Simplified quantification of PET myocardial blood flow: The need for technical standardization

  • Jonathan B. Moody
  • Edward P. Ficaro
  • Venkatesh L. Murthy

Noninvasive estimation of absolute myocardial blood flow (MBF) by positron emission tomography (PET) has a long history in clinical imaging.1, 2, 3 Many different flow models have been developed over the years,4, 5, 6, 7, 8, 9, 10 which have prompted numerous model comparison studies.11, 12, 13 In recent years, several of these validated flow models have now been implemented in a new generation of software applications.14, 15, 16, 17, 18 Thus, the focus has gradually shifted from development and validation of flow models to implementation and wider dissemination of clinical tools for MBF estimation in routine patient care. This new focus has emphasized the pressing need for technical standardization.

In this issue, the study of Chang et al19 continues this trend. The authors compared 13N-ammonia MBF quantification using two previously validated MBF models provided by two FDA-approved commercial software packages: the simplified retention model8 implemented in HeartSee (University of Texas, Houston), and compartmental modeling with the two-tissue model5 implemented in syngoMBF (Siemens Healthineers). The authors’ aims were to validate the retention model for 13N-ammonia cardiac PET/CT with the compartmental modeling serving as a reference standard, and to establish normal MBF reference values using the shorter retention model PET protocol.

The HeartSee software used in this study was developed at Dr K.L. Gould’s lab at the University of Texas, which may be considered the reference implementation of the simplified retention model.8,14 The retention model has also been implemented in other academic and commercial software packages, including FlowQuant (Ottawa Heart), ImagenQ (CVIT), MunichHeart (Technical University of Munich), and 4DM (INVIA). Since the 1990s, several research groups have performed detailed validation studies of the retention model for both 13N-ammonia6, 7, 8 and 82Rb.8,20 The major advantage of the retention model for clinical measurement of MBF is the simplified PET protocol and greatly reduced computing demands for image reconstruction and processing compared with compartmental modeling. The retention model trades flexibility for simplicity and efficiency, which, for earlier generations of clinical PET scanners, addressed the inherent technological limitations that made dynamic PET and full compartmental modeling impractical for routine use.8,21 Another important potential advantage of the retention model is reduced variability of the MBF estimates 22,23 at the expense of increased bias due to the use of approximations and fixed correction factors.8 In routine clinical applications, utility is very often determined by physiological and methodological variability rather than systematic error (bias).

There are three primary limitations of the retention model. First, the model assumes that tracer taken up by the myocardium is irreversibly trapped and does not subsequently wash out.8 This approximation is sufficiently accurate for normally perfused, viable myocytes, but in practice, tracer washout may occur in the presence of severe ischemia or non-transmural scar.8,24 Our own recent work has demonstrated that 82Rb washout can in fact be used to reliably assess myocardial viability in ischemic cardiomyopathy patients.25

Second, the retention model assumes that the entire integrated arterial input function can be captured during the initial fixed two-minute blood pool image, which may be inaccurate in cases with low cardiac function, or when physiological delay of the initial tracer bolus is exacerbated by the lack of a saline flush in the infusion system.26 A hybrid approach27 which largely overcomes this limitation (when list-mode acquisition is available on the PET scanner) consists in performing preliminary image reconstructions of short dynamic frames during the blood pool phase to determine the time course of activity in the left ventricle, and them summing the appropriate dynamic images to generate the integrated arterial input image. Although this procedure undermines the simplicity of the retention model, it can largely be automated, and additionally provides the opportunity to perform motion correction between dynamic frames during this crucial period of the PET acquisition.28

Finally, the retention model requires partial volume-correction factors which must generally be determined locally by phantom scans for each PET scanner and radionuclide.8 Previous determinations have been described for 82Rb and 13N-ammonia for the University of Texas cesium-fluoride PET scanner,8 and for 82Rb and 18F for the BGO-based GE Discovery ST PET/CT scanner.29 In the present study,19 although the manufacturer was noted (GE Healthcare), the PET scanner model was not reported, and the authors did not report the determination of partial volume-correction factors appropriate for their PET scanner or for 13N-ammonia, although partial volume corrections were performed by the HeartSee software.19 It is common that the same fixed partial volume-correction factors are used for all patients, assuming a mean diastolic wall thickness of 1 cm, which may become inaccurate at the extremes of very thin or very thick myocardial walls. This could be improved by utilizing patient-specific correction factors,29 but again this may reduce the simplicity and appeal of the retention model.

