Biologically-Constrained Multi-Label Classification with Learnable Domain Knowledge

Nabil Mouadden, Véronique Vergé, Ahmadreza Arbab, Jean-Baptiste Micol, Elsa Bernard, Aline Renneville, Stergios Christodoulidis, Maria Vakalopoulou
Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, PMLR 301:1196-1212, 2026.

Abstract

Although recent foundation models trained in a self-supervised setting have shown promise in cellular image analysis, they often produce biologically impossible predictions when handling multiple concurrent abnormalities. This is a problem, as the biological information that may be needed for the different clinical-oriented problems is not directly presented in the images. In this study, we present a novel and modular approach to enforce biological constraints in multi-label medical imaging classification. Building on the powerful and rich representations of the DinoBloom hematological foundation model, our method combines learnable constraint matrices with adaptive thresholding, effectively preventing contradictory predictions while maintaining high sensitivity.Extensive experiments on three datasets, two public and one in-house on neutrophil classification, demonstrate significant improvements over different foundation models and the state-of-the-art methods. Through detailed ablation studies and hyperparameter interpretation, we show that our approach successfully captures biological relationships between different abnormalities.

Cite this Paper

BibTeX
@InProceedings{pmlr-v301-mouadden26a, title = {Biologically-Constrained Multi-Label Classification with Learnable Domain Knowledge}, author = {Mouadden, Nabil and Verg\'e, V\'eronique and Arbab, Ahmadreza and Micol, Jean-Baptiste and Bernard, Elsa and Renneville, Aline and Christodoulidis, Stergios and Vakalopoulou, Maria}, booktitle = {Proceedings of The 8th International Conference on Medical Imaging with Deep Learning}, pages = {1196--1212}, year = {2026}, editor = {Tasdizen, Tolga and Elhabian, Shireen and Summers, Ronald and Chen, Chen and Koch, Lisa and Zhuang, Yan}, volume = {301}, series = {Proceedings of Machine Learning Research}, month = {09--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v301/main/assets/mouadden26a/mouadden26a.pdf}, url = {https://proceedings.mlr.press/v301/mouadden26a.html}, abstract = {Although recent foundation models trained in a self-supervised setting have shown promise in cellular image analysis, they often produce biologically impossible predictions when handling multiple concurrent abnormalities. This is a problem, as the biological information that may be needed for the different clinical-oriented problems is not directly presented in the images. In this study, we present a novel and modular approach to enforce biological constraints in multi-label medical imaging classification. Building on the powerful and rich representations of the DinoBloom hematological foundation model, our method combines learnable constraint matrices with adaptive thresholding, effectively preventing contradictory predictions while maintaining high sensitivity.Extensive experiments on three datasets, two public and one in-house on neutrophil classification, demonstrate significant improvements over different foundation models and the state-of-the-art methods. Through detailed ablation studies and hyperparameter interpretation, we show that our approach successfully captures biological relationships between different abnormalities.} }
Endnote
%0 Conference Paper %T Biologically-Constrained Multi-Label Classification with Learnable Domain Knowledge %A Nabil Mouadden %A Véronique Vergé %A Ahmadreza Arbab %A Jean-Baptiste Micol %A Elsa Bernard %A Aline Renneville %A Stergios Christodoulidis %A Maria Vakalopoulou %B Proceedings of The 8th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Tolga Tasdizen %E Shireen Elhabian %E Ronald Summers %E Chen Chen %E Lisa Koch %E Yan Zhuang %F pmlr-v301-mouadden26a %I PMLR %P 1196--1212 %U https://proceedings.mlr.press/v301/mouadden26a.html %V 301 %X Although recent foundation models trained in a self-supervised setting have shown promise in cellular image analysis, they often produce biologically impossible predictions when handling multiple concurrent abnormalities. This is a problem, as the biological information that may be needed for the different clinical-oriented problems is not directly presented in the images. In this study, we present a novel and modular approach to enforce biological constraints in multi-label medical imaging classification. Building on the powerful and rich representations of the DinoBloom hematological foundation model, our method combines learnable constraint matrices with adaptive thresholding, effectively preventing contradictory predictions while maintaining high sensitivity.Extensive experiments on three datasets, two public and one in-house on neutrophil classification, demonstrate significant improvements over different foundation models and the state-of-the-art methods. Through detailed ablation studies and hyperparameter interpretation, we show that our approach successfully captures biological relationships between different abnormalities.
APA
Mouadden, N., Vergé, V., Arbab, A., Micol, J., Bernard, E., Renneville, A., Christodoulidis, S. & Vakalopoulou, M.. (2026). Biologically-Constrained Multi-Label Classification with Learnable Domain Knowledge. Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 301:1196-1212 Available from https://proceedings.mlr.press/v301/mouadden26a.html.

Related Material