A Semi-Supervised Framework for Breast Ultrasound Segmentation with Training-Free Pseudo-Label Generation and Label Refinement

A Semi-Supervised Framework for Breast Ultrasound Segmentation with Training-Free Pseudo-Label Generation and Label Refinement explores A semi-supervised breast ultrasound segmentation framework leveraging vision-language models for training-free pseudo-label generation, achieving near fully-supervised performance with minimal labeled data, enabling scalable medical image segmentation.. Commercial viability score: 8/10 in Medical AI.