View in EDS

C3MT: Confidence-Calibrated Contrastive Mean Teacher for semi-supervised medical image segmentation.

Saved in:
Bibliographic Details
Title: C3MT: Confidence-Calibrated Contrastive Mean Teacher for semi-supervised medical image segmentation.
Authors: Wang, Xianmin1 (AUTHOR), Lin, Mingfeng1 (AUTHOR), Li, Jing1,2 (AUTHOR) lijing@gzhu.edu.cn
Source: Computerized Medical Imaging & Graphics. Feb2026, Vol. 128, pN.PAG-N.PAG. 1p.
Subjects: Image segmentation, Feature extraction, Supervised learning, Data augmentation
Abstract: Semi-supervised learning is crucial for medical image segmentation due to the scarcity of labeled data. However, existing methods that combine consistency regularization and pseudo-labeling often suffer from inadequate feature representation, suboptimal subnetwork disagreement, and noisy pseudo-labels. To address these limitations, this paper proposed a novel C onfidence- C alibrated C ontrastive M ean T eacher (C3MT) framework. First, C3MT introduces a Contrastive Learning-based co-training strategy, where an adaptive disagreement adjustment mechanism dynamically regulates the divergence between student models. This not only preserves representation diversity but also stabilizes the training process. Second, C3MT introduces a Confidence-Calibrated and Category-Aligned uncertainty-guided region mixing strategy. The confidence-calibrated mechanism filters out unreliable pseudo-labels, whereas the category-aligned design restricts region swapping to patches of the same semantic category, preserving anatomical coherence and preventing semantic inconsistency in the mixed samples. Together, these components significantly enhance feature representation, training stability, and segmentation quality, especially in challenging low-annotation scenarios. Extensive experiments on ACDC, Synapse, and LA datasets show that C3MT consistently outperforms recent state-of-the-art methods. For example, on the ACDC dataset with 20% labeled data, C3MT achieves up to a 4.3% improvement in average Dice score and a reduction in HD95 of more than 1.0 mm compared with strong baselines. The implementation is publicly available at https://github.com/l1654485/C3MT. [Display omitted] • Proposes C3MT: a Confidence-Calibrated contrastive mean teacher framework. • Contrastive Learning-based co-training stabilizes training and diversifies features. • Confidence-Calibrated, Category-Aligned UMIX generates reliable training samples. • Achieves superior segmentation on ACDC, Synapse, and LA under low labels. [ABSTRACT FROM AUTHOR]
Copyright of Computerized Medical Imaging & Graphics is the property of Pergamon Press - An Imprint of Elsevier Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:Semi-supervised learning is crucial for medical image segmentation due to the scarcity of labeled data. However, existing methods that combine consistency regularization and pseudo-labeling often suffer from inadequate feature representation, suboptimal subnetwork disagreement, and noisy pseudo-labels. To address these limitations, this paper proposed a novel C onfidence- C alibrated C ontrastive M ean T eacher (C3MT) framework. First, C3MT introduces a Contrastive Learning-based co-training strategy, where an adaptive disagreement adjustment mechanism dynamically regulates the divergence between student models. This not only preserves representation diversity but also stabilizes the training process. Second, C3MT introduces a Confidence-Calibrated and Category-Aligned uncertainty-guided region mixing strategy. The confidence-calibrated mechanism filters out unreliable pseudo-labels, whereas the category-aligned design restricts region swapping to patches of the same semantic category, preserving anatomical coherence and preventing semantic inconsistency in the mixed samples. Together, these components significantly enhance feature representation, training stability, and segmentation quality, especially in challenging low-annotation scenarios. Extensive experiments on ACDC, Synapse, and LA datasets show that C3MT consistently outperforms recent state-of-the-art methods. For example, on the ACDC dataset with 20% labeled data, C3MT achieves up to a 4.3% improvement in average Dice score and a reduction in HD95 of more than 1.0 mm compared with strong baselines. The implementation is publicly available at https://github.com/l1654485/C3MT. [Display omitted] • Proposes C3MT: a Confidence-Calibrated contrastive mean teacher framework. • Contrastive Learning-based co-training stabilizes training and diversifies features. • Confidence-Calibrated, Category-Aligned UMIX generates reliable training samples. • Achieves superior segmentation on ACDC, Synapse, and LA under low labels. [ABSTRACT FROM AUTHOR]
ISSN:08956111
DOI:10.1016/j.compmedimag.2026.102721