Self-Supervised Capsule-Generative Networks for Medical Image Anomaly Detection
Keywords:
Capsule Networks, Medical Imaging, Self-Supervised Learning, Unsupervised Pretraining, Interpretable Deep Learning, Anomaly Detection, Generative Pretext Tasks, Latent Representation Learning, Multimodal Reconstruction, Domain-Agnostic Encoding, Perceptual Consistency, Explainable AI, Clinical Decision Support, Robust Feature Learning, Transfer Learning, Capsule Representation, Reconstruction-Based Learning, Healthcare AI Deployment, Model Interpretability, Clinical Imaging Analytics.Abstract
Capsule-based medical imaging has shown promise as a compact, interpretable deep-learning architecture. Unsupervised pretraining is often employed to avoid expensive annotation costs. Self-supervised models can leverage unlabeled data while remaining perceptually aligned with the downstream task, improving robustness and transferability. This work proposes a self-supervised learning framework for these models. For both capsule representation and interpretation, pretext tasks are chosen to maximize the latent learning signal available in unlabeled data. Capsule representation is a key aspect of the approach. Using capsule-based images for these generative tasks allows reconstruction of the input modality or any other modality in the dataset, enabling domain-agnostic and perceptually consistent capsule encoding. The entire setup is articulated for anomaly detection and for generating explanations for classification outcomes. Finally, the entire pipeline is evaluated on multiple datasets, demonstrating its potential. Results show that these properties deliver a clinically relevant increase in performance and establish the groundwork for practical, clinical deployments.
References
1. Han, C., Rundo, L., Murao, K., Noguchi, T., Shimahara, Y., Milacski, Z. Á., Koshino, S., Sala, E., Nakayama, H., & Satoh, S. (2021). MADGAN: Unsupervised medical anomaly detection GAN using multiple adjacent brain MRI slice reconstruction. BMC Bioinformatics, 22, 31.
2. Quan, H., Xu, X., Zheng, T., Li, Z., Zhao, M., & Cui, X. (2021). DenseCapsNet: Detection of COVID-19 from X-ray images using a capsule neural network. Computers in Biology and Medicine, 133, 104399.
3. Jogi, N. L., Kumar, K. R., Mangala, N., M.Govindarajan, Gamini, P., & Bundhe, S. (2026). Machine Learning-Based Random Forest Model for Sustainable Supply Chain Sales Forecasting. In 2026 6th International Conference on Intelligent Technologies (CONIT) (pp. 1–6). IEEE. 2026 6th International Conference on Intelligent Technologies (CONIT). https://doi.org/10.1109/conit69683.2026.11620788
4. LaLonde, E., Torabi, A., Dutil, F., Pal, C., & Cohen, J. P. (2021). Capsules for biomedical image segmentation. Medical Image Analysis, 68, 101889.
5. Zhao, Q., Liu, Z., Adeli, E., & Pohl, K. M. (2021). Longitudinal self-supervised learning. Medical Image Analysis, 71, 102051.
6. Lakhanpal, S., Tunjungsari, H. K., Raj, S., Chukka, A., Dawadi, D., & Mangalampalli, B. M. (2026, April). The Digital Transformation of Healthcare: Balancing Opportunities, Risks, and Ethical Considerations. In 2026 International Conference on Multidisciplinary Innovations For Smart & Sustainable Future (MISSF) (pp. 1-6). IEEE.
7. He, Y., Carass, A., Zuo, L., Dewey, B. E., Roy, S., Resnick, S. M., Prince, J. L., & Landman, B. A. (2021). Autoencoder based self-supervised test-time adaptation for medical image analysis. Medical Image Analysis, 72, 102136.
8. Zhou, Z., Sodha, V., Pang, J., Gotway, M. B., & Liang, J. (2021). Models Genesis. Medical Image Analysis, 67, 101840.
9. Recharla, M., Garapati, R. S., Bandi, V. D. V. K., Yandamuri, U. S., & Mangalampalli, B. M. (2026, March). Generative AI-Enhanced Data Engineering Pipelines for Predictive Biomarker Discovery in Alzheimer’s and Kidney Disease. In 2026 IEEE International Conference on AI Engineering and Innovations (AIEI) (pp. 1-5). IEEE.
10. Caron, M., Touvron, H., Misra, I., Jégou, H., Mairal, J., Bojanowski, P., & Joulin, A. (2021). Emerging properties in self-supervised vision transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 9650–9660).
11. Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., & Guo, B. (2021). Swin Transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 10012–10022).
12. Davuluri, P. S. L. (2023). AI-Augmented Sanctions Screening: Enhancing Accuracy and Latency in Real Time Compliance Systems. AI-Augmented Sanctions Screening: Enhancing Accuracy and Latency in Real Time Compliance Systems (December 15, 2023).
13. Xu, J. (2021). A review of self-supervised learning methods in the field of medical image analysis. International Journal of Image, Graphics and Signal Processing, 13(4), 33–46.
