannulus only.

SCATTERED TRUE COI4CE*NCE ACCCENTAL COINCIDENCE

(1)

where Eis the detection efficiency for annihilation photons (including detector packing fraction), p is the activity density in jzCi per axial centimeter, G@is the ef

fec@iveshielding gap (cm), H is the shielding depth (cm),p (Fig. 2) is the patient port diameter (cm), and B1is aconstant that incorporatesthe averageattenuation andnumerical factors. R = H + @Pis the detector ring radius. For activity distributed in a 20-cm cylinder ofwater, B1 = (37,000 sec'@Ci') X (average attenualion) X (1/4) = 1850sec'zCi'. Due to edge penetration, the effective shielding gap, G@,is slightly largerthan the physical shielding gap, G:

G@= G + ÔG. (2)

A differentanalyticalexpressionwasderivedin(16) forthe overall rate of coincident scattered events (Cs):

C - Bs€2pG@S@H(H +@P)

972 THE JOURNAL OF NUCLEAR MEDICINE

by on January 11, 2020. For personal use only. jnm.snmjournals.org Downloaded from

HDTFD1d5/d1dA/did1/d@0Shielding

depth (cm)Observedtotalt(s@c1)Deadtimeloss (%)Imageevent

rate (sec1)Scatter/ImageratioAccld./lmageratioImage contrastQualityfactor

(sec insi)

BASIC SCIENCESPHYSICS AND RADIATION BIOLOGY

The overall rate of accidental events (CA) given in(16) is:

CA BA€2rp2G@/H2, (4)

where r is the full coincidence-time window. in nanoseconds, and BA @Sa constant that incorporates pulseheight thresholds, detector efficiencies for scattered andunscattered photons, and numerical factors.

These rates are reduced by system deadtime, whichis a combination of the deadtime of the detectors, thetiming and pulse-height discriminators, the coincidencecircuits, and memory. We assume that for a particularscattering medium (i.e., a 20-cm cylinder of water) theratio of photon interactions to coincident events is fixed,and that we can define an effective nonparalyzing systemdeadtime that applies to the total coincident-event rateonly. Before deadtime losses, the system detects C1 C@CA in the on-time coincidence window and CA in anoff-time window. Thus the total coincidence rate is CT= C1 C@ 2CA, and the fraction of events, F, that is lost

to deadtime is given by:

F = tCT/(1 tCT),

where t is the deadtime per event. The observed systemrates are:

D1 = (1 —F)C1

D@= (1 —F)Cs

DA=(1-F)CA. (6)

In the central region of the reconstructed image of acylinder of activity in water, the intensity of unscatteredcoincident events per square centimeter (d1) is givenby:

d1 = (1@ F)b1€2pG@/(H+ @P).

For a 20-cm cylinder, b1 = 29.44 reconstructed eventssec' cm2@zCi'. The intensity of scattered coincidentevents per square centimeter (ds) is given by:

(I— F)bsc2pG@d@= H(H+@P) ‘ (8)

and the intensity of accidental events per square centimeter (dA) is given by:

dA (1 —F)bAE2rp2G@/H2. (9)

Before background subtraction, the total intensity, dT,is given by:

dT=dIdSdA. (10)

DEFINITION OF THE FIGURE OF MERIT

A figure of merit (Q) can be defined as the product ofthe unscattered coincidence rate (Dj) and the imagecontrast (dI/dT) using the arguments of Beck (17):

Q=DI(dI/dT) (11)

(5) Q may also be called an “effective―image event rate,since the same signal-to-noise ratio would be obtainedin an ideal tomograph with D1' = Q and d5' = dA' = 0.Note that d1,d5, and dA all undergo the same deadtimeeffects, attenuation correction, and error propagationin the reconstruction process.

From Ref. 18, it is possible to relate the value of Q(and the imaging time, T) to the statistical uncertaintyin the reconstructed image. In the case of a 1-cm2 cellnear the center of a 20-cm cylinder of uniform activity,the fractional rms uncertainty is given by 90/@/@!

