Not All Pixels Are Equal: Pixel-wise Meta-Learning for Medical Segmentation with Noisy Labels

TL;DR

MetaDCSeg introduces a pixel-wise meta-learning framework that dynamically weights pixels and models boundary uncertainty to improve medical image segmentation accuracy amidst noisy labels and ambiguous boundaries.

Contribution

It presents a novel pixel-wise weighting method with boundary uncertainty modeling, addressing local variations and improving robustness in noisy medical segmentation tasks.

Findings

Outperforms state-of-the-art methods across four benchmark datasets.

Effectively suppresses noisy labels while preserving reliable annotations.

Enhances boundary segmentation accuracy in noisy conditions.

Abstract

Medical image segmentation is crucial for clinical applications, but it is frequently disrupted by noisy annotations and ambiguous anatomical boundaries, limiting its application in real-world scenarios. Existing methods often directly adapt noisy label learning techniques designed for instance classification, overlooking the pixel-wise heterogeneity in medical segmentation with its spatially and anatomically varying difficulties. Consequently, global assumptions or simple confidence metrics fail to address these local variations, leaving boundary ambiguities unresolved. To address this issue, we propose MetaDCSeg, a robust framework that dynamically learns optimal pixel-wise weights to suppress the influence of noisy labels while preserving reliable annotations. By explicitly modeling boundary uncertainty through a Dynamic Center Distance (DCD) mechanism, our approach utilizes weighted feature distances for foreground, background, and boundary centers, directing the model's attention toward hard-to-segment pixels near ambiguous boundaries. This strategy enables more precise handling of structural boundaries, which are often overlooked by existing methods, and significantly enhances segmentation performance. Extensive experiments across four benchmark datasets with varying noise levels demonstrate that MetaDCSeg outperforms existing state-of-the-art methods.