Molecular Imaging and BiologyVol. 6, No. 3, 131–138. 2004

Copyright � 2004 Elsevier Inc.Printed in the USA. All rights reserved.

1536-1632/04 $–see front matter

doi:10.1016/j.mibio.2004.02.004

ARTICLE

Intravascular Probe forDetection of Vulnerable Plaque

Martin Janecek1, Bradley E. Patt2,Jan S. Iwanczyk2, Lawrence MacDonald2, Yuko Yamaguchi2,

H. William Strauss3,4, Ross Tsugita4, Vartan Ghazarossian4, Edward J. Hoffman1

1Washington University School of Medicine, St. Louis, MO, USA, 2Photon Imaging, Inc., Northridge, CA, USA,3Stanford University School of Medicine, Standford, CA, USA, 4Imetrx, Inc., Mountain View, CA, USA

PURPOSE: Coronary angiography defines geometry of lumen of artery. However, perhaps70% of heart attacks occur when minimally obstructive thin capped fibroatheroma rupture,causing thrombus and arterial occlusion. We have developed an intravascular imagingdetector to identify vulnerable coronary artery plaque.PROCEDURE: Detector measures beta or conversion electron emissions from plaque-binding radiotracers. Detector assembly fits into a 2-mm diameter catheter and overcomestechnical constraints of size, sensitivity, and conformance to intravascular environment.RESULTS: Device was tested by stepping test point sources past detector to verify function.System resolution is 6.7 mm and sensitivity is 400 cps/µCi one mm from detector.CONCLUSION: This prototype is a first step in imaging of labeled vulnerable plaque incoronary arteries. This type of system may assist in development of targeted and cost effectivetherapies to lower incidence of acute coronary artery diseases (CAD) such as unstable angina,acute myocardial infarction, and sudden cardiac death. � 2004 Elsevier Inc. All rights reserved.

Key Words: Imaging probe; Intraoperative probe; Vulnerable plaque; Beta-emitting isotope;Coronary artery disease; Nuclear medicine; Atheroma.

Introduction

Coronary Artery Disease

Coronary artery disease (CAD) is an inflammatorydisease in the arterial wall. Although the initiat-ing factors are poorly characterized, irritating

lipoprotein cholesterol complexes localize beneath thevascular endothelium. The inflammation caused bythese lipids attracts phagocytic cells, monocytes, andmacrophages to the area to engulf and digest theseirritating lipid complexes. These cells are rarely success-ful at completely cleaning up the lipid, and the le-sions often progress. If the inflammation is severe,the lesion ruptures the cap separating the atheromafrom the blood. When the irritating contents of theatheroma comes in contact with blood, the clotting cas-cade is initiated, leading to formation of a thrombus, andpossible occlusion of the vessel. Coronary disease is theleading cause of death in most developed countries.1

Address correspondence to: Martin Janecek, Ph.D., Washington Uni-versity School of Medicine, Dept. of Radiological Sciences, CampusBox 8225, 510 South Kingshighway, St. Louis, MO 63110. E-mail:janecekm@mir.wustl.edu

A large portion of atheroma are undetected, until apatient has a sudden (often fatal) coronary event.

Atheroma are classified histologically into six catego-ries based on the size of the lipid lakes and the numberof inflammatory cells. Atheroma in Class IV and V arethose at risk of acute rupture. In patients below theage of 60, these lesions are often difficult to detectbecause the inflammation is contained in the wall ofthe vessel, and are therefore difficult to detect withcoronary angiography. This method only displays theluminal dimensions of the blood vessels and providesno information about the plaque contents or status.Up to 70% of all heart attacks are believed to be causedby vulnerable plaques.2,3 An overwhelming majority ofvulnerable plaques do not restrict the blood flowduring their build up and are thus not observed byangiography.