OPTIMIZATION OF THE FIGURE OF MERIT

(7) Equations1—10showthatforagivenimagingsitua

TABLE1. SHIELDINGOPTIMIZATiONFOR NaI(Tl)

5118,58211.98,8050.972.59.221,9321040,0004.08,2200.480.760.453,6721522,3502.27,3250.320.380.594,29520@15,3521.56,5580.240.240.674,4182511,7181.25,9240.190.170.734,337309,5131.05,3970.160.130.774,175406,9710.74,5790.120.090.833.789505,5410.63,9740.100.060.863,422604,6170.53,5100.080.050.883,102

a Port 50cm, effectIve shieldIng gap, G. 2 cm, pulse-height threshold 100 keV, detector efficiency e 45% , p 200

@Cl/cm,deadtlme t = 1 @isec,coIncidence-tIme window r@ 15 nsec.t Total rate in on-time and off-time coincidence windows combined.: H 19.8cmformaximum0.

Volume 21, Number 10 973

by on January 11, 2020. For personal use only. jnm.snmjournals.org Downloaded from

STEPHEN E. DERENZO

tion, reducing the shielding depth, H, improves theimaging rate, D1, but also decreases the image contrast,

dI/dT. Choosing a value of H that maximizes Q (Eq. 11)ensures the best tradeoff between sensitivity and imagecontrast.

By combining Eqs. 1- 11 we show the dependence ofQontheshieldingdepth,H:Q(H) =

dQdH

A6H3[4Ao + 3A1H + 2A2H2 + A3H3 A5H51

[Ao + A1H + A2H2 + A3H3 + A4H4 + A5H5J2@

(14)

An analytical expression for an optimum value of H (i.e.,a formula for H in terms ofAo, A1, A2, A3, and A5 suchthat dQ/dH = 0) requires the general solution to the

(12) quintic equation, which has not been accomplished.However, since the coefficients Ao, . . . , A6 are positive,it maybeshownfromEq.14thattheslopeof Q isequalto zero for only one value of H, is greater than zero forsmaller values of H, and is less than zero for larger valuesof H. As a result Q has only one extremum—a maximum.

In this work the optimum value of H was determinedin each case by using an iterative quadratic interpolation

algorithm on a small digital computer.

EXAMPLES

NaI(TI). The constants €,Bg, BA, bS, bA, and thepenetration factor ô@were determined for NaI(Tl) byfitting Equations 1-9 to measurements of 20-cm phantoms made by the Donner 280-crystal positron tomograph. The overall rates D1,D@,and DA were measuredfor a 20-cm-diameter cylinder of activity in water (6),

and ds/(ds + d1)and dA/(ds + d1)were measured at the(I 3) center of reconstructed images of a 5-cm-diameter cyl

inder containing only water, surrounded by a 20-cmdiameter annulus of activity in water (Fig. 2). Thepulse-height threshold was 100 keV, the coincidencetime window was 20 nsec, the activity was varied from100 to 300 MCi/cm, and the shielding gap was variedfrom I to 3 cm. A good fit was obtained with €= 45%,B5= 5100,BA =0.73,b@= 7l,bA 5.3 X 103,andôG= 2. 1 mm. The penetration factor ÔG was necessary in

each of Equations 1, 3, and 4 (D1, D5, and DA, respectively) for an adequate fit to the data. Using thesecoefficients in Eqs. 1-10, the shielding depth, H, wasvaried to maximize the figure of merit, Q, in Eq. 11.

As an example, Table I lists D1,ds/d1, dA/dI, dl/dT,and Q as a function of shielding depth, H, for a 50-cmpatient port, p = 200 jzCi/cm, G@ 2 cm effectiveshielding gap, a 20-cm-diameter water cylinder, and a1-zsec deadtime. Figure 3 presents curves of Q as afunction of H for the same conditions except for 25-cmand 50-cm patient ports.

Other detector materials. The relative rates D1, D5,and DAfor bismuth germanate (BOO) detector crystalshave been measured in this system with a 300-keVpulse-height threshold (6), and led to the constants €67%, B@ = 5100,BA 0.36,b5 = 71,bA 2.6X i0@.Note that the photopeak selection reduces BAand bA,but not B5 and b@,since the scatter background consists

A6H4

A0 + A1H + A2H2 + A3H3 + A4H4 + A5H5'

where

A6 B1b1E2pG@

A5 = b1

A4 b5G@b1P/2 pG@(tB1b1e2bAr)

A3 = Gc[barpGcP/2 + (tb1€2pG@)x (B5 + BArpG@)+ (P/2 + tB1E2pG@)

x (b@+ bArpGe)]

A2 pG@(bArP/2) (P/2 + B1t€2pG@)

tpE2G@4[BAbIrpP/2 + (B5 + BATPGC)

(b5 + bArpG@)]

A1 (tm2p2G@5P/2) [bA(BS + BATPGC)

+ BA(bS+ bArpGe)]

and

A0 = BAbAtr2E2p3G@6P2/4.

The slope of Q is given by:

‘Ua,

0

0

030

02 5 10 20 50ShieldIng depth (cm)

FiG. 3. Effective event rate, 0, as function of shielding depth forNal(fl) detectors with 45% detectIon efficiency, 2-cm Shieldinggap,200 @zCi/cm,1-,@secdeadtime, and coincidence resolving time of15 nsec. Optimum values are indicated with arrows for 25- and50-cm patient ports.

974 THE JOURNAL OF NUCLEAR MEDICINE

by on January 11, 2020. For personal use only. jnm.snmjournals.org Downloaded from

TABLE2. OPTIMUMVALUESOF H AND 0 FOR VARIOUSDETECTORMATERIALSWire

chamberNal(Tl)BGOCsFGe(Li)convertersPlastic

BASIC SCIENCESPHYSICS AND RADIATION BIOLOGY

Pulse-height threshold: 100 keV 300 keV 100 keV 500 keV 200 keV 100 keV511-keVefficiency: 45%t 67%t 50%t 15%t 20% 20%tCoincidence-tIme window: 15 ns 30 ns 3.3 ns 15 ns 50 ns 3.3 nsB5 5,100 5,100 5,100 0 5,100 5,100BA 0.73 0.36 0.73 0.15 0.50 0.73b5 71 71 71 0 71 71bA 5.3 X iO@ 2.6 X iO@ 5.3 X iO@ 1.0 X iO@ 3.5 X iO@ 5.3 X iO@p = 100 MCi/cm

Shielding depth: H (cm) 16.3 16.5 12.7 6.9 19.8 12.5Image contrast: d1/d@ 0.68 0.69 0.70 0.88 0.66 0.69Qualftyfactor:Q(sec1) 2,460 5,417 3,390 459 435 546

p = 200 MCi/cmShielding depth: H (cm) 19.8 20.3 14.1 8.9 24.7 13.8Image contrast: d,/d@ 0.67 0.68 0.70 0.85 0.63 0.69Qualfty factor: 0 (@@1) 4,419 9,648 6,477 835 750 1, 1

p = 500 MCi/cmShielding depth: H (cm) 27.6 29.4 17.6 12.5 34.6 16.7Image contrast: d1/d@ 0.65 0.67 0.70 0.80 0.58 0.68Quality factor: Q (@@1) 8,846 18,858 14,450 1,775 1,42 2,399

p = 1000MCi/cmShielding depth: H (cm) 37.9 42.1 22.4 16.4 46.4 20.4Image contrast: d,/d@ 0.63 0.66 0.70 0.76 0.53 0.67Qualityfactor: 0 (sec1) 13,827 28,561 24,894 3,031 2, 9 4,284

. Port 50 cm, effective shielding gap Ga, 2 cm, deadtime t 1 @sec per coincident event.

t Detector size I cm wide X 5 cm deep.

primarily of photons above 415 keY that have scatteredthrough <40°(5).

