%0 Journal Article
%A Chatterjee, Soumick
%A Yassin, Hadya
%A Dubost, Florian
%A Nürnberger, Andreas
%A Speck, Oliver
%T Weakly-supervised segmentation using inherently-explainable classification models and their application to brain tumour classification
%J Neurocomputing
%V 682
%@ 0925-2312
%C Amsterdam
%I Elsevier
%M DZNE-2026-00388
%P 133460
%D 2026
%X Deep learning has demonstrated significant potential in medical imaging; however, the opacity of “black-box” models hinders clinical trust, while segmentation tasks typically necessitate laborious, hard-to-obtain pixel-wise annotations. To address these challenges simultaneously, this paper introduces a framework for three inherently explainable classifiers (GP-UNet, GP-ShuffleUNet, and GP-ReconResNet). By integrating a global pooling mechanism, these networks generate localisation heatmaps that directly influence classification decisions, offering inherent interpretability without relying on potentially unreliable post-hoc methods. These heatmaps are subsequently thresholded to achieve weakly-supervised segmentation, requiring only image-level classification labels for training. Validated on two datasets for multi-class brain tumour classification, the proposed models achieved a peak F1-score of 0.93. For the weakly-supervised segmentation task, a median Dice score of 0.728 (95
%F PUB:(DE-HGF)16
%9 Journal Article
%R 10.1016/j.neucom.2026.133460
%U https://pub.dzne.de/record/286092