Dual Agreement Consistency Learning for Semi-Supervised Fetal Ultrasound Segmentation

2026-06-23

AI Summary

This work addresses the challenge of insufficient segmentation accuracy in fetal ultrasound imaging caused by the scarcity of pixel-level annotations. To this end, we propose DACL, a semi-supervised segmentation framework that jointly trains a lightweight convolutional network and a Transformer. The approach integrates supervised learning on labeled data with cross-pseudo supervision (CPS) on unlabeled data, and introduces a novel dual consistency loss that simultaneously aligns both prediction distributions and uncertainty estimates. Furthermore, mixup-based interpolation is employed to enhance consistency learning on unlabeled samples. Experimental results demonstrate that with only 5% labeled data, our method achieves up to a 2.77% improvement in Dice score and reduces the Hausdorff distance (HD95) by as much as 14.69 mm, significantly enhancing the boundary delineation accuracy for fetal head and abdominal circumferences.

Abstract

Maternal-fetal US is the primary imaging modality for monitoring fetal development, yet accurate automated segmentation remains challenging due to the scarcity of pixel-level annotations. To address this issue, we propose DACL, a semi-supervised framework for robust fetal US image segmentation. DACL jointly trains a deployment-oriented lightweight convolutional network (1.47\thinsp\mathrm{M} parameters) and a Transformer-based network, leveraging labeled data for supervised learning and unlabeled data via CPS. To enhance prediction stability, we introduce a dual-agreement consistency loss that couples pixel-wise probabilistic divergence with entropy-guided confidence alignment. Unlike conventional CPS methods that enforce agreement only at the prediction level, DACL explicitly regularizes both distributional alignment and uncertainty, thereby suppressing unreliable pseudo-labels and enabling stable cross-architecture pseudo-label learning under extreme annotation scarcity. Furthermore, an interpolation-based consistency strategy using mixup is applied to unlabeled samples to enhance robustness. Under 5% labeled data, DACL improves Dice by up to 2.77% and reduces HD95 by up to 14.69 mm compared with the strongest recent semi-supervised methods, demonstrating significant improvements in boundary accuracy on both fetal head and abdomen datasets. These results demonstrate the effectiveness of agreement-based consistency learning for annotation-efficient fetal US segmentation. Our code is on GitHub.

Problem

• semi-supervised learning
• fetal ultrasound segmentation
• annotation scarcity
• medical image segmentation
• pseudo-labeling

Innovation

• semi-supervised learning
• dual-agreement consistency
• pseudo-labeling
• fetal ultrasound segmentation
• cross-architecture training