Variational autoencoders can detect coronary artery disease in myocardial perfusion SPECT images
5 affiliations
- 1Department of Radiation Science, Hirosaki University Graduate School of Health Sciences, 66-1 Hon-cho, Hirosaki-shi, Aomori 036-8564 Japan
- 2Department of Nuclear Medicine, Kanazawa University Graduate School of Medical Sciences, Kanazawa, Japan
- 3Department of Electrical, Electronic and Computer Engineering, Gifu University, Gifu, Japan
- 4Center for One Medicine Innovative Translational Research, Gifu University, Gifu, Japan
- 5Department of Radiological Technology, Kanazawa University Hospital, Kanazawa, Japan
Abstract
Purpose: Variational autoencoders (VAEs) can detect unsupervised anomalies in medical images. We aimed to determine whether a VAE could automatically detect anomalies in myocardial perfusion SPECT (MPS) images. We then interpreted the training and generative processes of the VAE.
Methods: We trained a VAE on polar maps to learn the variability of normal myocardial perfusion profiles in latent space. The training dataset comprised 3,432 polar map images (49% male) without coronary artery disease (CAD) and the validation set contained 111 polar map images (59% male), among whom, (43% had obstructive CAD). This latent representation allowed the VAE to reconstruct synthetic normal images that served as references to detect anomalies. Abnormal myocardial perfusion was automatically identified by measuring differences between generated and original images using cosine similarity. We also summed stress scores (SSS) using sex-segregated normal databases.
Results: The trained VAE generated reference normal polar maps, regardless of normal and defect profiles and sex differences. The areas under the receiver-operating characteristics curve to detect CAD were 0.89 ± 0.041 for cosine similarity and 0.81 ± 0.030 for SSS (p = 0.047). The VAE learned the characteristics of normal male and female myocardial perfusion distributions in the latent space.
Conclusion: This VAE-based approach improved CAD detection in MPS images and eliminated the need to determine sex-specific myocardial perfusion characteristics.
Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1007/s00259-026-08161-1.
Related
- Discordance Between Gated Single-Photon Emission Computed Tomography Myocardial Perfusion Imaging and Coronary Angiography in Dilated Cardiomyopathy
- Left ventricular function and volumes from cardiac magnetic resonance and gated [13N]-ammonia positron emission tomography myocardial perfusion imaging in patients with ischemia with no obstructive coronary arteries: a prospective head-to-head comparison using hybrid positron emission tomography/magnetic resonance
- Hemodynamic Significance of APEXCT Coronary Angiography Combined With Dynamic CT Myocardial Perfusion Imaging in Assessing Restenosis After Coronary Stenting
- Association between coronary artery calcium score and myocardial ischemia assessed by stress perfusion cardiac magnetic resonance imaging
- Absolute stress myocardial blood flow is reduced in patients with prior cardiac disease despite normal perfusion and ejection fraction on rubidium-82 PET: an Arabian Gulf cohort
BibTeX
@article{okuda2026variational,
title = {{Variational autoencoders can detect coronary artery disease in myocardial perfusion SPECT images}},
author = {Koichi Okuda and Kenichi Nakajima and Takeshi Hara and Hiroto Yoneyama and Seigo Kinuya},
journal = {European Journal of Nuclear Medicine and Molecular Imaging},
year = {2026},
month = aug,
doi = {10.1007/s00259-026-08161-1},
url = {https://doi.org/10.1007/s00259-026-08161-1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Variational autoencoders can detect coronary artery disease in myocardial perfusion SPECT images
AU - Koichi Okuda
AU - Kenichi Nakajima
AU - Takeshi Hara
AU - Hiroto Yoneyama
AU - Seigo Kinuya
PY - 2026
DA - 2026/08/29
JO - European Journal of Nuclear Medicine and Molecular Imaging
DO - 10.1007/s00259-026-08161-1
UR - https://doi.org/10.1007/s00259-026-08161-1
AB - Purpose: Variational autoencoders (VAEs) can detect unsupervised anomalies in medical images. We aimed to determine whether a VAE could automatically detect anomalies in myocardial perfusion SPECT (MPS) images. We then interpreted the training and generative processes of the VAE. Methods: We trained a VAE on polar maps to learn the variability of normal myocardial perfusion profiles in latent space. The training dataset comprised 3,432 polar map images (49% male) without coronary artery disease (CAD) and the validation set contained 111 polar map images (59% male), among whom, (43% had obstructive CAD). This latent representation allowed the VAE to reconstruct synthetic normal images that served as references to detect anomalies. Abnormal myocardial perfusion was automatically identified by measuring differences between generated and original images using cosine similarity. We also summed stress scores (SSS) using sex-segregated normal databases. Results: The trained VAE generated reference normal polar maps, regardless of normal and defect profiles and sex differences. The areas under the receiver-operating characteristics curve to detect CAD were 0.89 ± 0.041 for cosine similarity and 0.81 ± 0.030 for SSS (p = 0.047). The VAE learned the characteristics of normal male and female myocardial perfusion distributions in the latent space. Conclusion: This VAE-based approach improved CAD detection in MPS images and eliminated the need to determine sex-specific myocardial perfusion characteristics. Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1007/s00259-026-08161-1.
SN - 1619-7070
SN - 1619-7089
ER -