Annotation-Efficient Task Guidance for Medical Segment Anything

Medical image segmentation is a key task in the imaging workflow, influencing many image-based decisions. Fully-supervised segmentation models rely on large amounts of labeled training data, typically obtained through manual anno-tation, which can be an expensive, time-consuming, and error-prone process. This signals a need for accurate, automatic, and annotation-efficient methods to train these models. We propose SAM-Mix, a novel multitask learning framework for medical image segmentation that uses class activation maps produced by an auxiliary classifier to guide predictions of the semi-supervised segmentation branch. Experimental evaluations on the public LiTS and TotalSegmentator datasets demonstrate the effectiveness of SAM-Mix in segmenting liver from abdominal computed tomography (CT) images. Our code is available at https://github.com/tbwa233/SAM-Mix.