Scribble-Supervised Medical Image Segmentation via Dual-Branch Network and Dynamically Mixed Pseudo Labels Supervision

TL;DR

This paper introduces a dual-branch network that leverages scribble annotations and dynamically mixed pseudo labels to improve medical image segmentation, reducing annotation costs while achieving superior performance.

Contribution

The proposed method combines scribble supervision with dynamically generated pseudo labels in a dual-branch network for efficient end-to-end training.

Findings

Outperforms existing scribble-supervised segmentation methods

Achieves better results than several semi-supervised approaches

Effective on cardiac MRI segmentation with the ACDC dataset

Abstract

Medical image segmentation plays an irreplaceable role in computer-assisted diagnosis, treatment planning, and following-up. Collecting and annotating a large-scale dataset is crucial to training a powerful segmentation model, but producing high-quality segmentation masks is an expensive and time-consuming procedure. Recently, weakly-supervised learning that uses sparse annotations (points, scribbles, bounding boxes) for network training has achieved encouraging performance and shown the potential for annotation cost reduction. However, due to the limited supervision signal of sparse annotations, it is still challenging to employ them for networks training directly. In this work, we propose a simple yet efficient scribble-supervised image segmentation method and apply it to cardiac MRI segmentation. Specifically, we employ a dual-branch network with one encoder and two slightly different decoders for image segmentation and dynamically mix the two decoders' predictions to generate pseudo labels for auxiliary supervision. By combining the scribble supervision and auxiliary pseudo labels supervision, the dual-branch network can efficiently learn from scribble annotations end-to-end. Experiments on the public ACDC dataset show that our method performs better than current scribble-supervised segmentation methods and also outperforms several semi-supervised segmentation methods.