What question did this study set out to answer?

The aim is to improve ultrasound image segmentation by leveraging limited labeled data from a single center to enhance generalization across multi-center datasets.

February 28, 2026IEEE Journal of Biomedical and Health Informatics

Rethinking Multi-center Semi-supervised Breast Cancer Ultrasound Image Segmentation: An Intermediate-domain Perspective

View Full Paper

Ask AI

Bookmark

The aim is to improve ultrasound image segmentation by leveraging limited labeled data from a single center to enhance generalization across multi-center datasets.
Developed the Cross-domain Few-label Generalization (CFG) framework for better segmentation.
Created an Intermediate Domain Generator (IDG) to generate intermediate samples linking source and target domains.
Utilized the Swin Masked Autoencoder (MAE) to enhance image diversity by simulating noise during image acquisition.
Integrated the Kolmogorov-Arnold Network (KAN) with UNet for multi-scale feature perception.
Achieved a Kappa value of 77.17%, surpassing ten state-of-the-art methods.
Outperformed the second-best method by 0.78% over four different multi-center datasets.

Discussion

No takes yet. Share an insight, caveat, or question.

Implication

Multi-center semi-supervised segmentation shows improved model generalization in breast ultrasound images, suggesting an effective solution for diverse datasets.

Cite This Study

Mei et al. (2026) studied this question.

synapsesocial.com/papers/69a285da0a974eb0d3c00b8b https://doi.org/10.1109/jbhi.2026.3668217

View Full Paper

Ask AI

Bookmark

Implication

Multi-center semi-supervised segmentation shows improved model generalization in breast ultrasound images, suggesting an effective solution for diverse datasets.

Rethinking Multi-center Semi-supervised Breast Cancer Ultrasound Image Segmentation: An Intermediate-domain Perspective

Key Points

Discussion

Implication

Cite This Study

Rethinking Multi-center Semi-supervised Breast Cancer Ultrasound Image Segmentation: An Intermediate-domain Perspective

Key Points

Discussion

Implication

Cite This Study

Also Consider

Also Consider