By using the values for efficiency, time resolution, andenergy resolution for other detector materials, it is possible to optimize the shielding depth for each material.Table 2 lists the results for six detector materials at fouractivity levels. See (19) for a more extensive tabulation.The detection efficiency for Ge(Li) and plastic was determined from Monte Carlo calculations that traced theinteractions of a beam of 51l-keV photons through agroup of 1-cm-wide detectors (20). The detection efficiency was defined as the fraction of incident photonsthat produced a signal above threshold in only one detector.

Note that Table 2 was intended to provide examplesof results of the optimization method only for a varietyof detector materials, and does not necessarily representthe best that can be done.

DISCUSSION

In the examples considered in Table 2 (a circular detector array, parallel shields with a 2-cm gap, and a50-cm patient port) the highest optimum value of Q isachieved with BGO at each of the four activity levels.

The second best material is CsF, which has a lower detection efficiency than BGO but significantly better timeresolution.

Ge(Li) is unique in having sufficient energy resolutionto reject almost all tissue-scattered photons. This elim

mates the coincident scattered background and greatly

reduces the accidental background. The low full-energyefficiency for 5 11-keV photons, however, results in lowvalues of Q.

Plastic scintillators are unique in having such excellenttiming resolution that the coincidence-time window of2—3nsec is determined by the size of the patient port andthe speed oflight. If time resolutions of the order of ±100psec could be realized for tomographic systems, thetiming information and the use of shorter detectors couldlocalize the annihilation point along the line of flight to±1.5 cm. This possibility was suggested by Anger in1966 (2!) and is used in a 3-dimensional imaging systembuilt by Nickles and coworkers (22). The use of CsF fortime-of-flight positron tomography has been suggested

by Allemand et al. (23) and by Mullani et al (24). InRef. 23 a 0.5-nsec FWHM timing resolution is reported,and this is estimated to improve the signal-to-noise ratioin the reconstructed image of a 32-cm cylinder of activityin water by a factor of 2.9.

Volume 21, Number 10 975

by on January 11, 2020. For personal use only. jnm.snmjournals.org Downloaded from

STEPHEN E. DERENZO

For the case of wire chambers with lead converters,rather optimistic values of efficiency and time resolutionhave been used, but the resulting quality factors are stillrelatively low. The situation may improve, as severalgroups are investigating schemes to improve theirproperties (25-28).

CONCLUSIONS

For positron-emission tomographs, one can define aquality factor that describes the signal-to-noise ratio inthe reconstructed image, and it is then possible to choosea side-shielding depth and detector-circle diameter thatoptimize the tradeoff between sensitivity and imagecontrast. The method requires knowledge of the scaling

coefficients for the scattered and accidental backgroundrates as well as an adequate description of the detectorsbeing considered.

Under these conditions, the comparison of optimalquality factors permits a direct comparison of the suitability of different detector materials for positronemission tomography in terms of the signal-to-noiseratio. This approach may be expanded to include nonparallel shields and multislice geometries.

ACKNOWLEDGMENTS

I thank T. F. Budinger, L. Carrol, and R. H. Huesman for helpfuldiscussionsand T. Vuletich for valuable technical assistance.This workwas reported in part at the 26th Annual Meeting of the Society ofNuclear Medicine, Atlanta, Georgia, June 26-29, 1979.

This work was supported by the Div.of Biomedical Research of theU.S. Dept. of Energyunder Contract No. W-7405-ENG-48and byNIH Grant No. CA 17566-3.

