Consistency-based Semi-supervised Evidential Active Learning Framework for Robust Classification of Radiology Images

Consistency-based Semi-supervised Evidential Active Learning Framework for Robust Classification of Radiology Images

IEEE Journal of Biomedical and Health Informatics

10.1109/JBHI.2026.3696868

https://doi.org/10.1109/jbhi.2026.3696868

Shafa Balaram, Manh Cuong Nguyen, Yang Yu, Ivan Ho Mien, Huijuan Yang, Navya Sindhu Vanjavaka, Nurul Hafidzah Rahim, Alistair Giap Chuan Koh, Shaheen Low, Ashish Jith Sreejith Kumar, Feri Guretno, Yih Yian Sitoh, Ashraf Kassim, Pavitra Krishnaswamy

Active learning, Medical image annotation, Noisy labels, semi-supervised learning, Theory of Evidence

25 May 2026

Balaram, S., Nguyen, M. C., Yu, Y., Mien, I. H., Yang, H., Vanjavaka, N. S., Rahim, N. H., Koh, A. G. C., Low, S., Kumar, A. J. S., Guretno, F., Sitoh, Y. Y., Kassim, A., & Krishnaswamy, P. (2026). Consistency-based Semi-supervised Evidential Active Learning Framework for Robust Classification of Radiology Images. IEEE Journal of Biomedical and Health Informatics, 1–13. https://doi.org/10.1109/jbhi.2026.3696868

Deep learning offers high performance for radiology image classification, but relies on large, expert-annotated datasets. Semi-supervised learning and active learning approaches can leverage unlabelled samples and mitigate the annotation burden. Combining these techniques via semi-supervised active learning (SSAL) can compound their benefits, yet the effectiveness of this approach may be hindered by unreliable uncertainty estimation and consistency enforcement. To address these challenges, we propose Consistency-based Semi-supervised Evidential Active Learning (CSEAL). CSEAL is a principled SSAL framework leveraging evidential learning for reliable estimation of predictive uncertainty for consistency enforcement and prioritised sampling. First, we develop evidential counterparts of leading semi-supervised methods with different consistency enforcement mechanisms: Pseudo-labelling, Virtual Adversarial Training, Mean Teacher, and NoTeacher, and demonstrate customisability of CSEAL. Second, we introduce Noise Robust-evidential NoTeacher, an enhancement over evidential NoTeacher, that uses consensus principles and a small-loss inclusion mechanism to learn with noisy annotations. In extensive experiments on many X-ray, CT, and MRI datasets, we demonstrate that CSEAL offers substantial performance gains over competitive SSAL baselines and \cyan{enhances} robustness in noisy annotation scenarios. Third, we translate CSEAL into an annotation platform and demonstrate its value as an aid for real-world radiology image annotation. Specifically, our CSEAL-assisted platform performs close to fully-supervised learning with very small labelled datasets, enables accurate auto-labelling, and maintains performance even when only noisy labels from junior annotators are available for training. Our work offers new opportunities to enhance the efficiency of clinical image annotation and model development workflows.

Publisher Copyright

This research / project is supported by the A*STAR - Advanced Manufacturing and Engineering (AME) Programmatic Fund
Grant Reference no. : A20H6b0151

© 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

https://oar.a-star.edu.sg/communities-collections/articles/23240

2168-2194
2168-2208

Institute for Infocomm Research