Quality-Guided Semi-Supervised Learning for Medical Image Segmentation

2026-06-01

AI Summary

Medical image segmentation typically relies on large-scale, meticulously annotated datasets, which are costly to acquire. Existing semi-supervised approaches often lack explicit mechanisms to assess the reliability of pseudo-labels. To address this limitation, this work proposes a quality-guided semi-supervised learning framework that introduces, for the first time, an explicit segmentation quality prediction network. By generating diverse training samples through synthetic mask perturbations and leveraging intermediate model features, the framework objectively estimates pseudo-label reliability. These quality predictions are then integrated into the training process via a regularized loss function and dynamic reweighting of pseudo-labels. Extensive experiments demonstrate that the proposed method consistently outperforms current semi-supervised techniques across five medical imaging datasets and multiple backbone architectures, and it functions effectively as a plug-and-play module to enhance segmentation accuracy.

Abstract

Training accurate medical image segmentation models requires large amounts of densely annotated data, which is costly and time-consuming to obtain. Semi-supervised learning (SSL) alleviates this by learning from both abundant unlabeled data and limited labeled data. However, most modern SSL methods rely on pseudolabels for unlabeled data, and typically assess their reliability through model confidence or uncertainty, measures that are self-referential and lack explicit grounding in segmentation quality. Instead, we propose a quality-guided SSL framework that trains a dedicated network to estimate segmentation quality from image-mask pairs. The predictor is trained on variable-quality masks generated through synthetic corruptions augmented with imperfect outputs from partially trained segmentation models, capturing realistic error patterns encountered during training. We integrate the quality predictor into SSL through two complementary mechanisms: a quality-aware regularization loss and a quality-based pseudolabel sample reweighting scheme. We show that our method serves as a drop-in enhancement to existing SSL frameworks. Extensive experiments across five datasets and multiple architectures demonstrate consistent improvements over competing SSL methods, advancing the state-of-the-art in semi-supervised medical image segmentation.

Problem

• medical image segmentation
• semi-supervised learning
• pseudolabel reliability
• annotation scarcity
• segmentation quality

Innovation

• quality-guided
• semi-supervised learning
• medical image segmentation
• pseudolabel reweighting
• segmentation quality estimation