Cardiovascular disease is the leading cause of morbidity and mortality in developed countries, with coronary artery disease (CAD) being the largest contributor . Assessment of myocardial perfusion is an important step for determining the prognosis of patients with CAD. Rubidium-82 (82Rb) chloride is a positron-emitting myocardial perfusion tracer that has gained popularity as positron emission tomography (PET) scanners have become more available, particularly since the combination of computed tomography (CT) and PET scanners in the same gantry [2–4]. Although this tracer has a very short half-life (76 s), it can be produced in-house using a generator  rather than a cyclotron. The generator has a 25-day half-life  and is infused using a computerized infusion system . Previous studies have shown the ability of 82Rb PET myocardial perfusion imaging to detect serial changes in absolute perfusion and quantitatively measure regional myocardial blood flow (MBF) in ml/min/g [8–12]. Quantitative assessment of myocardial perfusion can be an asset, especially in patients with diffuse multivessel CAD [6,9], where techniques that rely on relative differences in myocardial perfusion may not be adequate . Quantification of flow is also useful in defining early microvascular disease without or before the development of obstructive coronary stenoses [14,15].
Many studies investigating 82Rb perfusion imaging have been performed using two-dimensional (2D) imaging. With the development of improved detector materials and newer scanners, some with only three-dimensional (3D) capabilities, the efficacy of 3D imaging for quantification of myocardial perfusion has gained interest [8,16–18]. Three-dimensional imaging provides increased sensitivity and may result in decreased costs and reduced patient dose in comparison with 2D imaging owing to a reduction in the required injected activity of radiotracer [16,19]. If a substantially lower injected activity of radiotracer is required, a 82Rb generator could either be loaded with considerably less 82Sr or used for substantially longer . Further, owing to the short half-life, 2D images obtained with 82Rb tend to be relatively count poor [20,21]. An alternative to increase the injected dose to overcome this issue is to acquire the perfusion data in 3D rather than 2D mode.
The purpose of this study, therefore, was to compare quantitative MBF results from images obtained in both 3D and 2D mode on a GE Discovery LS PET/CT hybrid scanner (GE Healthcare, Milwaukee, Wisconsin, USA) under a variety of flow states.
Animal preparation and surgery
This study was approved by the Animal Use Subcommittee of the Canadian Council on Animal Care at the University of Western Ontario (ethics approval ♯2003-108-02). A total of 12 mongrel dogs were used for this study. Anesthesia was induced using 1% propofol injected intravenously and maintained with 1.5–2% isoflurane. A left lateral thoracotomy was performed in the fourth intercostal space. A region of the left anterior descending (LAD) coronary artery was dissected free of the myocardium and a snare ligature was placed in this region. The snare was kept loose and externalized through an opening in the chest wall and the incision was closed. The animals were then transported to the clinical PET/CT hybrid scanner.
Creation of stunned and infarcted myocardium
This study employed our canine model of stunned  and infarcted [23,24] myocardium, and images were obtained both at rest and during dobutamine hyperemia (stress). Stunned myocardium was created by occluding the LAD coronary artery for 15 min, using the snare ligature, followed by reperfusion [25,26]. Resting blood flow measurements were acquired 45 min after the release of the occlusion (stunned group, n=12), and blood flow during dobutamine hyperemia was measured a further 30 min later (following a step-wise increase of dobutamine up to 30 μg/kg/min over 8 min, image acquisition initiated 3 min after the start of 30 μg/kg/min dose, same dobutamine dose was used for both 3D and 2D imaging). Pharmacological hyperemia was maintained throughout image acquisition except when not tolerated by the animals. Two hours after reperfusion, the LAD was again occluded and either released 2 h later (acute reperfused, n=6) or permanently occluded (acute occluded, n=6). Resting blood flow measurements were acquired 2.5 h after the release of the occlusion (acute reperfused) or 4.5 h after permanent occlusion (acute occluded). Blood flow was measured again during dobutamine (30 μg/kg/min) hyperemia 30 min later. The animals were recovered and then brought back 8 weeks later for rest, dobutamine stress, and repeat rest imaging to measure blood flow in chronic myocardial infarction (chronic reperfused, n=6; chronic occluded, n=6). A schematic representation of this protocol is shown in Fig. 1.
