ERSR: An Ellipse-constrained pseudo-label refinement and symmetric regularization framework for semi-supervised fetal head segmentation in ultrasound images

2025-08-27

AI Summary

Fetal head segmentation in ultrasound images is challenged by poor image quality and scarce annotated data, leading to unreliable pseudo-labels and ineffective consistency constraints. To address this, we propose a semi-supervised framework that first enforces anatomically plausible geometry on initial pseudo-labels via least-squares elliptical fitting. Subsequently, we introduce a multi-level symmetry regularization mechanism integrating boundary consistency, contour regularity assessment, and dual-view symmetry constraints to strengthen structural prior learning from unlabeled data. Furthermore, a dual-scoring adaptive filtering strategy is designed to refine pseudo-label selection. Evaluated on the HC18 and PSFH benchmarks, our method achieves Dice scores of 92.05% and 91.68%, respectively, using only 10% labeled data—substantially outperforming existing semi-supervised approaches. This demonstrates the critical role of shape-guided supervision and explicit symmetry modeling in enhancing robustness for ultrasound fetal head segmentation.

Technology Category

• Machine Learning: Semi-Supervised Learning
• Computer Vision: Segmentation
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Automated segmentation of the fetal head in ultrasound images is critical for prenatal monitoring. However, achieving robust segmentation remains challenging due to the poor quality of ultrasound images and the lack of annotated data. Semi-supervised methods alleviate the lack of annotated data but struggle with the unique characteristics of fetal head ultrasound images, making it challenging to generate reliable pseudo-labels and enforce effective consistency regularization constraints. To address this issue, we propose a novel semi-supervised framework, ERSR, for fetal head ultrasound segmentation. Our framework consists of the dual-scoring adaptive filtering strategy, the ellipse-constrained pseudo-label refinement, and the symmetry-based multiple consistency regularization. The dual-scoring adaptive filtering strategy uses boundary consistency and contour regularity criteria to evaluate and filter teacher outputs. The ellipse-constrained pseudo-label refinement refines these filtered outputs by fitting least-squares ellipses, which strengthens pixels near the center of the fitted ellipse and suppresses noise simultaneously. The symmetry-based multiple consistency regularization enforces multi-level consistency across perturbed images, symmetric regions, and between original predictions and pseudo-labels, enabling the model to capture robust and stable shape representations. Our method achieves state-of-the-art performance on two benchmarks. On the HC18 dataset, it reaches Dice scores of 92.05% and 95.36% with 10% and 20% labeled data, respectively. On the PSFH dataset, the scores are 91.68% and 93.70% under the same settings.

Problem

• Semi-supervised fetal head ultrasound image segmentation
• Addressing poor image quality and limited annotations
• Generating reliable pseudo-labels with ellipse constraints

Innovation

• Dual-scoring adaptive filtering for teacher outputs
• Ellipse-constrained refinement of pseudo-labels
• Symmetry-based multi-level consistency regularization