Enhancing generalization of medical image segmentation via game theory-based domain selection

Zuyu Zhang, Yan Li, Byeong-Seok Shin

Abstract: Highlights•Introduces a game-theoretic approach to enhance medical image segmentation generalization.•Proposes a Beta distribution-based domain selection optimized via reinforcement learning.•Achieves balanced performance across domains by reaching a Nash equilibrium during training.

External IDs:doi:10.1016/j.jbi.2025.104802