82Rb PET imaging and regional myocardial blood flow calculation
Three-dimensional and 2D PET imaging was performed on a GE Discovery LS PET/CT hybrid scanner. A 10 min 68Ge static transmission scan was acquired for attenuation correction.
Dynamic emission imaging was performed with 18 frames acquired over 10.5 min (12×10 s, 3×30 s, 1×60 s, 1×120 s, 1×240 s). 82Rb was infused over 30 s intravenously using a custom-built infusion system [6,27] and the scan was started when counts were first observed in the field of view. On average, 378 MBq was infused for 3D image acquisitions and 617 MBq for 2D image acquisitions. First, 3D imaging was performed followed by the 2D image acquisition; the infusion of 82Rb for the 2D acquisition was started as soon as the scan setup was complete (approximately 12 min between the start of each acquisition). Image sets were reconstructed, using the iterative reconstruction software available on the commercial system (GE Medical Systems, software version 16.01) including attenuation and scatter correction, into a 128×128 array with a 5.5 mm full width at half maximum Gaussian postfilter, pixel size of 1.17 mm and 35 slices with a slice thickness of 4.25 mm. The 2D data were reconstructed using ordered-subset expectation maximization  and the 3D data were reconstructed using Fourier rebinning plus ordered-subset expectation maximization .
An automatic program reoriented the dynamic 82Rb images into the short axis plane  to sample the left ventricular myocardium. The left ventricular and atrial cavities were identified automatically to obtain a median arterial input function, used for compartmental modeling and correction of partial volume effects, as follows. The input function was computed as the median of three time–activity curves measured in small cylindrical image regions placed automatically in the center of the left ventricle cavity, base, and left atrium . In each of 496 mid myocardial sectors, dynamic polar maps containing myocardial time–activity curves were generated. For each time–activity curve, a 1-compartment tracer kinetic model was used to quantify regional MBF in ml/min/g [19,31,32] using an extraction fraction correction of the uptake rate . The extraction fraction was measured versus ammonia in healthy volunteers and validated in patients with CAD. This function was then used to estimate blood flow using the standard relationship: rubidium uptake rate, K1 = MBF×e(MBF). The blood flow polar maps were analysed using a standard 17-segment model . Segmental blood flow was then classified as either region at-risk (RAR) or remote tissue. Each segmental blood flow value (17 segments for each image acquisition) was then included in the analysis, resulting in the numbers seen in Table 1.
In the reperfused animals, the RAR was determined by measuring blood flow at rest during the 2-hour occlusion. In the occluded animals, the RAR was determined using the resting blood flow measurements acquired 4.5 h postocclusion. Tissue was defined as at-risk if average segmental blood flow was ≤0.4 ml/min/g. All other segmental blood flow values were classified as remote. As MBF is dependent on heart rate and blood pressure [34,35], the rate–pressure product (RPP) was determined and the values at the time of 3D and 2D rest and stress acquisitions were compared when data were available. However, blood pressure measurements were not always available and thus, for the overall analysis, individual rest and stress flow values were not normalized to their respective RPPs.
In a subset of animals, microspheres were also used to measure blood flow concurrently with 3D blood flow measurements: 21 rest and 10 stress measurements were compared with 3D blood flow measurements. A limited number of microspheres were available (46Sc, 51Cr, 57Co, 85Sr, 95Nb, 103Ru, 141Ce, or 153Gd), and thus only comparisons with 3D blood flow measurements were possible. Microsphere injection coincided with the beginning of the 82Rb infusion for 3D imaging.
The total numbers of animals and segments included in each experimental and imaging group are shown in Fig. 1 and Table 1, respectively.
In the stunning group, one 2D rest study was not completed, leaving 11 with complete rest data. In the same group, dobutamine was terminated early in four animals as a result of an increase in heart rate greater than 200 bpm, leaving eight animals with complete stress data. These same four animals did not tolerate dobutamine stress during acute reperfusion, leaving two of six animals with complete stress data. One acute occluded stress study was excluded owing to imaging artifact. All 12 animals survived the surgery and acute imaging session.
In the chronic reperfused group, one animal was euthanized 2 days after the surgery and one died suddenly before image acquisition, leaving four animals available for imaging. Three animals died after the first chronic rest measurement (one chronic reperfused and two chronic occluded). In the chronic occluded group, the repeat rest data were not available in two other animals for technical reasons.
Statistical analysis was performed using GraphPad Prism (San Diego, California, USA). Spearman correlation analysis was used to compare the 3D and 2D results. Correlations were considered strong if 1.0>ρ>0.7, moderate if 0.7>ρ>0.3, weak if 0.3>ρ>0.1. Matched paired t-tests (Wilcoxon rank tests when the data were not normally distributed) were employed to determine whether there were any significant differences between the 3D and 2D parameters (e.g. average values). Analysis of variance with Dunn's multiple comparison post-hoc testing was performed on segmental blood flow results in each tissue group for comparison of flow results from rest to hyperemia with either the 2D or 3D technique. Bland–Altman plots of the average segmental values versus the difference between techniques were used to evaluate bias and dispersion. Average microsphere and 3D blood flow results in the remote and RAR tissue were analyzed for significance using a Mann–Whitney t-test. Average values are reported as mean±standard error of the mean, NS = not significantly different (P>0.05).
A total of 57 comparative image sets from all animals (37 rest and 20 stress image sets) with 17 segments (n=969) were acquired, as shown in Table 1. The average 3D versus 2D results for rest and stress values were 0.63±0.01 versus 0.68±0.01 (P<0.001) and 4.51±0.07 versus 4.04±0.08 ml/min/g (P<0.001), respectively. The RPP was 5497±318.9 and 5251±268.4 (NS; n=22) at the time of resting 3D and 2D blood flow measurements, respectively. For the 3D and 2D stress blood flow measurements, the RPP was 13011±738 and 9874±692 (P<0.05, n=11), respectively. A significant and strong correlation with the RPP was present for both the 3D and 2D (ρ=0.87, P<0.0001 for both) blood flow values.
Figure 2 shows the correlation results for segments at rest (i) as well as those during dobutamine stress (ii) and all values (rest and stress) (iii). Strong to moderate, significant correlations were found for each (all, ρ=0.95; rest, ρ=0.88; stress, ρ=0.62; P<0.001). Figure 3 shows the average blood flow results in each tissue group at rest and during dobutamine hyperemia. Each tissue group had a significant increase in blood flow from rest to stress that was observed with both 3D and 2D imaging.
Figure 4 shows the Bland–Altman plots for the 3D versus 2D rest and stress data, which showed a negative bias for the resting data (−8%) and a positive bias for the stress data (+11%). When rest and stress data were combined, a 7% bias was present. Figure 5 shows representative segmental flow values obtained in one dog at rest and during stress using both 3D and 2D image acquisitions. Although some fluctuation is noted between the two techniques, very good regional agreement is present overall. Overall, both the techniques identified the same at-risk regions , that is, segmental flow values less than 0.4 ml/min/g during the LAD occlusion, although the 2D technique tended to have higher segmental blood flow values in the remote regions (supplied by normal coronary arteries) and lower values in the at-risk regions (supplied by the LAD) compared with the 3D technique during dobutamine stress.
In the subset of animals with microsphere flow results, average resting microsphere and 3D flow values in the remote tissue were 0.66±0.02 and 0.69±0.02 ml/min/g, respectively (P=NS). In the RAR tissue, average resting microsphere and 3D flow values were 0.50±0.03 and 0.39±0.02 ml/min/g (P<0.05). Average hyperemic microsphere and 3D flow values in the remote tissue were 2.44±0.16 and 2.89±0.17 ml/min/g (P=NS). In the RAR tissue, average hyperemic microsphere and 3D flow values were 1.41±0.14 and 1.70±0.21 ml/min/g (P=NS).
This study compared regional MBF results obtained in the setting of stunned and infarcted canine myocardium using 3D and 2D 82Rb PET imaging. Although some small differences were noted in the average values, the results were highly correlated at rest and during dobutamine stress. Further, blood flow results in stunned and infarcted myocardium, whether reperfused, occluded, acute or chronic, were similar and showed comparable responses from rest to dobutamine hyperemia. These results suggest that quantitative 3D PET perfusion imaging using 82Rb is feasible, providing greater access for clinical centers and reducing cost and patient radiation dose owing to a decrease in the injected activity of radiotracer required.