14. Shurrab, S., & Duwairi, R. (2022). Self-supervised learning methods and applications in medical imaging analysis: A survey. PeerJ Computer Science, 8, e1045.
15. Krishnan, R., Rajpurkar, P., & Topol, E. J. (2022). Self-supervised learning in medicine and healthcare. Nature Biomedical Engineering, 6, 1346–1352.
16. Tiu, E., Talius, E., Patel, P., Langlotz, C. P., Ng, A. Y., & Rajpurkar, P. (2022). Expert-level detection of pathologies from unannotated chest X-ray images via self-supervised learning. Nature Biomedical Engineering, 6, 1399–1406.
17. Kolla, S. K., Bandi, V. D. V. K., & Meda, R. (2026). Comment on “From laboratory promise to decision-grade practice: strengthening reproducibility and casework relevance in AI-assisted bloodstain ageing”. Forensic Science, Medicine and Pathology, 1-2.
18. Tang, Y., Yang, D., Li, W., Roth, H. R., Landman, B., Xu, D., Nath, V., & Hatamizadeh, A. (2022). Self-supervised pre-training of Swin transformers for 3D medical image analysis. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 20730–20740).
19. Pinaya, W. H. L., Tudosiu, P.-D., Gray, R., Rees, G., Nachev, P., Ourselin, S., & Cardoso, M. J. (2022). Unsupervised brain imaging 3D anomaly detection and segmentation with transformers. Medical Image Analysis, 79, 102475.
20. Tehrani Zade, A. A., Jalili Aziz, M., Masoudnia, S., Mirbagheri, A., & Ahmadian, A. (2022). An improved capsule network for glioma segmentation on MRI images: A curriculum learning approach. Computers in Biology and Medicine, 148, 105917.
21. Tian, B., Xia, S., Ren, J., Yang, L., Wang, H., Yu, H., et al. (2022). Fully automatic pipeline of convolutional neural networks and capsule networks to distinguish COVID-19 from community-acquired pneumonia via CT images. Computers in Biology and Medicine, 141, 105182.
22. Pal, A., Chaturvedi, A., Chandra, A., Chatterjee, R., Senapati, S., Frangi, A. F., & Garain, U. (2022). MICaps: Multi-instance capsule network for machine inspection of Munro's microabscess. Computers in Biology and Medicine, 140, 105071.
23. Zhou, H.-Y., Chen, X., Zhang, Y., Luo, R., Wang, L., & Yu, Y. (2022). Generalized radiograph representation learning via cross-supervision between images and free-text radiology reports. Nature Machine Intelligence, 4, 32–40.
24. Wang, Z., Wu, Z., Agarwal, D., & Sun, J. (2022). MedCLIP: Contrastive learning from unpaired medical images and text. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing (pp. 3876–3887). Association for Computational Linguistics.
25. Ghesu, F. C., Georgescu, B., Mansoor, A., Yoo, Y., Neumann, D., Horst, F., et al. (2022). Contrastive self-supervised learning from 100 million medical images with optional supervision. Journal of Medical Imaging, 9(6), 064503.
26. Tran, M., Wagner, S. J., Boxberg, M., & Peng, T. (2022). S5CL: Unifying fully-supervised, self-supervised, and semi-supervised learning through hierarchical contrastive learning. In Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 (pp. 99–108). Springer.
27. Zhu, J., Li, Y., Ding, L., & Zhou, S. K. (2022). Aggregative self-supervised feature learning from limited medical images. In Medical Image Computing and Computer Assisted Intervention – MICCAI 2022. Springer.
28. Wu, Y., Zeng, D., Wang, Z., Shi, Y., & Hu, J. (2022). Distributed contrastive learning for medical image segmentation. Medical Image Analysis, 81, 102564.
29. Zhou, J., Wei, C., Wang, H., Shen, W., Xie, C., Yuille, A., & Kong, T. (2022). iBOT: Image BERT pre-training with online tokenizer. In International Conference on Learning Representations.
30. Nandan, B. P., Kumar, M. V. K., Garapati, R. S., Bandi, V. D. V. K., Davuluri, P. N., & Mangalampalli, B. M. (2026, March). AI-Enhanced Semiconductor Yield Optimization Using Hybrid Deep Learning and Edge Data Analytics. In 2026 IEEE International Conference on AI Engineering and Innovations (AIEI) (pp. 1-6). IEEE.
31. He, K., Chen, X., Xie, S., Li, Y., Dollár, P., & Girshick, R. (2022). Masked autoencoders are scalable vision learners. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 16000–16009).
32. Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2022). High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 10684–10695).
33. Huang, S.-C., Pareek, A., Jensen, M., Lungren, M. P., Yeung, S., & Chaudhari, A. S. (2023). Self-supervised learning for medical image classification: A systematic review and implementation guidelines. npj Digital Medicine, 6, 74.
34. Zhang, C., Zheng, H., & Gu, Y. (2023). Dive into the details of self-supervised learning for medical image analysis. Medical Image Analysis, 89, 102879.
