Weakly Supervised Nuclei Segmentation via Instance Learning

TL;DR

This paper introduces a modular deep learning approach for weakly supervised nuclei segmentation that decouples semantic and instance segmentation, significantly improving performance on pathological image benchmarks.

Contribution

It proposes a novel two-branch network with instance-sensitive loss to enhance instance-aware learning from point annotations.

Findings

Achieves state-of-the-art results on two public pathological image datasets.

Effectively handles crowded nuclei with improved instance differentiation.

Reduces labeling costs by using weak supervision.

Abstract

Weakly supervised nuclei segmentation is a critical problem for pathological image analysis and greatly benefits the community due to the significant reduction of labeling cost. Adopting point annotations, previous methods mostly rely on less expressive representations for nuclei instances and thus have difficulty in handling crowded nuclei. In this paper, we propose to decouple weakly supervised semantic and instance segmentation in order to enable more effective subtask learning and to promote instance-aware representation learning. To achieve this, we design a modular deep network with two branches: a semantic proposal network and an instance encoding network, which are trained in a two-stage manner with an instance-sensitive loss. Empirical results show that our approach achieves the state-of-the-art performance on two public benchmarks of pathological images from different types of organs.