Deep learning-based attenuation map generation for myocardial perfusion SPECT



Attenuation correction using CT transmission scanning increases the accuracy of single-photon emission computed tomography (SPECT) and enables quantitative analysis. Current existing SPECT-only systems normally do not support transmission scanning and therefore scans on these systems are susceptible to attenuation artifacts. Moreover, the use of CT scans also increases radiation dose to patients and significant artifacts can occur due to the misregistration between the SPECT and CT scans as a result of patient motion. The purpose of this study is to develop an approach to estimate attenuation maps directly from SPECT emission data using deep learning methods.


Both photopeak window and scatter window SPECT images were used as inputs to better utilize the underlying attenuation information embedded in the emission data. The CT-based attenuation maps were used as labels with which cardiac SPECT/CT images of 65 patients were included for training and testing. We implemented and evaluated deep fully convolutional neural networks using both standard training and training using an adversarial strategy.


The synthetic attenuation maps were qualitatively and quantitatively consistent with the CT-based attenuation map. The globally normalized mean absolute error (NMAE) between the synthetic and CT-based attenuation maps were 3.60% ± 0.85% among the 25 testing subjects. The SPECT reconstructed images corrected using the CT-based attenuation map and synthetic attenuation map are highly consistent. The NMAE between the reconstructed SPECT images that were corrected using the synthetic and CT-based attenuation maps was 0.26% ± 0.15%, whereas the localized absolute percentage error was 1.33% ± 3.80% in the left ventricle (LV) myocardium and 1.07% ± 2.58% in the LV blood pool.


We developed a deep convolutional neural network to estimate attenuation maps for SPECT directly from the emission data. The proposed method is capable of generating highly reliable attenuation maps to facilitate attenuation correction for SPECT-only scanners for myocardial perfusion imaging.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7


  1. 1.

    van Dijk J, Mouden M, Ottervanger J, van Dalen J, Knollema S, Slump C, et al. Value of attenuation correction in stress-only myocardial perfusion imaging using CZT-SPECT. J Nucl Cardiol. 2017;24:395–401.

    Article  Google Scholar 

  2. 2.

    Patchett ND, Pawar S, Sverdlov A, Miller EJ. Does Improved Technology in SPECT Myocardial Perfusion Imaging Reduce Downstream Costs? An Observational Study. International Journal of Radiology and Imaging Technology. 2017;3.

  3. 3.

    Abdollahi H, Shiri I, Salimi Y, Sarebani M, Mehdinia R, Deevband MR, et al. Radiation dose in cardiac SPECT/CT: an estimation of SSDE and effective dose. Eur J Radiol. 2016;85:2257–61.

    Article  Google Scholar 

  4. 4.

    Technavio. Global SPECT Market 2017-2021. 2017. Accessed 5 Nov 2019.

  5. 5.

    Jha AK, Zhu Y, Clarkson E, Kupinski MA, Frey EC. Fisher information analysis of list-mode SPECT emission data for joint estimation of activity and attenuation distribution. arXiv preprint. arXiv:180701767. 2018.

  6. 6.

    Wu J, Liu C. Recent advances in cardiac SPECT instrumentation and imaging methods. Phys Med Biol. 2019;64:06TR1.

    Google Scholar 

  7. 7.

    Shimizu M, Fujii H, Yamawake N, Nishizaki M. Cardiac function changes with switching from the supine to prone position: analysis by quantitative semiconductor gated single-photon emission computed tomography. J Nucl Cardiol. 2015;22:301–7.

    Article  Google Scholar 

  8. 8.

    Pan T-S, King MA, Luo D-S, Dahlberg ST, Villegas BJ. Estimation of attenuation maps from scatter and photopeak window single photon-emission computed tomographic images of technetium 99m-labeled sestamibi. J Nucl Cardiol. 1997;4:42–51.

    CAS  Article  Google Scholar 

  9. 9.

    Zaidi H, Hasegawa B. Determination of the attenuation map in emission tomography. J Nucl Med. 2003;44:291–315.

    PubMed  Google Scholar 

  10. 10.

    Núñez M, Prakash V, Vila R, Mut F, Alonso O, Hutton BF. Attenuation correction for lung SPECT: evidence of need and validation of an attenuation map derived from the emission data. Eur J Nucl Med Mol Imaging. 2009;36:1076–89.

    Article  Google Scholar 

  11. 11.

    Pan T-S, King MA, de Vries DJ, Ljungberg M. Segmentation of the body and lungs from Compton scatter and photopeak window data in SPECT: a Monte-Carlo investigation. IEEE Trans Med Imaging. 1996;15:13–24.

    CAS  Article  Google Scholar 

  12. 12.

    Cade SC, Arridge S, Evans MJ, Hutton BF. Use of measured scatter data for the attenuation correction of single photon emission tomography without transmission scanning. Med Phys. 2013;40:082506.

    Article  Google Scholar 

  13. 13.

    Gourion D, Noll D, Gantet P, Celler A, Esquerré J-P. Attenuation correction using SPECT emission data only. IEEE Trans Nucl Sci. 2002;49:2172–9.

    Article  Google Scholar 

  14. 14.

    Yan Y, Zeng GL. Attenuation map estimation with SPECT emission data only. Int J Imaging Syst Technol. 2009;19:271–6.

    Article  Google Scholar 

  15. 15.