Does the retention model continue to offer advantages today? The retention model was primarily developed and utilized on 2D PET scanners,6,7,14,20 and may still be preferred or necessary in the case of older BGO-based PET-only and PET/CT scanners without list-mode capabilities. Some remanufactured 2D systems have now added list-mode features, and for such systems, the retention model with modifications mentioned above may offer some advantages over full compartmental modeling. However, all contemporary 3D PET/CT scanners have list-mode capabilities, and newer tracer delivery systems and clinical software have largely mitigated the technical challenges that previously limited the role of routine compartmental modeling and MBF estimation in the clinic.

Finally, did Chang et al19 achieve their stated aims in this study? The mean flows are consistent with values from the literature (see Ref. 30 online appendix). However, it is necessary to consider the methodological variability of each approach to determine whether the MBF estimates from the two models are “close enough.” The short-term repeatability (within minutes) of the retention model (in terms of the 95% repeatability coefficient, RPC) has only been reported for Rb-82 (15-20%),22,23 while the 95% RPC of compartmental modeling is 20% for 13N-ammonia.31 Considering the global flow estimates in Figure 1 of Ref. 19, and using 95% RPC limits of ± 20%, we find that 27% of rest scans and 32% of stress scans were beyond these limits (Figure 1), suggesting a notable lack of agreement in as many as one-third of patients studied. Possible reasons for this disagreement might be the possible use of incorrect partial volume-correction factors in the retention model, or that metabolite correction was performed for the retention model but not for the compartmental model, which could lead to MBF underestimation by 5-10%.5,32 Moreover, regarding clinical efficiency, it is noteworthy that validation5,33 of the compartmental model used by Chang et al19 was originally demonstrated with PET scan durations of only 10 min (rather than the 16 min used here), and there are other validated compartmental models available for 13N-ammonia that require even shorter durations.12,13
Figure 1

Bland-Altman global left ventricular flow data from Figure 1b of reference,19 replotted with 95% repeatability limits (RPC ± 20%, dashed lines).22,23,31R-flow Simplified retention model, C-flow compartment model. 27% of rest scans and 32% of stress scans were outside the 95% repeatability limits

If the underlying goal is local implementation of reliable, routine clinical MBF measurement, Chang et al may be on the right track, but we must conclude that additional protocol optimization may be necessary. The difficulty is not a lack of tools, but a lack of technical standardization. The implementation process could be made truly simpler if all stakeholders engaged collaboratively to develop technical standards for MBF quantification that will ultimately bring quantifiable clinical value and tangible benefits for patient care.



J.B. Moody is an employee of INVIA. E.P. Ficaro is a stockholder of INVIA, which produces 4DM, a clinical software package for cardiac PET analysis. V.L. Murthy owns stock in General Electric and Cardinal Health and stock options in Ionetix. He has received consulting fees from Ionetix and Jubilant DraxImage. He has received grant funding from Siemens Medical Imaging. He has further received funding under #Grant R01HL136685 from the National Heart, Lung, and Blood Institute; and Grant #R01AG059729 from the National Institute on Aging.