The positive bias found in the stress values is most likely a result of changes in blood flow owing to the continued presence of dobutamine over time. The 3D and 2D acquisitions were not obtained simultaneously and physiological markers cannot be assumed to be steady for both imaging sessions during dobutamine hyperemia. The drop in the RPP from 3D to 2D acquisition during dobutamine infusion likely caused the observed decrease in segmental blood flow. In fact, in the subset of data where correction of the blood flow results by the RPP was possible, a negative bias in the 3D stress data (−13%) was found. In addition, the negative bias noted in the resting data was maintained (−5%) when corrected by the RPP. This negative bias in the rest and stress data may be indicative of residual scatter present in the 3D acquisitions arterial input function resulting from incomplete scatter correction. That is, scatter from the surrounding myocardium into the left ventricle cavity region would increase the apparent arterial blood concentration at late time points (more so during stress than at rest), causing underestimations of flow that are greater at stress than at rest, as observed in the RPP-corrected data.
Previous studies have investigated the feasibility of 3D image acquisition compared with 2D for myocardial imaging using 82Rb [16,18,20]. Although one study concluded that image quality of 3D and 2D images were comparable, the other study suggested that 82Rb studies should only be performed in 2D. The conclusions of these studies were primarily based on contrast values and image quality from 3D and 2D static images. The results presented here provide another dimension to this topic as quantitative blood flow measurements were determined in a variety of tissue types. Further, the bias found in this study at rest and stress is of somewhat smaller magnitude compared with that found in a previous study using the same 13N-ammonia data with different tracer kinetic models, where the average bias was 13% . Therefore, the bias in the rest and stress blood flow values from this study seem to be within the expected range of variability when different techniques are investigated. In addition, 3D blood flow values were found to not be significantly different from microsphere flow results in the remote tissue at rest and during hyperemia or during hyperemia in the RAR tissue, further indicating the reliability of 3D blood flow measurements using 82Rb. The difference in the resting microsphere and 3D RAR blood flow values, while statistically significant, is small in magnitude and probably not of clinical importance.
Registration of the transmission/emission images was assured because (i) the animals are anesthetized and therefore do not move between the transmission and emission image acquisitions, and (ii) the same detector array is used for both image sets. Major sources of motion that might lead to misregistration are physiological – respiratory motion in particular. However, the use of radioisotope transmission imaging averages the acquisition over the respiratory cycle in the same fashion as the emission acquisition, minimizing this difference. A further potential for error still exists if the breathing cycle were to change between the emission and transmission scans. However, as the dogs were ventilated, the respiration is held consistent and is thus not a concern.
The canine model was chosen for these studies given that this model creates stable myocardial stunning and infarction and that repeated blood flow measurements could be obtained over 8 weeks in a controlled manner. However, we realize that the amount of photon scatter is dependent on patient habitus, which is obviously smaller in canines than in patients. This may be of particular importance in 3D imaging as scatter detection is also increased versus 2D . Thus, patient studies should also be performed in a similar manner as these studies to evaluate the ability of 3D PET to accurately quantify blood flow in various tissue states. Nonetheless, these results provide favorable evidence for 3D dynamic imaging in normal, stunned, and infarcted myocardium.
The design of this study purposefully involved acquiring all 3D images before the 2D images. By consistently acquiring the 3D data before the 2D data, we could minimize variance within the data (e.g. consistent length of anesthetic for each 3D or 2D acquisition in each group) and be able to explain any systematic biases that existed between the two techniques, as we did with the RPP. Further, our project design allowed us to inject microspheres concurrently with the injection of 82Rb for the 3D measurements, providing a second method to validate the 3D results as well as comparison with the 2D results. While there is the potential for increased background activity at the time of 2D imaging, the amount of activity remaining in the tissue (less than 3 kBq) would be negligible compared with the activity injected for the 2D image acquisition, which was almost two times greater than that originally used for the 3D measurement. Furthermore, randomization would have degraded the correlations between the 3D and 2D techniques and may have concluded that 3D is not sufficiently similar to 2D imaging. Finally, given the nature of animal studies, we could not know from the outset how many animals would survive and might have resulted with uneven group populations.
Three-dimensional dynamic 82Rb imaging for quantification of regional MBF seems to yield similar results to 2D dynamic imaging. As such, 3D 82Rb dynamic imaging can be performed for determination of regional MBF in ml/min/g while potentially reducing operational costs and patient dose.