REFERENCES

I. ROBERTSON iS, MARR RB, ROSENBLUM B, et al:Thirty-two crystal positron transverse section detector. InTomographic Imaging in Nuclear Medicine. FreedmanOS,Ed. New York, Society of Nuclear Medicine, 1973, pp142—153

2. HOFFMANEJ,PHELPSME, MULLANINA, etal:Designand performance characteristics of a whole-body positrontransaxial tomograph. J Nuci Med I7:493-502, 1976

3. CH0 ZH, COHEN MB, SINGH M, et al: Performance andevaluation of the circular ring transverse axial positron camera(CRTAPC). IEEE Trans Nuci Sci NS-24, No I :532-543,I977

4. ERIKSSON L, BOHM C, BERGSTROMM, et al: One yearexperience with a high resolution ring detector positroncamera system: present status and future plans. IEEE TransNuciSciNS-27,No 1:435-444,1980

5. DERENZO SE, BUDINGER TF, CAHOON JL, et al: Highresolution computed tomography of positron emitters. IEEETrans NuciSci NS-24, No 1:544-558, 1977

6. DERENZOSE,BUDINGERTF, CAHOONJL,et al:TheDonner 280-crystal high resolution positron tomograph. IEEETrans Nuci Sci NS-26, No 2:2790-2793, 1979

7. THOMPSON CJ, YAMAMOTO YL, MEYER E: Positome II:a a high-efficiency positron imaging device for dynamic brain

studies. IEEE Trans NuciSci NS-26, No 1:583-589, 1979

8. RICCI AR, HOFFMAN EJ, PHELPS ME, Ct aI: Emissioncomputed tomography simulator. IEEE Trans Nuci SciNS-27, No 1:479-484, 1980

9. BROWNELL0, BURNHAM C, CORREIAJ. Ctal: Transversesection imaging with the MGH positron camera. IEEE TransNuci Sd NS-26, No 2:2698-2702, 1979

10. MULLANI NA, HIGGINS CS, HooD JT, et al: PETT IV:design analysis and performance characteristics. IEEE TransNuciSciNS-25,No 1:180—183,1978

11. TER-P000SSIAN MM, MULLANI NA, HooD J, Ctal: Amuitislice positron emission computed tomograph (PElT IV)yielding transverse and longitudinal images. Radiology128:477—484,1978

12. TER-POGOSSIAN MM, MULLANI NA, HooD iT, Ct al:Design considerations for a positron emission transverse tomograph (PElT V) for imaging ofthe brain. I Comput Assist

Tomogr 2:539-544, 197813. CARROLL LR: Design and performance characteristics of

a production model positron imaging system. IEEE TransNuc/Sci NS-25, No 1:606-614, 1978

14. CH0 ZH, NALCIOGLU 0, FARUKHI MR: Analysis of acylindrical hybrid positron camera with bismuth germanate(BGO) scintillation crystals. IEEE Trans Nuci Sd NS-25,No 2:952—963,1978

15. CARROLL LR, HENDRY GO, CURRIN JD: Design criteriafor multislice positron emission-computed tomography detector systems. IEEE Trans NuclSci NS-27, No 1:485-488,1980

16. DERENZO SE, ZAKLAD H, BUDINGER TF: Analytical studyof a high-resolution positron ring detector system for transaxial reconstruction tomography. I Nuci Med 16:1166-1173,1975

17. BECKRN: A theoryof radioisotopescanningsystems.InMedical Radioisotope Scanning. vol 1, Vienna, IAEA, 1964,pp35-56

18. BUDINGER TF, DERENZO SE, GREENBERG WL, et al:Quantitative potentials of dynamic emission computed tomography.JNuclMed 19:309-315, 1978

19. DERENZO SE: Tabulation ofside shielding optimization andcomparison of detector materials for circular positron emissiontomographs. Berkeley, CA., Lawrence Berkeley Lab. ReportLBL-9584, 1980

20. DERENZO SE: Monte Carlo calculations of the detectionefficiency oflinear arrays of NaI(Tl), BGO, CsF, Ge(Li) andplastic for 51 1 keV photons. Berkeley, CA., LawrenceBerkeley Lab. Report LBL-9201, 1980

21 . ANGER HO: Survey of radioisotope cameras. ISA Trans 5:311—334,1966

22. NICKLESRi, MEYERHO: Designof a three-dimensionalpositron camera for nuclear medicine. Phys Med Biol 23:686—695,1978

23. ALLEMAND R, GRESSET C, VACHER J: Potential advantages of a cesium fluoride scintillator for a time-of-flightpositroncamera.JNuclMed 21:153—155,1980

