Superpixel-guided Iterative Learning from Noisy Labels for Medical Image Segmentation

TL;DR

This paper introduces a superpixel-guided iterative learning method that effectively handles noisy labels in medical image segmentation by leveraging structural priors and noise-aware training, leading to improved robustness and accuracy.

Contribution

The proposed approach uniquely integrates superpixel representations with iterative noise refinement, exploiting structural constraints to enhance segmentation from noisy labels.

Findings

Outperforms state-of-the-art methods on benchmark datasets.

Demonstrates robustness across various levels of label noise.

Effectively leverages structural priors to improve segmentation accuracy.

Abstract

Learning segmentation from noisy labels is an important task for medical image analysis due to the difficulty in acquiring highquality annotations. Most existing methods neglect the pixel correlation and structural prior in segmentation, often producing noisy predictions around object boundaries. To address this, we adopt a superpixel representation and develop a robust iterative learning strategy that combines noise-aware training of segmentation network and noisy label refinement, both guided by the superpixels. This design enables us to exploit the structural constraints in segmentation labels and effectively mitigate the impact of label noise in learning. Experiments on two benchmarks show that our method outperforms recent state-of-the-art approaches, and achieves superior robustness in a wide range of label noises. Code is available at https://github.com/gaozhitong/SP_guided_Noisy_Label_Seg.