Chronological Contrastive Learning: Few-Shot Progression Assessment in Irreversible Diseases

Clemens Watzenböck, Daniel Aletaha, Michaël Deman, Thomas Deimel, Jana Eder, Ivana Janı́čková, Robert Janiczek, Peter Mandl, Philipp Seeböck, Gabriela Supp, Paul Weiser, Georg Langs
Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, PMLR 315:2795-2817, 2026.

Abstract

Quantitative disease severity scoring in medical imaging is costly, time-consuming, and subject to inter-reader variability. At the same time, clinical archives contain far more longitudinal imaging data than expert-annotated severity scores. Existing self-supervised methods typically ignore this chronological structure. We introduce ChronoCon, a contrastive learning approach that replaces label-based ranking losses with rankings derived solely from the visitation order of a patient’s longitudinal scans. Under the clinically plausible assumption of monotonic progression in irreversible diseases, the method learns disease-relevant representations without using any expert labels. This generalizes the idea of Rank-N-Contrast from label distances to temporal ordering. Evaluated on rheumatoid arthritis radiographs for severity assessment, the learned representations substantially improve label efficiency. In low-label settings, ChronoCon significantly outperforms a fully supervised baseline initialized from ImageNet weights. In a few-shot learning experiment, fine-tuning ChronoCon on expert scores from only five patients yields an intraclass correlation coefficient of 86% for severity score prediction. These results demonstrate the potential of chronological contrastive learning to exploit routinely available imaging metadata to reduce annotation requirements in the irreversible disease domain.

Cite this Paper

BibTeX
@InProceedings{pmlr-v315-watzenbock26a, title = {Chronological Contrastive Learning: Few-Shot Progression Assessment in Irreversible Diseases}, author = {Watzenb{\"o}ck, Clemens and Aletaha, Daniel and Deman, Micha{\"e}l and Deimel, Thomas and Eder, Jana and Jan\'{\i}\v{c}kov\'{a}, Ivana and Janiczek, Robert and Mandl, Peter and Seeb{\"o}ck, Philipp and Supp, Gabriela and Weiser, Paul and Langs, Georg}, booktitle = {Proceedings of The 9th International Conference on Medical Imaging with Deep Learning}, pages = {2795--2817}, year = {2026}, editor = {Huo, Yuankai and Gao, Mingchen and Kuo, Chang-Fu and Jin, Yueming and Deng, Ruining}, volume = {315}, series = {Proceedings of Machine Learning Research}, month = {08--10 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v315/main/assets/watzenbock26a/watzenbock26a.pdf}, url = {https://proceedings.mlr.press/v315/watzenbock26a.html}, abstract = {Quantitative disease severity scoring in medical imaging is costly, time-consuming, and subject to inter-reader variability. At the same time, clinical archives contain far more longitudinal imaging data than expert-annotated severity scores. Existing self-supervised methods typically ignore this chronological structure. We introduce ChronoCon, a contrastive learning approach that replaces label-based ranking losses with rankings derived solely from the visitation order of a patient’s longitudinal scans. Under the clinically plausible assumption of monotonic progression in irreversible diseases, the method learns disease-relevant representations without using any expert labels. This generalizes the idea of Rank-N-Contrast from label distances to temporal ordering. Evaluated on rheumatoid arthritis radiographs for severity assessment, the learned representations substantially improve label efficiency. In low-label settings, ChronoCon significantly outperforms a fully supervised baseline initialized from ImageNet weights. In a few-shot learning experiment, fine-tuning ChronoCon on expert scores from only five patients yields an intraclass correlation coefficient of 86% for severity score prediction. These results demonstrate the potential of chronological contrastive learning to exploit routinely available imaging metadata to reduce annotation requirements in the irreversible disease domain.} }
Endnote
%0 Conference Paper %T Chronological Contrastive Learning: Few-Shot Progression Assessment in Irreversible Diseases %A Clemens Watzenböck %A Daniel Aletaha %A Michaël Deman %A Thomas Deimel %A Jana Eder %A Ivana Janı́čková %A Robert Janiczek %A Peter Mandl %A Philipp Seeböck %A Gabriela Supp %A Paul Weiser %A Georg Langs %B Proceedings of The 9th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Yuankai Huo %E Mingchen Gao %E Chang-Fu Kuo %E Yueming Jin %E Ruining Deng %F pmlr-v315-watzenbock26a %I PMLR %P 2795--2817 %U https://proceedings.mlr.press/v315/watzenbock26a.html %V 315 %X Quantitative disease severity scoring in medical imaging is costly, time-consuming, and subject to inter-reader variability. At the same time, clinical archives contain far more longitudinal imaging data than expert-annotated severity scores. Existing self-supervised methods typically ignore this chronological structure. We introduce ChronoCon, a contrastive learning approach that replaces label-based ranking losses with rankings derived solely from the visitation order of a patient’s longitudinal scans. Under the clinically plausible assumption of monotonic progression in irreversible diseases, the method learns disease-relevant representations without using any expert labels. This generalizes the idea of Rank-N-Contrast from label distances to temporal ordering. Evaluated on rheumatoid arthritis radiographs for severity assessment, the learned representations substantially improve label efficiency. In low-label settings, ChronoCon significantly outperforms a fully supervised baseline initialized from ImageNet weights. In a few-shot learning experiment, fine-tuning ChronoCon on expert scores from only five patients yields an intraclass correlation coefficient of 86% for severity score prediction. These results demonstrate the potential of chronological contrastive learning to exploit routinely available imaging metadata to reduce annotation requirements in the irreversible disease domain.
APA
Watzenböck, C., Aletaha, D., Deman, M., Deimel, T., Eder, J., Janı́čková, I., Janiczek, R., Mandl, P., Seeböck, P., Supp, G., Weiser, P. & Langs, G.. (2026). Chronological Contrastive Learning: Few-Shot Progression Assessment in Irreversible Diseases. Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 315:2795-2817 Available from https://proceedings.mlr.press/v315/watzenbock26a.html.

Related Material