24. MULLANI NA, FICKEDC, TER-P0GOSSIANMM: Cesiumfluoride: a new detector for positron emission tomography.IEEE Trans Nucl Sci NS-27, No 1:572-575, 1980

25. CHUD,TAMKC,PEREZ-MENDEZV, Ctal:High-efficiencycollimator-converters for neutral particle imaging withMWPC. IEEE Trans Nuci Sd NS-23, No 1:634-639,I976

26. NEUMANN Mi: A gas-filled channel converter for fast coincidence cameras. IEEE Trans Nuci Sci NS-24, No I:515—520,1977

27. JEAVONSAP, TOWNSENDDW, FOkDNL, etal:A highresolution proportional chamber positron camera and its applications. IEEE Trans Nucl Sd NS-25, No 1:164-173,

976 THE JOURNAL OF NUCLEAR MEDICINE

by on January 11, 2020. For personal use only. jnm.snmjournals.org Downloaded from

BASIC SCIENCESPHYSICS AND RADIATION BIOLOGY

glass tubing converters for gamma detection in MWPC. IEEETransNuclSciNS-27,No 1:157-165,1980

197828. LUM GK, GREEN MI, PEREZ-MENDEZV, et al: Lead oxide

F.bruary 6-7, 1981 N•wOrleans, Louisiana

The Computer Council and Instrumentation Council of the SOCIetYof Nuclear Medicine will meet February 6 and 7,1981 at the Marriott Hotel in New Orleans, Louisiana.

A topical symposium is being sponsored by the SNMComputer Council consisting of invited presentations, contnbuted papers, educational sessions, and active attendee participation. Submitted papers are encouraged on Functional Mapping of Organ Systems. Submitted papers willalso beconsidered in otheraspectsofthe use of instrumentsand computers in nuclear medicine.

The Councils welcome the submission of abstracts from members and nonmembers of the Society of Nuclear Medicine. The title, author, and institutional affiliations should be included at the top of the first page. The name of theauthor presentingthe paper mustbe underlined.Abstractsshouldcontaina statementof purpose, the methodsused,results, and conclusion.

Originalabstracts and supporting data should be sent in duplicate to:

Computer CouncilRonald R. Price, Ph.D.

Dept. of Radiology and Radiological SciencesVanderbiltUniversityMedicalCenter

Nashville, TN 37232Phone: (615) 322-2394

Instrumentation CouncilL Stephen Graham, Ph.D.Nuclear Medicine Service

Veterans Administration Medical Center16111 Plummer

Sepulveda, CA 91343Phone: (213) 891-2485

Abstracts must b. recslvsd no later than October 1, 1950.

Volume 21, Number 10 977

SOCIETY OF NUCLEAR MEDICINECOMPUTER COUNCIL ANNUAL MEETINGINSTRUMENTATION COUNCIL MEETING

Marrioft HotI

by on January 11, 2020. For personal use only. jnm.snmjournals.org Downloaded from

1980;21:971-977.J Nucl Med.   Stephen E. Derenzo  Comparing Detector MaterialsMethod for Optimizing Side Shielding in Positron-Emission Tomographs and for

http://jnm.snmjournals.org/content/21/10/971.2This article and updated information are available at:

  http://jnm.snmjournals.org/site/subscriptions/online.xhtml

Information about subscriptions to JNM can be found at:  

http://jnm.snmjournals.org/site/misc/permission.xhtmlInformation about reproducing figures, tables, or other portions of this article can be found online at:

(Print ISSN: 0161-5505, Online ISSN: 2159-662X)1850 Samuel Morse Drive, Reston, VA 20190.SNMMI | Society of Nuclear Medicine and Molecular Imaging

is published monthly.The Journal of Nuclear Medicine

© Copyright 1980 SNMMI; all rights reserved.

by on January 11, 2020. For personal use only. jnm.snmjournals.org Downloaded from