Several methods of detecting vulnerable plaque arebeing explored.4 Our approach, nuclear medicine im-aging, has the advantage that specific characteristicsof vulnerable plaque can be interrogated—such as themetabolism of inflammatory cells, localization of lipo-protein cholesterol, the degree of apoptosis, or endothe-lial cell activation—with radioactive tracers.5

131

132 Molecular Imaging and Biology, Volume 6, Number 3

Materials and Methods

Design Aspects

We have designed and built a miniature, compact, andflexible intravascular probe. To be able to image vulner-able plaques (which are extremely small) in the coro-nary arteries (which are highly likely to accumulatesignificant amounts of tracer), our strategy is to use beta-emitting radionuclides. Since gamma radiation is verypenetrating and a collection of tracer in the heart couldcreate a potentially overwhelming background, it wouldbe difficult if not impossible to see the relatively weaksignal from the vulnerable plaques. Beta-emitting com-pounds, on the other hand, are not very penetrating;therefore, one can be certain that—even though theplaques may not accumulate large amounts of tracer—the detected beta radiation is emitted from plaqueswithin a very close proximity to the detector. Also, sincethe beta range is small, the beta-radiation detectormust be brought into close proximity to the vulnerableplaque (�1 mm) and this leads to a very high sensitivity(since the geometric factor will be very high because thedetector is virtually touching the plaque).

Since the detector must be inserted into the samespace as an angiographic catheter, its dimensions mustbe similar. The size must eventually be restricted to 1.5mm in diameter or less, it must be maneuverable intoplace with a guidewire, and it must be safe (e.g., no highvoltages or hazardous materials). The size-restrictingproblem is addressed in a stepwise manner where ineach step the diameter of the system is decreased. Thecatheter detector described in this article is 1.9 mm indiameter, and the ultimate aim for this project is tohave a catheter smaller than one mm in diameter.

System Overview

The prototype catheter is shown in Figure 1. In practice,the catheter will be introduced into the artery systemat the groin, and passed over a conventional guidewireunder fluoroscopic guidance into the coronary arteryto be examined. The detector, sitting at the tip of thecatheter, detects the beta-labeled plaques and the signalis collected at the photomultiplier tube (PMT) that sitsat the interface. The part of the device from the distaldetector to the PMT is of single use. The signal is de-coded into a response image that is displayed on thecomputer screen. The linear response display is a mea-sure of the counts detected in each of the six individualdetectors units. The total instantaneous count rate isput to an audio output whose tone corresponds to thecount rate.

Figure 1. Picture of catheter system.

Detector

The front-end of the detector consists of six scintillatingfibers, see Figure 2, each offset from the next to allowsimultaneous readings along a length of the coronaryartery, giving a crude image of the plaque. Each scintil-lating fiber acts as a pixel in a one-dimensional array.The scintillating fibers are arranged in a spiral aroundthe guidewire. Each fiber is 6.7 mm long and offsetlongitudinally to the next by 6 mm, which should createa resolution of about 6 to 7 mm and a field-of-view(FOV) of 4 cm. The scintillating fibers are 0.5-mm diam-eter plastic fibers with polystyrene-based cores. The coreis doped with scintillating phosphor and is surroundedby a one-layer acrylic cladding. The scintillation yield isabout 8000 photons per MeV of absorbed energy andthe trapping efficiency ranges from 3.44% for eventsat the center of the core to ∼7 % for events at the core-cladding interface. The scintillating fibers absorb thebeta radiation and emit light photons in the visible lightregion. The peak photon yield in the light spectrumis at 435 nm, with a spread of about 60 nm, which

Figure 2. Front-end of detector.

Vulnerable Plaque Detection/ Janecek et al. 133

matches the peak of the PMT response sensitivity curve.The light photons are guided to the position-sensitivephotomultiplier tube (PSPMT) by clear double-claddedplastic fibers, 1.5 m in length. The scintillating fibersand the clear fibers have the same core material, whichmakes the fused bond strong and minimizes signal lossat the optical interface. To increase the light yield, thedistal end surfaces of the scintillating fibers have beencoated with white reflective acrylic paint. Also, the endsof the scintillating fibers and the optical fibers have allbeen polished to minimize any loss of light signal. Boththe scintillating fibers and the clear optical fibers areoff-the-shelf products from Saint-Gobain Crystals &Detectors (formally Bicron: Newbury, OH).

Positron Range

Compounds labeled with positron emitters, such as 18F,are potential candidates for the detection of vulnerableplaque. The detection system must be sensitive to thepositron during its period as a positively charged betaparticle and insensitive to it as a 511 keV gamma-rayafter annihilation. Because it is highly penetrating, the511 keV gammas can reach the detector from any partof the body. Both distance and low sensitivity to thisradiation will be useful in reducing this potential sourceof background. Approximately 1% of impinging 511keV gammas will interact with the individual fibers.Figure 3 shows the energy range and energy loss forbetas in polystyrene plastic. To detect positrons in therange 0.1–1.0 MeV with high efficiency, while sup-pressing the gamma background, the detector material

Figure 3. Beta range and stopping power in polystyrene.

must have a high stopping power for betas while having alow stopping power for 511 keV gammas. Scintillatingplastic fibers have these properties and in addition arerelatively easy to cut, polish, and fuse to form the finaldetector system.

Catheter

A multiple compartment catheter was designed and fab-ricated for the detector, see Figure 4. The catheter in-cludes three peripheral chambers housing two opticalfibers each, and a central compartment for the guide-wire. An opaque dark blue material was chosen toprevent light from the outside from entering into thefibers and also to minimize crosstalk between neigh-boring optical fibers. The wall thickness of the catheterwas kept thin to minimize attenuation of the betas andalso to increase flexibility of the catheter.

Electronics

The PSPMT used is a Hamamatsu bialkali cross-wirePSPMT (Hamamatsu Corp., Bridgewater, NJ). The sen-sitivity peak for the PSPMT is located at 420 nm. ThePSPMT is mounted inside a light tight fixture calledthe interface, as shown in Figure 1. The fibers are cou-pled to the PSPMT with a spring-loaded snap-on connec-tor to ensure good mechanical coupling. The fibers arearranged in a hexagonal pattern on the PSPMT front,this to maximize the separation between the fibers andto minimize optical crosstalk, see Figure 5.

A resistive dividing network takes the signal from thePSPMT and produces the position signals X�, X�, Y�,and Y�. The signals are then preamplified and shapedto produce Gaussian-shaped outputs. An additional fastsum signal is generated to produce a timing signal for

Figure 4. Cross-section of the multicompartmental catheter.The peripheral chambers house six optical fibers and thecenter compartment is for the guidewire.

134 Molecular Imaging and Biology, Volume 6, Number 3

Figure 5. Picture of the interface connection. On the left, thefibers, arranged in a hexagonal pattern, are visible as white dotson the end piece. On the right, the faceplate of the PSPMTis visible as a reflection inside of the interface.

the analog-to-digital converter (ADC). The four positionsignals are digitized by a quad ADC in the PC. UsingAnger logic and a look-up table, the signals from the sixscintillation fibers are decoded into a linear image thatis displayed on the computer screen. The typical framerate is about one second. In addition to the dynamicprofile, a tone, with its frequency dependent on theinstantaneous count rate, is generated in the audiooutput.

Results

The detector’s efficiency is strongly dependent on itsability to detect low energy betas. The emitted betas areattenuated by the tissue and the blood, as well asthe catheter tubing, before reaching the detector. Whilethe beta spectrum is a continuum, only the betas withan energy higher than the cut-off energy will be counted.The cut-off energy is set to be just above the electronicnoise level. The catheter thickness is minimized in orderto minimize the attenuation of the betas, while stillmaintaining a light tight compartment for the detectors.

To test the detector efficiency as a function of maxi-mum beta energy, the device was tested with severalbeta sources. Some interesting beta sources are listed inTable 1. Technitium-99m, 18F, 204Tl, and 32P are isotopes

that are accessible in the Biological Imaging lab or sur-rounding labs at UCLA. Thallium-204 has a maximumbeta energy of 765 keV, which is close to 18F’s 633 keV,which is important if 2-deoxy-2-[18F]fluoro-D-glucose(FDG) is used as a tracer. Due to 204Tl’s long half-life andits ease of handling (since it exists as a sealed source inour lab), most of the experiments were performed withthis isotope and it is also used as a gold standard inthe analysis. In addition, three nonstandard isotopeshave been added to Table 1 that have potential as betaemitters for this detector system. They are available asthe decay products of long-lived parents and are avail-able commercially from isotope generator systems.These are 113mIn, 188Re, and 68Ga. Each of these ele-ments has a history of use in nuclear medicine.6 Indium-111 is currently employed in a wide range of products,67Ga has been employed for years for various types ofnuclear medicine studies, and 188Re is a chemical analogof 99mTc.

Beta Penetration

In order to determine if FDG might be a good tracerfor this probe or if a higher energy beta emitter isrequired, initial penetration tests were performed with204Tl (close to 18F in energy) and a higher energy betaemitter, 32P. A point source was placed at a fixed distancefrom the detector and the beta energy spectrum wascollected using a multichannel analyzer (MCA) con-nected to a CAMAC unit and a PC. Thin plastic films,each 51 microns in thickness, were inserted betweenthe source and the detector and the resulting energyspectra were collected. The penetration tests were re-peated with various film thicknesses for both pointsources. The detected beta energy spectra for 204Tl and32P are displayed in Figure 6. The 204Tl beta spectrumshows that there is a reasonable signal through aboutone mm of plastic film. For 32P, the signal is reasonablethrough about three mm of plastic film.