    Nuyts J, Dupont P, Stroobants S, Benninck R, Mortelmans L, Suetens P. Simultaneous maximum a posteriori reconstruction of attenuation and activity distributions from emission sinograms. IEEE Trans Med Imaging. 1999;18:393–403.

    CAS  Article  Google Scholar 

  16. 16.

    Krol A, Bowsher JE, Manglos SH, Feiglin DH, Tornai MP, Thomas FD. An EM algorithm for estimating SPECT emission and transmission parameters from emission data only. IEEE Trans Med Imaging. 2001;20:218–32.

    CAS  Article  Google Scholar 

  17. 17.

    Nie D, Trullo R, Lian J, Wang L, Petitjean C, Ruan S, et al. Medical image synthesis with deep convolutional adversarial networks. IEEE Trans Biomed Eng. 2018;65:2720–30.

    Article  Google Scholar 

  18. 18.

    Hwang D, Kang SK, Kim KY, Seo S, Paeng JC, Lee DS, et al. Generation of PET attenuation map for whole-body time-of-flight 18F-FDG PET/MRI using a deep neural network trained with simultaneously reconstructed activity and attenuation maps. J Nucl Med. 2019:jnumed. 118.219493.

  19. 19.

    Han X. MR-based synthetic CT generation using a deep convolutional neural network method. Med Phys. 2017;44:1408–19.

    CAS  Article  Google Scholar 

  20. 20.

    Shi L, Onofrey JA, Revilla EM, Toyonaga T, Menard D, Ankrah J, et al. A novel loss function incorporating imaging acquisition physics for PET attenuation map generation using deep learning. International Conference on Medical image computing and computer-assisted intervention: Springer International Publishing; 2019. p. 723–31.

  21. 21.

    Goodfellow I, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozair S, et al. Generative adversarial nets. Advances in neural information processing systems; 2014. p. 2672–80.

  22. 22.

    Mao X, Li Q, Xie H, Lau RY, Wang Z, Paul Smolley S. Least squares generative adversarial networks. Proceedings of the IEEE International Conference on Computer Vision; 2017. p. 2794–802.

  23. 23.

    Shi L, Onofrey J, Liu H, Liu Y-H, Liu C. Generating attenuation map for SPECT-only systems using generative adversarial networks. J Nucl Med. 2019;60:572.

    Google Scholar 

  24. 24.

    Hudson HM, Larkin RS. Accelerated image reconstruction using ordered subsets of projection data. IEEE Trans Med Imaging. 1994;13:601–9.

    CAS  Article  Google Scholar 

  25. 25.

    Isola P, Zhu J-Y, Zhou T, Efros AA. Image-to-image translation with conditional adversarial networks. Proceedings of the IEEE conference on computer vision and pattern recognition; 2017. p. 1125–34.

  26. 26.

    Ronneberger O, Fischer P, Brox T. U-net: convolutional networks for biomedical image segmentation. International Conference on Medical image computing and computer-assisted intervention: Springer; 2015. p. 234–41.

  27. 27.

    Milletari F, Navab N, Ahmadi S-A. V-net: fully convolutional neural networks for volumetric medical image segmentation. 2016 Fourth International Conference on 3D Vision (3DV): IEEE; 2016. p. 565–71.

  28. 28.

    Onofrey JA, Casetti-Dinescu DI, Lauritzen AD, Sarkar S, Venkataraman R, Fan RE, et al. Generalizable multi-site training and testing of deep neural networks using image normalization. Biomedical Imaging (ISBI), 2019 IEEE 16th International Symposium on; 2019. p. pp. 1–4.

  29. 29.

    Abadi M, Barham P, Chen J, Chen Z, Davis A, Dean J, et al. Tensorflow: a system for large-scale machine learning. 12th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 16); 2016. p. 265–83.

Download references


This study was supported by American Heart Association award 18PRE33990138 and National Institute of Health grant R01HL123949. We would like to thank Dr. Eric Frey from Johns Hopkins University for the image reconstruction tool and Dr. Stephanie Thorn from Yale University and Christopher Weyman from Yale New Haven Hospital for helpful discussions. We would also like to thank Drs. Edward J. Miller and Albert J. Sinusas from Yale University for their help with data acquisition and results interpretation.


This study was supported by American Heart Association award 18PRE33990138 and National Institute of Health grant R01HL123949.

Author information




All authors contributed to the study conception and design. Conception and design of the study: Luyao Shi and Chi Liu; algorithm implementation: Luyao Shi and John A. Onofrey; data acquisition: Luyao Shi and Chi Liu; data analysis: Luyao Shi, Hui Liu, and Yi-Hwa Liu; writing of the first draft of the manuscript: Luyao Shi. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Chi Liu.

Ethics declarations

Conflict of interest

Chi Liu, Luyao Shi, and John Onofrey are named inventors on a provisional patent application that Yale University has filed on this work. No other potential conflicts of interest relevant to this article exist.

Ethical approval retrospective studies

The retrospective use of the anonymized data in this study was approved by Yale Institutional Review Board under protocol number 2000026790.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This article is part of the Topical Collection on Cardiology

Electronic supplementary material


(DOCX 33 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Shi, L., Onofrey, J.A., Liu, H. et al. Deep learning-based attenuation map generation for myocardial perfusion SPECT. Eur J Nucl Med Mol Imaging 47, 2383–2395 (2020).

Download citation


  • Synthetic attenuation map
  • Deep learning
  • Myocardial perfusion imaging