Medical Imaging

This article presents a novel approach that leverages SAM 2 video segmentation capabilities to reduce the number of prompts needed to segment a complete volume of medical images. The study first compares the performance of SAM and SAM 2 in medical image segmentation. Then, our novel approach is presented. The results show that SAM 2 achieves an average improvement of 1.76% in Jaccard Index and 1.49% in Dice Score compared to SAM.

Segment Anything Model (SAM), developed by Meta, demonstrates exceptional performance in medical image segmentation, particularly for lung CT scans and chest X-rays, achieving an average Jaccard index of 91.45% and an average Dice score of 94.95% with minimal user intervention. In particular, SAM exceeds the 70% threshold recommended in the literature and rivals state-of-the-art models explicitly designed for medical segmentation.