Graph- and transformer-guided boundary aware network for medical image segmentation

Shanshan Xu, Lianhong Duan, Yang Zhang, Zhicheng Zhang, Tiansheng Sun, Lixia Tian

Abstract: Highlights•Research highlights 1GTBA-Net is designed to tackle the challenges of complex backgrounds, target-irrelevant noise, and ambiguous boundaries in medical image segmentation.•Research highlights 2Global Feature Aggregation modules employ an efficient self-attention mechanism to facilitate target localization from complex backgrounds.•Research highlights 3Graph-based Dynamic Feature Fusion modules utilize graph attention network to suppress target-irrelevant noises and preserve spatial details.•Research highlights 4Uncertainty-based Boundary Refinement modules introduce an uncertainty quantification strategy and auxiliary loss to discern ambiguous boundaries.