Leveraging Uncertainty for Deep Interpretable Classification and Weakly-Supervised Segmentation of Histology Images
This work addresses interpretability and segmentation challenges in histology image analysis, which is crucial for medical diagnostics, but it is incremental as it builds on existing weakly-supervised methods with specific improvements.
The paper tackled the problem of high false-positive rates in weakly-supervised segmentation of histology images, where regions of interest are visually similar to background, by proposing novel regularization terms that model both discriminative and non-discriminative regions, resulting in reduced false positives and accurate ROI segmentation on two datasets.
Trained using only image class label, deep weakly supervised methods allow image classification and ROI segmentation for interpretability. Despite their success on natural images, they face several challenges over histology data where ROI are visually similar to background making models vulnerable to high pixel-wise false positives. These methods lack mechanisms for modeling explicitly non-discriminative regions which raises false-positive rates. We propose novel regularization terms, which enable the model to seek both non-discriminative and discriminative regions, while discouraging unbalanced segmentations and using only image class label. Our method is composed of two networks: a localizer that yields segmentation mask, followed by a classifier. The training loss pushes the localizer to build a segmentation mask that holds most discrimiantive regions while simultaneously modeling background regions. Comprehensive experiments over two histology datasets showed the merits of our method in reducing false positives and accurately segmenting ROI.