SELF-REINFORCING FOR FEW-SHOT MEDICAL IMAGE SEGMENTATION

Few-shot semantic segmentation is usually a solution to data scarcity, so it has great potential in the field of medical images. For this work, we propose a simple and effective foreground-background self-reinforcing segmentation method for medical image few-shot segmentation, self-reinforcing prototype network(SRPNet). In the case of no additional introduction of prior knowledge, the model can autonomously adjust the segmentation effect of foreground and background, so that the model can better segment the unseen class. Experiments on CT dataset and MRI dataset show that the proposed method outperforms other advanced methods.