Autoencoders to detect manifestation shift in medical images

• Samuel Armbrust Freitas Unisinos
• Cristiano André da Costa Unisinos
• Gabriel de Oliveira Ramos Unisinos

DOI: https://doi.org/10.5753/sbcas.2025.7227

Abstract

Advanced intelligent models can potentially assist medical professionals with their routine tasks. Nonetheless, a recognized challenge is that these algorithms often excel in operating in-distribution settings but may underperform when confronted with Out-of-Distribution situations from diverse sources, mainly caused by manifestation shifts. This inconsistency underscores the disparity between lab-based findings and real-world clinical applications, necessitating reliable methods for measuring uncertainty. This study introduces a novel approach proposing self-awareness in detecting manifestation shifts within medical images, designing the challenge as a classification of lesion subtypes. The proposed method achieved competitive performance compared with recent literature utilizing reconstruction thresholds based on histogram analysis. Furthermore, a review of erroneous predictions uncovered various confounding factors, contributing to a better understanding of the models’ limitations.

References

Alanazi, M. F., Ali, M. U., Hussain, S. J., Zafar, A., Mohatram, M., Irfan, M., AlRuwaili, R., Alruwaili, M., Ali, N. H., and Albarrak, A. M. (2022). Brain tumor/mass classification framework using magnetic-resonance-imaging-based isolated and developed transfer deep-learning model. Sensors, 22(1):372.

Amran, G. A., Alsharam, M. S., Blajam, A. O. A., Hasan, A. A., Alfaifi, M. Y., Amran, M. H., Gumaei, A., and Eldin, S. M. (2022). Brain tumor classification and detection using hybrid deep tumor network. Electronics, 11(21):3457.

An, J. and Cho, S. (2015). Variational autoencoder based anomaly detection using reconstruction probability. Special lecture on IE, 2(1):1–18.

Castro, D. C., Walker, I., and Glocker, B. (2020). Causality matters in medical imaging. Nature Communications, 11(1):3673.

Chen, C., Tang, L., Liu, F., Zhao, G., Huang, Y., and Yu, Y. (2022). Mix and reason: Reasoning over semantic topology with data mixing for domain generalization. In Advances in Neural Information Processing Systems.

Cui, C., Wang, Y., Bao, S., Tang, Y., Deng, R., Remedios, L. W., Asad, Z., Roland, J. T., Lau, K. S., Liu, Q., et al. (2023). Feasibility of universal anomaly detection without knowing the abnormality in medical images. In Workshop on Medical Image Learning with Limited and Noisy Data, pages 82–92. Springer.

Esmaeili, M., Toosi, A., Roshanpoor, A., Changizi, V., Ghazisaeedi, M., Rahmim, A., and Sabokrou, M. (2023). Generative adversarial networks for anomaly detection in biomedical imaging: A study on seven medical image datasets. IEEE Access, 11:17906–17921.

Gómez-Guzmán, M. A., Jiménez-Beristaín, L., García-Guerrero, E. E., López-Bonilla, O. R., Tamayo-Perez, U. J., Esqueda-Elizondo, J. J., Palomino-Vizcaino, K., and Inzunza-González, E. (2023). Classifying brain tumors on magnetic resonance imaging by using convolutional neural networks. Electronics, 12(4):955.

Graham, M. S., Tudosiu, P. D., Wright, P., Pinaya, W. H. L., Teikari, P., Patel, A., U-King-Im, J. M., Mah, Y. H., Teo, J. T., Jäger, H. R., Werring, D., Rees, G., Nachev, P., Ourselin, S., and Cardoso, M. J. (2023). Latent transformer models for out-of-distribution detection. Medical Image Analysis, 90.

Guo, X., Gichoya, J. W., Purkayastha, S., and Banerjee, I. (2022). Margin-aware intraclass novelty identification for medical images. Journal of Medical Imaging, 9.

Haq, A. U., Li, J. P., Kumar, R., Ali, Z., Khan, I., Uddin, M. I., and Agbley, B. L. Y. (2023). Mcnn: a multi-level cnn model for the classification of brain tumors in iot-healthcare system. Journal of Ambient Intelligence and Humanized Computing, 14(5):4695–4706.

Huang, X. and Belongie, S. (2017). Arbitrary style transfer in real-time with adaptive instance normalization. In Proceedings of the IEEE international conference on computer vision, pages 1501–1510.

Kang, J., Ullah, Z., and Gwak, J. (2021). Mri-based brain tumor classification using ensemble of deep features and machine learning classifiers. Sensors, 21(6):2222.

Kascenas, A., Sanchez, P., Schrempf, P., Wang, C., Clackett, W., Mikhael, S. S., Voisey, J. P., Goatman, K., Weir, A., Pugeault, N., et al. (2023). The role of noise in denoising models for anomaly detection in medical images. Medical Image Analysis, 90:102963.

Ledig, C., Theis, L., Huszár, F., Caballero, J., Cunningham, A., Acosta, A., Aitken, A., Tejani, A., Totz, J., Wang, Z., et al. (2017). Photo-realistic single image super resolution using a generative adversarial network. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4681–4690.

Lewis, P. R., Chandra, A., Parsons, S., Robinson, E., Glette, K., Bahsoon, R., Torresen, J., and Yao, X. (2011). A survey of self-awareness and its application in computing systems. In 2011 Fifth IEEE conference on self-adaptive and self-organizing systems workshops, pages 102–107. IEEE.

Li, C., Lin, X., Mao, Y., Lin, W., Qi, Q., Ding, X., Huang, Y., Liang, D., and Yu, Y. (2022). Domain generalization on medical imaging classification using episodic training with task augmentation. Computers in biology and medicine, 141:105144.

Pan, X., Luo, P., Shi, J., and Tang, X. (2018). Two at once: Enhancing learning and generalization capacities via ibn-net. In Proceedings of the European Conference on Computer Vision (ECCV), pages 464–479.

Petrovska, A. (2021). Self-awareness as a prerequisite for self-adaptivity in computing systems. In 2021 IEEE International Conference on Autonomic Computing and Self-Organizing Systems Companion (ACSOS-C), pages 146–149. IEEE.

Rafiee, N., gholamipoorfard, R., and Kollmann, M. (2023). Abnormality detection for medical images using self-supervision and negative samples. bioRxiv, pages 2023–05.

Tschuchnig, M. E. and Gadermayr, M. (2022). Anomaly Detection in Medical Imaging - A Mini Review, pages 33–38. Springer Fachmedien Wiesbaden.

Xu, R., Zhang, X., Shen, Z., Zhang, T., and Cui, P. (2022). A theoretical analysis on independence-driven importance weighting for covariate-shift generalization. In International Conference on Machine Learning, pages 24803–24829. PMLR.

Yang, J., Wang, P., Zou, D., Zhou, Z., Ding, K., Peng, W., Wang, H., Chen, G., Li, B., Sun, Y., et al. (2022). Openood: Benchmarking generalized out-of-distribution detection. Advances in Neural Information Processing Systems, 35:32598–32611.

Yang, J., Zhou, K., Li, Y., and Liu, Z. (2021). Generalized out-of-distribution detection: A survey.

Zhang, K., Schölkopf, B., Muandet, K., and Wang, Z. (2013). Domain adaptation under target and conditional shift. In International conference on machine learning, pages 819–827. PMLR.

Zhou, K., Liu, Z., Qiao, Y., Xiang, T., and Loy, C. C. (2022). Domain generalization: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence.