Hierarchical Reliability-Guided Mutual Learning for Semi-Supervised Ultrasound Image Segmentation
IEEE Transactions on Image Processing · 2026 · 35 · 11377-11391
DOI: 10.1109/tip.2026.3746986 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Image Processing, 2026. Verfasst von Qiujie Lv und 8 weiteren Autor:innen. Die Arbeit nennt 57 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Henan Province, Henan Provincial Science and Technology Research Project.
Abstract (Originalsprache)
The high cost of manually annotating ultrasound images has motivated semi-supervised learning approaches that leverage abundant unlabeled data. Mutual learning, which integrates consistency regularization and pseudo labeling through cross-network supervision, provides a promising solution for ultrasound image segmentation. However, the inherent characteristics of ultrasound images, such as speckle noise and artifacts, may lead to structurally inaccurate and semantically ambiguous pseudo labels, resulting in unreliable supervision and error propagation during iterative training. We propose Hierarchical Reliability-guided Mutual Learning (HRML), a dual-network framework that improves both pseudo label generation and reliability-aware pseudo label selection. HRML incorporates a Detail Refinement Module (DRM) to enhance fine-grained local structural representations and a Semantic Decoupling Module (SDM) to encourage semantically clearer high-level representations, thereby improving pseudo label quality. It further introduces a Multi-Dimensional Dynamic Reliability Weighting (MDRW) mechanism, which jointly evaluates the historical stability and semantic consistency of pseudo labels to suppress unreliable supervision and mitigate error propagation during mutual learning. Extensive experiments on four public ultrasound datasets confirm the superior performance of HRML over state-of-the-art fully supervised and semi-supervised methods.
Förderung
- National Natural Science Foundation of China
- Natural Science Foundation of Henan Province
- Henan Provincial Science and Technology Research Project
- China Postdoctoral Science Foundation
Zitieren
APA
Lv, Q., Liu, X., Liu, Z., Song, H., Lee, C. H., Lin, Y., Shen, L., Guo, S., & Yang, L. T. (2026). Hierarchical Reliability-Guided Mutual Learning for Semi-Supervised Ultrasound Image Segmentation. IEEE Transactions on Image Processing, 35, 11377-11391. https://doi.org/10.1109/tip.2026.3746986
Vancouver
Lv Q, Liu X, Liu Z, Song H, Lee CH, Lin Y, et al. Hierarchical Reliability-Guided Mutual Learning for Semi-Supervised Ultrasound Image Segmentation. IEEE Transactions on Image Processing. 2026;35:11377-11391. doi:10.1109/tip.2026.3746986
BibTeX
@article{Lv2026,
author = {Lv, Qiujie and Liu, Xinlong and Liu, Zhaoshan and Song, Hongyu and Lee, Chau Hung and Lin, Yusong and Shen, Lei and Guo, Saidi and Yang, Laurence T.},
title = {Hierarchical Reliability-Guided Mutual Learning for Semi-Supervised Ultrasound Image Segmentation},
journal = {IEEE Transactions on Image Processing},
year = {2026},
volume = {35},
pages = {11377--11391},
doi = {10.1109/tip.2026.3746986}
}
EUVIMED