Beta Versus Gamma Sensitivity

A tracer emitting positrons, such as FDG, will also pro-duce 511 keV gammas. One of the requirements for

Table 1. β emitters

Isotope Half-life bmax or EC energy b abundance(daughter/parent) (daughter/parent) (isotope or daughter) (isotope or daughter)

99mTc/99Mo 6.03 hr/66.7 hr 119 & 138 keV, EC 10.3%18F 109 min 633 keV, βmax 97%204Tl 3.8 years 765 keV, βmax 98%32P 14.3 days 1710 keV, βmax 100%188Re/188W 17 hr/69 days 2130, 1980 keV, βmax 70%/26%68Ga/68Ge 68.3 min/287 days 1898 keV, βmax 89%113mIn/113Sn 99.4 min/115 days 360, 390 keV, EC 36%131I 8.06 days 606 keV, βmax 90%

Vulnerable Plaque Detection/ Janecek et al. 135

12

10

8

6

4

2

0100 150 200

Channel (Energy)

Lig

ht o

utpu

t (cp

s)

300250

Plastic film thickness

0 mm

204Tl

0.41 mm0.82 mm1.23 mm2.05 mm

350

A

12

14

16

10

8

6

4

2

0100 150 200

Channel (Energy)

Lig

ht o

utpu

t (cp

s)

300250

Plastic film thickness

0 mm

32P

0.41 mm0.82 mm1.23 mm1.64 mm2.46 mm3.69 mm

350

B

Figure 6. (A) Thallium-204 beta-energy spectra for variousfilm thicknesses. (B) Phosphorus-32 beta-energy spectra forvarious plastic film thicknesses.

in vivo beta imaging, especially in close proximity tothe heart, is the suppression of the associated gammabackground. To measure the sensitivity of gammasversus betas for the detector, measurements were per-formed with a low-energy emitting gamma source, 99mTc,and the results were compared with 204Tl. The beta-to-gamma sensitivity was measured with the point sourcesone mm from the detector. The source strengths wereselected to be low to avoid dead-time problems. Thegamma emitter, 99mTc, was evaporated onto a piece ofpaper, making sure that the 99mTc spot would be smallenough to be considered a point source. The emittedconversion electrons from 99mTc have insufficientenergy (10% at 120 keV7) to penetrate the catheter

tubing. Thallium-204 was used as a beta emitter. Whencomparing the results from the two measurements, thesource strengths were normalized. Integrating the sig-nals from the cut-off energy, which was set to be justabove the noise level, the beta-to-gamma sensitivity wascalculated to be �1100:1, see Figure 7.

The gamma-to-beta sensitivity was also measured with18F, a positron emitter. The positron interacts as a posi-tive electron until it is slowed and captured by a negativeelectron, at which time the positron and electron anni-hilate and emit two 511-keV gamma rays. A single scintil-lation fiber was placed in a 2-cm3 vial containing 18F.A second measurement was performed with the scintil-lating fiber outside of the vial, shielded with 2 mm ofpolytetrafluoroethylene (PTFE) to eliminate beta pene-tration. Both measurements were performed with thesame electronic set-up as described in the beta penetra-tion section. From the vial measurement, a beta �gamma count rate was received, and from the externalmeasurement a gamma-only count rate was received.With these two count rates, the beta-to-gamma sensitivitycould be calculated:

([(total events�gammas)/gammas])to 11.1 � 1.1 EQ1

This number is not a really useful result because thescintillating-fiber detector will work in a completely dif-ferent environment. The result can be greatly alteredby having a larger or smaller vial, thereby changing thegeometry of the experiment. The experiment shouldin the future, therefore, be performed with a realistic

Figure 7. Detection efficiency for 204Tl (betas) and 99mTc(gammas). Notice that the 99mTc spectrum is magnified by afactor of 100.

136 Molecular Imaging and Biology, Volume 6, Number 3

artery phantom and with isotope concentrations ofmore realistic values. The realistic values have not yetbeen measured in man or animals. The conclusionto be drawn here is that the best result will be seen withradiopharmaceuticals that are pure beta emitters or thatdo not extract into the myocardium.

