A Deep Reinforcement Learning Framework for Rapid Diagnosis of Whole Slide Pathological Images

TL;DR

This paper introduces a weakly supervised deep reinforcement learning framework for rapid and accurate diagnosis of whole slide pathological images, significantly reducing inference time without requiring detailed annotations.

Contribution

It proposes a novel reinforcement learning approach inspired by clinical diagnosis, combining multi-instance learning and knowledge distillation for efficient whole slide image analysis.

Findings

Achieves fast inference on gigapixel images

Provides accurate predictions without pixel-level annotations

Reduces reliance on extensive labeled datasets

Abstract

The deep neural network is a research hotspot for histopathological image analysis, which can improve the efficiency and accuracy of diagnosis for pathologists or be used for disease screening. The whole slide pathological image can reach one gigapixel and contains abundant tissue feature information, which needs to be divided into a lot of patches in the training and inference stages. This will lead to a long convergence time and large memory consumption. Furthermore, well-annotated data sets are also in short supply in the field of digital pathology. Inspired by the pathologist's clinical diagnosis process, we propose a weakly supervised deep reinforcement learning framework, which can greatly reduce the time required for network inference. We use neural network to construct the search model and decision model of reinforcement learning agent respectively. The search model predicts the next action through the image features of different magnifications in the current field of view, and the decision model is used to return the predicted probability of the current field of view image. In addition, an expert-guided model is constructed by multi-instance learning, which not only provides rewards for search model, but also guides decision model learning by the knowledge distillation method. Experimental results show that our proposed method can achieve fast inference and accurate prediction of whole slide images without any pixel-level annotations.