GLCV-NET: An automatic diagnosis system for advanced liver fibrosis using global-local cross view in B-mode ultrasound images

Open Webpage

Bianzhe Wu, Ze-Rong Huang, Jinglin Liang, Hong Yang, Wei Wang, Shuangping Huang, Li-Da Chen, Qinghua Huang

Published: 01 Jan 2024, Last Modified: 12 May 2025Comput. Methods Programs Biomed. 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•GLCV-Net innovatively diagnoses advanced liver fibrosis using integrated global–local B-mode ultrasound data.•Introduced Segmentation-Enhanced Global Hybrid Feature Extractor for comprehensive global liver analysis.•Developed Heatmap-weighted Local Feature Extractor for automatic capture of local details in liver regions.•Proposed Scale-adaptive Fusion Module for integrating features across various scales in liver regions.•GLCV-Net validated on 1,000+ patients, showcasing enhanced diagnostic accuracy and clinical decision-making potential.