Spatial Resolution

The spatial resolution was measured by passing a slit-collimated 204Tl-point source in small increments overthe length of the detector. The spatial resolution fromthis measurement was estimated to be 6.69 � 0.24mm, which is essentially equal to the scintillating fiberlengths, see Figure 8.

The peak intensity for each scintillating fiber in Figure8 has been normalized. Without this normalization, thescintillating fiber at the catheter side that faces the pointsource has a much higher count rate compared withthe scintillating fiber that is on the furthest side of thecatheter. The increased distance to the radioactivesource and, even more importantly, the shielding bythe fibers that lie in between the scintillating fiberand the radioactive source (along with the guidewire)result in a lower count rate. For the furthest distantscintillating fibers, only the highest-energy emitted betaswill be detected, giving a low count rate, as well as alow beta spectrum. In the spatial resolution experiment,with the cut-off energy set to the electronic noise level,the count rates for the scintillating fibers varied at mostby a factor of five between the closest and the mostdistant scintillating fiber. This intensity ratio can belowered by: 1) using a low-noise PSPMT instead of thegeneral purpose PSPMT that was used in this probe; 2)using a higher-energy beta emitter; or 3) simply rotatingthe probe during the acquisition.

Sensitivity

The sensitivity was measured with a 204Tl-point sourceone mm from the detector. The count rate was mea-sured at the longitudinal center of the scintillating fibersand the six measurements were averaged. The sensitivitywas calculated to 404 � 43 cps/µCi.

Flood Image

A flood image was acquired by moving a 204Tl sourcepast the length of detector at a constant distance andat a very slow speed. The data were accumulated intoone image during the acquisition. From the acquiredflood image, a look-up table could be created. This look-up table is used in the image reconstruction routinementioned in the electronics section. The acquiredflood image is displayed in Figure 9A. Each pixel iseasily resolved from the next one, and the correspond-ing look-up table is displayed in Figure 9B.

Fiber Flexibility

A single 0.5-mm diameter optical fiber was tested forflexibility. A scintillation fiber at the distal end of theoptical fiber was irradiated with a 204Tl source. The lightproduced was guided through the optical fiber to aPMT, where the count rate was measured. The opticalfiber was twisted into three loops. After each measure-ment, the diameter of the twisted loops was decreasedand a new count rate measurement was performed. Thesignal strength was unchanged down to about a 1.5-cmradius of curvation. At this point, the fiber also startedcurling up. At 1-cm radius, the fiber became perma-nently distorted, and about half of the light was lost.

100

80

60

40

20

00 10 20

X - position (mm)

Nor

mal

ized

inte

nsity

30 40

Fiber 1

Fiber 2

Fiber 3

Fiber 4

Fiber 5

Fiber 6

Figure 8. Spatial resolution of probe measured with a collimated 204Tl-point source. The peak intensity for each scintillatingfiber has been normalized.

Vulnerable Plaque Detection/ Janecek et al. 137

Figure 9. (A) Flood image of detector exposed to 204Tl. (B)Corresponding look-up table.

Catheter Diameter Optimization

Three catheter probes with 0.5-mm diameter fibers werefabricated. Due to assembly problems with the fragilescintillating-optical fiber bonds (where the bonds wouldoccasionally break as the 1.5-m long fibers were in-serted in the very narrow catheter housing), the firstcatheter-probe had an outer diameter of 2.5 mm. Prof-iting from the experience of assembling the first probe,the diameter of the second catheter-probe was reduced.The third (and final) 0.5-mm fiber based catheter probehad an outer diameter of 1.9 mm.

In order to reduce the diameter further, a catheterprobe with 300-µm diameter fibers was fabricated. Theprobe was fabricated in the same manner as the 0.5-mm probe. The outer diameter of this catheter probewas measured to be 1.5 mm.

To compare the light yield between the 300-µm andthe 0.5-mm diameter scintillating fibers, the scintillatingfibers were fused to 1.5-m long optical fibers (300-µmdiameter scintillating fibers to 300-µm diameter clearoptical fibers, and 0.5-mm diameter scintillating fibersto 0.5-mm diameter clear optical fibers). The scintil-lating fibers were exposed to a 204Tl-point source andthe light was collected by a PMT at the distal end of the

optical fiber. The light yield for the 300-µm diameterscintillating fibers was measured to be 35.3% � 4.6% ofthe 0.5-mm scintillating fiber light yield. These resultsare consistent with theoretical considerations, as thedeposited energy should decrease with the square ofthe diameter,

Sensitivity �1/d2 EQ2

This measurement with the smaller diameter probe indi-cates that reducing the diameter of the imaging probefrom its current 0.5-mm configuration, without higherscintillation photon yields from the scintillating fibers,will severely reduce the sensitivity.