The authors acknowledge the financial assistance of the Nuclear Medicine Department at SJHC, London (Dr Jean-Luc Urbain). Lela Deans, Huafu Kong, Jennifer Hadway, Dominique Ouimet, and Terrie Ann Campbell provided experimental and animal assistance and expertise. The authors thank Dr Janice DeMoor and Dr Jean De Serres for editorial assistance. This study was funded by Canadian Institutes of Health Research (grant ♯R-04-368), Ontario Consortium of Cardiac Imaging (OCCI), which is supported by the Ontario Research and Development Challenge Fund, DRAXIMAGE, Kirkland, Quebec, Canada. Dr Lekx was supported by a co-institution, co-supervisor fellowship from OCCI (FP, RB, RdK).
1. Heart and Stroke Just the Facts, 2002/2003 edition. Heart and Stroke Foundation of Ontario. 2006.
2. Von Schulthess GK. Positron emission tomography versus positron emission tomography/computed tomography: from unclear to new-clear medicine. Mol Imaging Biol 2004; 6:183–187.
3. Von Schulthess GK, Steinert HC, Hany TF. Integrated PET/CT: current applications and future directions. Radiology 2006; 238:405–422.
4. Schwaiger M, Ziegler S, Nekolla SG. PET/CT: challenge for nuclear cardiology. J Nucl Med 2005; 46:1664–1678.
5. Alvarez-Diez TM, dekemp R, Beanlands RSB, Vincent J. Manufacture of strontium-82/rubidium-82 generators and quality control of rubidium-82 chloride for myocardial perfusion imaging in patients using positron emission tomography. Appl Radiat Isot 1999; 50:1015–1023.
6. Machac J. Cardiac positron emission tomography imaging. Semin Nucl Med 2005; 35:17–36.
7. Epstein J, Benelfassi A, Beanlands RSB, deKemp RA. An 82Rb infusion system for quantitative perfusion imaging with 3D PET. Appl Radiat Isot 2004; 60:921–927.
8. Dekemp RA, Ruddy TD, Hewitt T, Dalipaj MM, Beanlands RS. Detection of serial changes in absolute myocardial perfusion with 82Rb PET. J Nucl Med 2000; 41:1426–1435.
9. Parkash R, deKemp RA, Ruddy TD, Kitsikis A, Hart R, Beauchesne L, et al. Potential utility of rubidium-82 PET quantification in patients with 3-vessel coronary artery disease. J Nucl Cardiol 2004; 11:440–449.
10. Yoshida K, Mullani N, Gould KL. Coronary flow and flow reserve by PET simplified for clinical applications using rubidium-82 or nitrogen-13-ammonia. J Nucl Med 1996; 37:1701–1712.
11. Lin JW, Sciacca RR, Chou RL, Laine AF, Bergmann SR. Quantification of myocardial perfusion in human subjects using 82Rb and wavelet-based noise reduction. J Nucl Med 2001; 42:201–208.
12. El Fakhri G, Sitek A, Guerin B, Kijewski MF, Di Carli MF, Moore SC. Quantitative dynamic cardiac 82Rb PET using generalized factor and compartment analyses. J Nucl Med 2005; 46:1264–1271.
13. Camici PG, Rimoldi OE. Myocardial blood flow in patients with hibernating myocardium. Cardiovasc Res 2003; 57:302–311.
14. Dayanikli F, Grambow D, Muzik O, Mosca L, Rubenfire M, Schwaiger M. Early detection of abnormal coronary flow reserve in asymptomatic men at high risk for coronary artery disease using positron emission tomography. Circulation 1994; 90:808–817.
15. Geltman EM, Henes CG, Senneff MJ, Sobel BE, Bergmann SR. Increased myocardial perfusion at rest and diminished perfusion reserve in patients with angina and angiographically normal coronary arteries. J Am Coll Cardiol 1990; 16:586–595.
16. Votaw JR, White M. Comparison of 2-dimensional and 3-dimensional cardiac 82Rb PET studies. J Nucl Med 2001; 42:701–706.
17. Lewellen TK, Kohlmyer SG, Miyaoka RS, Kaplan MS. Investigation of the performance of the General Electric Advance positron emission tomograph in 3-D mode. IEEE Trans Nucl Sci 1996; 43:2199–2206.
18. Knesaurek K, Machac J, Krynyckyi BR, Almeida OD. Comparison of 2-dimensional and 3-dimensional 82Rb myocardial perfusion PET imaging. J Nucl Med 2003; 44:1350–1356.