  1. 1.
    Parker JA, Beller GA, Hoop B, Holman BL, Smith TW. Assessment of regional myocardial blood flow and regional fractional oxygen extraction in dogs, using 15O-water and 15O-hemoglobin. Circ Res. 1978;42:511–8.CrossRefGoogle Scholar
  2. 2.
    Schelbert HR, Phelps ME, Hoffman EJ, Huang S-C, Selin CE, Kuhl DE. Regional myocardial perfusion assessed with N-13 labeled ammonia and positron emission computerized axial tomography. Am J Cardiol. 1979;43:209–18.CrossRefGoogle Scholar
  3. 3.
    Budinger TF, Yano Y, Huesman RH, Sobel BE. Positron emission tomography of the heart. Physiologist. 1983;26:31–4.Google Scholar
  4. 4.
    Bergmann SR, Herrero P, Markham J, Weinheimer CJ, Walsh MN. Noninvasive quantitation of myocardial blood flow in human subjects with oxygen-15-labeled water and positron emission tomography. J Am Coll Cardiol. 1989;14:639–52.CrossRefGoogle Scholar
  5. 5.
    Hutchins G, Schwaiger M, Rosenspire K, Krivokapich J, Schelbert H, Kuhl D. Noninvasive quantification of regional blood flow in the human heart using N-13 ammonia and dynamic positron emission tomographic imaging. J Am Coll Cardiol. 1990;15:1032–42.CrossRefGoogle Scholar
  6. 6.
    Bellina CR, Parodi O, Camici P, Salvadori PA, Taddei L, Fusani L, et al. Simultaneous in vitro and in vivo validation of nitrogen-13-ammonia for the assessment of regional myocardial blood flow. J Nucl Med. 1990;31:1335–43.Google Scholar
  7. 7.
    Nienaber CA, Ratib O, Gambhir SS, Krivokapich J, Huang SC, Phelps ME, et al. A quantitative index of regional blood flow in canine myocardium derived noninvasively with N-13 ammonia and dynamic positron emission tomography. J Am Coll Cardiol. 1991;17:260–9.CrossRefGoogle Scholar
  8. 8.
    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;37:1701–12.Google Scholar
  9. 9.
    Lortie M, Beanlands RSB, Yoshinaga K, Klein R, Dasilva JN, DeKemp RA. Quantification of myocardial blood flow with 82Rb dynamic PET imaging. Eur J Nucl Med Mol Imaging. 2007;34:1765–74.CrossRefGoogle Scholar
  10. 10.
    Prior JO, Allenbach G, Valenta I, Kosinski M, Burger C, Delaloye FR, et al. Quantification of myocardial blood flow with 82Rb positron emission tomography: Clinical validation with 15O-water. Eur J Nucl Med Mol Imaging. 2012;39:1037–47.CrossRefGoogle Scholar
  11. 11.
    Choi Y, Huang S-C, Hawkins RA, Kuhle WG, Dahlbom M, Hoh CK, et al. A simplified method for quantification of myocardial blood flow using nitrogen-13-ammonia and dynamic PET. J Nucl Med. 1993;34:488–97.Google Scholar
  12. 12.
    DeGrado T, Hanson M, Turkington T, Delong DM, Brezinski DA, Vallée JP, et al. Estimation of myocardial blood flow for longitudinal studies with 13N-labeled ammonia and positron emission tomography. J Nucl Cardiol. 1996;3:494–507.CrossRefGoogle Scholar
  13. 13.
    Choi Y, Huang S-C, Hawkins RA, Kim JY, Kim BT, Hoh CK, et al. Quantification of myocardial blood flow using 13N-ammonia and PET: Comparison of tracer models. J Nucl Med. 1999;40:1045–55.Google Scholar
  14. 14.
    Johnson NP, Gould KL. Physiological basis for angina and ST-segment change: PET-verified thresholds of quantitative stress myocardial perfusion and coronary flow reserve. JACC Cardiovasc Imaging. 2011;4:990–8.CrossRefGoogle Scholar
  15. 15.
    Slomka PJ, Alexanderson E, Jácome R, Jiménez M, Romero E, Meave A, et al. Comparison of clinical tools for measurements of regional stress and rest myocardial blood flow assessed with 13N-ammonia PET/CT. J Nucl Med. 2012;53:171–81.CrossRefGoogle Scholar
  16. 16.
    deKemp RA, Declerck J, Klein R, Nakazato R, Tonge C, Arumugam P, et al. Multisoftware reproducibility study of stress and rest myocardial blood flow assessed with 3D dynamic PET/CT and a 1-tissue-compartment model of 82Rb kinetics. J Nucl Med. 2013;54(4):571–7.CrossRefGoogle Scholar
  17. 17.
    Nesterov SV, Deshayes E, Sciagrà R, Settimo L, Declerck JM, Pan XB, et al. Quantification of myocardial blood flow in absolute terms using 82Rb PET imaging: Results of RUBY-10 study. JACC Cardiovasc Imaging. 2014;7:1119–27.CrossRefGoogle Scholar
  18. 18.
    Harms HJ, Nesterov SV, Han C, Danad I, Leonora R, Raijmakers PG, et al. Comparison of clinical non-commercial tools for automated quantification of myocardial blood flow using oxygen-15-labelled water PET/CT. Eur Heart J Cardiovasc Imaging. 2014;15:431–41.CrossRefGoogle Scholar
  19. 19.