35. Tian, Y., Liu, F., Pang, G., Chen, Y., Liu, Y., Verjans, J. W., Singh, R., & Carneiro, G. (2023). Self-supervised pseudo multi-class pre-training for unsupervised anomaly detection and segmentation in medical images. Medical Image Analysis, 90, 102930.
36. Kolla, S. H., Nagabhyru, K. C., & Kummari, D. N. (2026). Letter to the Editor re:“CurvAssist: An AI assisted pipeline for penile shaft segmentation/curvature measurement in children with hypospadias”. Journal of Pediatric Urology.
37. Li, Y., Lao, Q., Kang, Q., Jiang, Z., Du, S., Zhang, S., & Li, K. (2023). Self-supervised anomaly detection, staging and segmentation for retinal images. Medical Image Analysis, 87, 102805.
38. Kascenas, A., Sanchez, P., Schrempf, P., Wang, C., Clackett, W., Mikhael, S. S., Voisey, J. P., Goatman, K., Weir, A., Pugeault, N., Tsaftaris, S. A., & O'Neil, A. Q. (2023). The role of noise in denoising models for anomaly detection in medical images. Medical Image Analysis, 90, 102963.
39. Konz, N., Dong, H., & Mazurowski, M. A. (2023). Unsupervised anomaly localization in high-resolution breast scans using deep pluralistic image completion. Medical Image Analysis, 87, 102836.
40. Georgescu, M.-I., Ionescu, R. T., Khan, F. S., Popescu, M., & Shah, M. (2023). Masked autoencoders for unsupervised anomaly detection in medical images. Procedia Computer Science, 225, 969–978.
41. Sahlsten, J., Jaskari, J., Kivinen, J., Turunen, L., Jaanio, E., Hietala, K., & Kaski, K. (2023). Unsupervised anomaly detection with generative adversarial networks in mammography. Scientific Reports, 13, 2925.
42. Gopinath, A., Davuluri, P. N., Kumar, U. B., MP, S., & Yamini, G. (2026, April). Communication Aware Malware Detection Using Network Flow Bytecode Fusion With Adaptive Focal Optimization. In 2026 2nd International Conference on Intelligent Systems and Computational Networks (ICISCN) (pp. 1-8). IEEE.
43. Wu, Y., Qi, Q., & Qi, S. (2023). Classification of COVID-19 from community-acquired pneumonia: Boosting the performance with capsule network and maximum intensity projection image of CT scans. Computers in Biology and Medicine, 154, 106567.
44. Ramesh, S., Srivastav, V., Alapatt, D., Yu, T., Murali, A., Sestini, L., Nwoye, C. I., Hamoud, I., Sharma, S., Fleurentin, A., Exarchakis, G., Karargyris, A., & Padoy, N. (2023). Dissecting self-supervised learning methods for surgical computer vision. Medical Image Analysis, 88, 102844.
45. Haghighi, F., Hosseinzadeh Taher, M. R., Gotway, M. B., & Liang, J. (2024). Self-supervised learning for medical image analysis: Discriminative, restorative, or adversarial? Medical Image Analysis, 94, 103086.
46. Zhang, T., Wei, D., Zhu, M., Gu, S., & Zheng, Y. (2024). Self-supervised learning for medical image data with anatomy-oriented imaging planes. Medical Image Analysis, 94, 103151.
47. VanBerlo, B., Hoey, J., & Wong, A. (2024). A survey of the impact of self-supervised pretraining for diagnostic tasks in medical X-ray, CT, MRI, and ultrasound. BMC Medical Imaging, 24, 79.
48. Zeng, X., Abdullah, N., & Sumari, P. (2024). Self-supervised learning framework application for medical image analysis: A review and summary. BioMedical Engineering OnLine, 23, 107.
49. Subramanian, N., & Kolla, T. Equity in Healthcare Access Policy Lessons from Universal Coverage Models.
50. Yu, J., Oh, H., Lee, Y., & Yang, J. (2024). Denoising diffusion model with adversarial learning for unsupervised anomaly detection on brain MRI images. Pattern Recognition Letters, 186, 229–235.
51. Tang, P., Yan, X., Hu, X., Wu, K., Lasser, T., & Shi, K. (2025). Anomaly detection in medical images using encoder-attention-2decoders reconstruction. IEEE Transactions on Medical Imaging, 44(8), 3370–3382.
52. Cai, Y., Zhang, W., Chen, H., & Cheng, K.-T. (2025). MedIAnomaly: A comparative study of anomaly detection in medical images. Medical Image Analysis, 102, 103500.
53. Bi, Y., Huang, L., Clarenbach, R., Ghotbi, R., Karlas, A., Navab, N., & Jiang, Z. (2025). Synomaly noise and multi-stage diffusion: A novel approach for unsupervised anomaly detection in medical images. Medical Image Analysis, 105, 103737.
54. Frotscher, A., Kapoor, J., Wolfers, T., & Baumgartner, C. F. (2026). Unsupervised anomaly detection in medical imaging using aggregated normative diffusion. Medical Image Analysis, 109, 103895.
Additional Files
Published
Data Availability Statement
None