Discussion and Conclusions

Currently, no pure beta-emitting tracer exists that willlabel vulnerable plaque. The penetration measure-ments were made to guide us in the development ofsuch a tracer. The penetration measurements show usthat a higher beta energy emitter is preferable as a tracer.Even though the 204Tl betas give a signal, the detectionis low when the source is on one side of the detectorwhile the scintillation fiber is on the opposite side. Ina non-normalized spatial resolution image, the differ-ence in counts per fiber between the closest scintillatingfiber and a scintillating fiber on the opposite side inthe catheter (relative to the point source) can be ashigh as a factor of five. Using a higher beta-energy tracerwould give a strong signal while still allowing a preciselocalization of the tracer. Using an internal conversion

Figure 10. Normalized beta-emission spectra for 18F and 113mIn.The 113mIn spectrum consists of three peaks (363.7 keV, 387.7keV, and 391.0 keV) convoluted with a Gaussian-line spreadfunction (FWHM � 9.5%). (The FWHM was calculated assum-ing 100% energy deposited in the scintillating fiber, 8000photons/MeV created in the scintillating fiber, and a 20%PMT quantum efficiency.) The area under the 113mIn spec-trum is 35.8% compared with the 18F’s area to account forthe internal conversion probability of 113mIn.

138 Molecular Imaging and Biology, Volume 6, Number 3

electron emitter might also work. A potential radioactivenuclide is 113mIn. In Figure 10, an 113mIn-emission spec-trum is compared to a 18F-emission spectrum. As onlythe higher energy betas will be registered as eventsin the detector in a typical beta-emission spectrum andthe 113mIn spectrum has only the higher energy elec-trons, its physical properties would be ideal as a vulnera-ble plaque tracer.

The original imaging probe was manufactured with∼7-mm long scintillating fibers, giving a resolution ofthe same order. The resolution of the catheter systemcan be increased or decreased by changing the scin-tillation fiber lengths. The FOV will also changeaccordingly.

The system described here is a first pass at addres-sing the problem of detecting vulnerable plaque di-rectly. Currently, there is limited success with 18F anda number of 99mTc-labeled compounds. As we haveshown, the sensitivity of a small beta detector for99mTc gammas is prohibitively low, and the backgroundfrom activity in the heart may be a serious drawbackfor 18F. Although 113mIn is not FDA approved, 111Inis familiar to the chemists that develop compoundsfor nuclear medicine and it may be possible to usethat knowledge with 113mIn. The 100-minute half-life of113mIn and the fact that is available from a generatorwith a 115-day half-life makes it very attractive for thisapplication. As usual, the success of a nuclear medicineprocedure is in the hands of the chemists.

The authors would like to thank Mr. Richard Luu for hisexpert assistance on various parts of the technical work.This work was supported in part by NIH SBIR R43 HL065837.

References1. American Heart Association. Heart disease and stroke sta-

tistics—2003 Update. Dallas, Texas: American Heart Associ-ation, 2002.

2. Falk, E. Pathological characterization. In: Chairs, ed. 72nd

Scientific Sessions of the American Heart Association.Atlanta, GA: American Heart Association, 1999.

3. Langer, A.; Fisher, M.; Califf, R.M.; et al. Higher rates ofcoronary angiography and revascularization following myo-cardial infarction may be associated with greater survival inthe United States than in Canada. The CARS Investigators(Coumadin/Aspirin Reinfarction Study). Can. J. Cardiol.15:1095–1102; 1999.

4. Naghavi, M.; Madjid, M.; Khan, M.R.; et al. New develop-ments in the detection of vulnerable plaque. Curr. Atheros-cler. Rep. 3:125–135; 2001.

5. Mari, C.; Nedelman, M.; Blankenberg, F.; et al. Detectionof active atheroma in a rabbit model: evaluation of 8 radio-tracers. J. Nucl. Med. 42:45P; 2001.

6. Wagner, H.N. Principles of nuclear medicine. Philadel-phia, PA: Saunders; 1995.

7. Dillman, L.T.; Von der Lage, F.C. Radionuclide decayschemes and nuclear parameters for use in radiation-doseestimation. J. Nucl. Med. (Pamphlet no. 10):62. 1975.