Weakly supervised training of pixel resolution segmentation models on whole slide images
This work addresses the challenge of noisy labels in medical image segmentation for pathologists, but it appears incremental as it builds on existing weakly supervised methods.
The authors tackled the problem of training pixel-resolution segmentation models on whole slide images using weakly supervised data, achieving promising results with strong morphological consistency in segmenting tumor areas on the CAMELYON 16 dataset.
We present a novel approach to train pixel resolution segmentation models on whole slide images in a weakly supervised setup. The model is trained to classify patches extracted from slides. This leads the training to be made under noisy labeled data. We solve the problem with two complementary strategies. First, the patches are sampled online using the model's knowledge by focusing on regions where the model's confidence is higher. Second, we propose an extension of the KL divergence that is robust to noisy labels. Our preliminary experiment on CAMELYON 16 data set show promising results. The model can successfully segment tumor areas with strong morphological consistency.