SAM-EG: Segment Anything Model with Egde Guidance Framework for Efficient Polyp Segmentation

Polyp segmentation, a critical concern in medical imaging, has promptednumerous proposed methods aimed at enhancing the quality of segmented masks.While current state-of-the-art techniques produce impressive results, the sizeand computational cost of these models pose challenges for practical industryapplications. Recently, the Segment Anything Model (SAM) has been proposed as arobust foundation model, showing promise for adaptation to medical imagesegmentation. Inspired by this concept, we propose SAM-EG, a framework thatguides small segmentation models for polyp segmentation to address thecomputation cost challenge. Additionally, in this study, we introduce the EdgeGuiding module, which integrates edge information into image features to assistthe segmentation model in addressing boundary issues from current segmentationmodel in this task. Through extensive experiments, our small models showcasetheir efficacy by achieving competitive results with state-of-the-art methods,offering a promising approach to developing compact models with high accuracyfor polyp segmentation and in the broader field of medical imaging.