19. Dekemp RA, Yoshinaga K, Beanlands RSB. Will 3-dimensional PET-CT enable the routine quantification of myocardial blood flow? J Nucl Cardiol 2007; 14:380–397.
20. Knešaurek K, Machac J, Kim JH. Comparison of 2D, 3D high dose and 3D low dose gated myocardial 82
Rb PET imaging. BMC Nucl Med 2007; 7:4.
21. Knešaurek K, Machac J, Zhang Z. Repeatability of regional myocardial blood flow calculation in Rb-82 PET imaging. BMC Nucl Med 2009; 9:2.
22. Thornhill RE, Prato FS, Pereira RS, Wisenberg G, Sykes J. Examining a canine model of stunned myocardium with Gd-DTPA-enhanced MRI. Magn Reson Med 2001; 45:864–871.
23. Pereira RS, Prato FS, Wisenberg G, Sykes J. The determination of myocardial viability using Gd-DTPA in a canine model of acute myocardial ischemia and reperfusion. Magn Reson Med 1996; 36:684–693.
24. Pereira RS, Prato FS, Lekx KS, Sykes J, Wisenberg G. Contrast-enhanced MRI for the assessment of myocardial viability after permanent coronary artery occlusion. Magn Reson Med 2000; 44:309–316.
25. Dourado PM, Tsutsui JM, Mathias W Jr, Andrade JL, da Luz PL, Chagas AC. Evaluation of stunned and infarcted canine myocardium by real time myocardial contrast echocardiography. Braz J Med Biol Res 2003; 36:1501–1509.
26. Juergens KU, Reimer P, Weber TP, Tombach B, Bremer C, Renger B, et al. Cine and tagged magnetic resonance imaging in short-term stunned versus necrotic myocardium. Int J Cardiovasc Imaging 2005; 21:271–282.
27. Klein R, Adler A, Beanlands RS, deKemp RA. Precision-controlled elution of a 82Sr/82Rb generator for cardiac perfusion imaging with positron emission tomography. Phys Med Biol 2007; 52:659–673.
28. Hudson HM, Larkin RS. Accelerated image reconstruction using ordered subsets of projection data. IEEE Trans Med Imag 1994; 13:601–609.
29. Liu X, Comtat C, Michel C, Kinahan P, Defrise M, Townsend D. Comparison of 3-D reconstruction with 3D-OSEM and with FORE+OSEM for PET. IEEE Trans Med Imaging 2001; 20:804–814.
30. Dekemp RA, Nahmias C. Automated determination of the left ventricular long axis in cardiac positron tomography. Physiol Meas 1996; 17:95–108.
31. Lortie M, Beanlands RS, Yoshinaga K, Klein R, Dasilva JN, deKemp RA. Quantification of myocardial blood flow with Rubidium-82 dynamic PET imaging. Eur J Nucl Med Mol Imaging 2007; 34:1765–1774.
32. Golanowski L, deKemp RA, Beanlands RS, Ruddy TD. Variance and covariance of 82Rb kinetic model parameters: computer simulations and dynamic PET studies. IEEE Engineering in Medicine and Biology Conference 2000; 2:23–28.
33. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 2002; 105:539–542.
34. Fritz-Hansen T, Hove JD, Kofoed KF, Kelbaek H, Larsson HBW. Quantification of MRI measured myocardial perfusion reserve in healthy humans: a comparison with positron emission tomography. JMRI 2008; 27:818–824.
35. Sawada S, Muzik O, Beanlands RSB, Wolfe E, Hutchins GD, Schwaiger M. Interobserver and interstudy variability of myocardial blood flow and flow-reserve measurements with nitrogen 13 ammonia-labeled positron emission tomography. J Nucl Cardiol 1995; 2:413–422.
36. Khorsand A, Graf S, Pirich C, Muzik O, Kletter K, Dudczak R, et al. Assessment of myocardial perfusion by dynamic N-13 ammonia PET imaging: comparison of 2 tracer kinetic models. J Nucl Cardiol 2005; 12:410–417.
Keywords:© 2010 Lippincott Williams & Wilkins, Inc.
canine model; myocardial blood flow; rubidium perfusion; three-dimensional positron emission tomography; two-dimensional positron emission tomography