CELNet: Evidence Localization for Pathology Images using Weakly Supervised Learning

TL;DR

CELNet introduces a weakly supervised learning method that accurately localizes cancerous evidence in pathology images, reducing annotation needs while maintaining high detection performance.

Contribution

The paper presents a novel weakly supervised approach for evidence localization in pathology images, improving interpretability without extensive annotations.

Findings

Reliable localization of cancer evidence achieved

Competitive performance on detection tasks

Reduces annotation requirements

Abstract

Despite deep convolutional neural networks boost the performance of image classification and segmentation in digital pathology analysis, they are usually weak in interpretability for clinical applications or require heavy annotations to achieve object localization. To overcome this problem, we propose a weakly supervised learning-based approach that can effectively learn to localize the discriminative evidence for a diagnostic label from weakly labeled training data. Experimental results show that our proposed method can reliably pinpoint the location of cancerous evidence supporting the decision of interest, while still achieving a competitive performance on glimpse-level and slide-level histopathologic cancer detection tasks.