    Chang C-Y, Hung G-U, Hsu B, Yang BH, Chang CW, Hu LH, et al. Simplified quantification of 13N-ammonia PET myocardial blood flow: A comparative study with the standard compartment model to facilitate clinical use. J Nucl Cardiol 2018;in press.Google Scholar
  20. 20.
    Lautamäki R, George R, Kitagawa K, Higuchi T, Merrill J, Voicu C, et al. Rubidium-82 PET-CT for quantitative assessment of myocardial blood flow: Validation in a canine model of coronary artery stenosis. Eur J Nucl Med Mol Imaging. 2009;36:576–86.CrossRefGoogle Scholar
  21. 21.
    Muzik O, Duvernoy C, Beanlands R, Sawada S, Dayanikli F, Wolfe ER, et al. Assessment of diagnostic performance of quantitative flow measurements in normal subjects and patients with angiographically documented coronary artery disease by means of nitrogen-13 ammonia and positron emission tomography. J Am Coll Cardiol. 1998;31:534–40.CrossRefGoogle Scholar
  22. 22.
    Kitkungvan D, Johnson NP, Roby AE, Patel MB, Kirkeeide R, Gould KL. Routine clinical quantitative rest stress myocardial perfusion for managing coronary artery disease: Clinical relevance of test-retest variability. JACC Cardiovasc Imaging. 2016;10:565–77.CrossRefGoogle Scholar
  23. 23.
    Klein R, Ocneanu A, Renaud JM, Ziadi MC, Beanlands RSB, deKemp RA. Consistent tracer administration profile improves test–retest repeatability of myocardial blood flow quantification with 82Rb dynamic PET imaging. J Nucl Cardiol. 2016;25:929–41.CrossRefGoogle Scholar
  24. 24.
    Gould KL, Yoshida K, Hess MJ, Haynie M, Mullani N, Smalling RW. Myocardial metabolism of fluorodeoxyglucose compared to cell membrane integrity for the potassium analogue rubidium-82 for assessing infarct size in man by PET. J Nucl Med. 1991;32:1–9.Google Scholar
  25. 25.
    Moody JB, Hiller KM, Lee BC, et al. The utility of 82Rb PET for myocardial viability assessment: Comparison with perfusion-metabolism 82Rb-18F-FDG PET. J Nucl Cardiol 2018;in press.Google Scholar
  26. 26.
    Renaud JM, Wu KY, Gardner K, Aung M, Beanlands RSB, deKemp RA. Saline-push improves rubidium-82 PET image quality. J Nucl Cardiol. 2018. Scholar
  27. 27.
    Renaud JM, DaSilva JN, Beanlands RSB, deKemp RA. Characterizing the normal range of myocardial blood flow with 82rubidium and 13N-ammonia PET imaging. J Nucl Cardiol. 2013;20:578–91.CrossRefGoogle Scholar
  28. 28.
    Lee BC, Moody JB, Poitrasson-Rivière A, Melvin AC, Weinberg RL, Corbett JR, et al. Blood pool and tissue phase patient motion effects on 82rubidium PET myocardial blood flow quantification. J Nucl Cardiol. 2018. Scholar
  29. 29.
    Johnson NP, Sdringola S, Gould KL. Partial volume correction incorporating Rb-82 positron range for quantitative myocardial perfusion PET based on systolic-diastolic activity ratios and phantom measurements. J Nucl Cardiol. 2011;18:247–58.CrossRefGoogle Scholar
  30. 30.
    Gould KL, Johnson NP, Bateman TM, Beanlands RS, Bengel FM, Bober R, et al. Anatomic versus physiologic assessment of coronary artery disease Role of coronary flow reserve, fractional flow reserve, and positron emission tomography imaging in revascularization decision-making. J Am Coll Cardiol. 2013;62:1639–53.CrossRefGoogle Scholar
  31. 31.
    Nagamachi S, Czernin J, Kim AS, Sun KT, Bottcher M, Phelps ME, et al. Reproducibility of measurements of regional resting and hyperemic myocardial blood flow assessed with PET. J Nucl Med. 1996;37:1626–31.Google Scholar
  32. 32.
    Bormans G, Maes A, Langendries W, Nuyts J, Vrolix M, Vanhaecke J, et al. Metabolism of nitrogen-13 labelled ammonia in different conditions in dogs, human volunteers and transplant patients. Eur J Nucl Med Mol Imaging. 1995;22:116–21.CrossRefGoogle Scholar
  33. 33.
    Muzik O, Beanlands RSB, Hutchins GD, Mangner TJ, Nguyen N, Schwaiger M. Validation of nitrogen-13-ammonia tracer kinetic model for quantification of myocardial blood flow using PET. J Nucl Med. 1993;34:83–91.Google Scholar

Copyright information

© American Society of Nuclear Cardiology 2019

Authors and Affiliations

  • Jonathan B. Moody
    • 1
  • Edward P. Ficaro
    • 1
    • 2
    • 4
  • Venkatesh L. Murthy
    • 2
    • 3
    • 4
  1. 1.INVIA Medical Imaging SolutionsAnn ArborUSA
  2. 2.Cardiac Imaging ProgramUniversity of MichiganAnn ArborUSA
  3. 3.Division of Cardiovascular Medicine, Department of Internal MedicineUniversity of MichiganAnn ArborUSA
  4. 4.Division of Nuclear Medicine, Department of RadiologyUniversity of MichiganAnn ArborUSA

Personalised recommendations