Graph confidence intercalibration network for intracranial aneurysm lesion instance segmentation in DSA

Open Webpage

Haili Ye, Yancheng Mo, Chen Tang, Mingqian Liao, Xiaoqing Zhang, Limeng Dai, Baihua Li, Jiang Liu

Published: 01 Jan 2025, Last Modified: 05 Nov 2025Displays 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•We proposed GCINet to address confidence bias and boundary errors in IA segmentation.•Introduces GCI and ESP modules to refine confidence and improve boundaries.•GCINet outperforms SOTA methods on IA-DSA and LiTS datasets with better accuracy.•Accurate IA segmentation improves surgery planning and enhances patient outcomes.