Consistency label-activated region generating network for weakly supervised medical image segmentation

Wei Du, Yongkang Huo, Rixin Zhou, Yu Sun, Shiyi Tang, Xuan Zhao, Ying Li, Gaoyang Li

Abstract: Highlights•The proposed model can generate lesion masks directly with the image-level label.•We design a one-phase weakly supervised medical image segmentation framework.•We design CBFNet, which can adjust lesion mask boundaries and generate fake images.•The proposed model can achieve state-of-the-art performance on various datasets.