03 May 2026
May. 2026. Our paper, BECM-Net: A Multi-granularity Collaborative Framework for Semi-Supervised Fetal Ultrasound Segmentation, has been accepted in the IEEE Journal of Biomedical and Health Informatics.
This work develops a multi-granularity collaborative framework for semi-supervised fetal ultrasound segmentation, with the goal of improving performance in challenging regions with blurred or low-contrast anatomical boundaries. BECM-Net integrates pixel-level, region-level, and structure-level learning to provide more reliable supervision under limited annotation settings, and demonstrates strong performance on fetal ultrasound datasets.
We are grateful to all co-authors for their hard work and contributions to this study. Congratulations to Wei Hu, Cong Tan, Wendong Wang, Zeheng Wang, Qibing Qin, and Wenfeng Zhang on this achievement.