Uncertainty-guided mutual consistency learning for semi-supervised medical image segmentation

Yichi Zhang, Rushi Jiao, Qingcheng Liao, Dongyang Li, Jicong Zhang

Abstract: Highlights•A novel uncertainty-guided mutual consistency learning framework is proposed for semi-supervised medical image segmentation to alleviate the heavy burden of acquiring expert-examined annotations.•Incorporating both intra-task and cross-task consistency learning with the guidance of estimated segmentation uncertainty to efficiently utilize unlabeled data.•Extensive experiments on two benchmark datasets show that our proposed method outperforms existing semi-supervised segmentation methods with promising performance.