Search papers, labs, and topics across Lattice.
This paper introduces EP-SAM, an edge-aware and prompt-enhanced adaptation of the Segment Anything Model (SAM) specifically designed for ultrasound image segmentation. By leveraging multi-block feature extraction and edge-aware supervision, EP-SAM improves the model's ability to delineate anatomical structures and lesions, overcoming challenges related to contour ambiguity and speckle noise. Experimental results show that EP-SAM significantly outperforms existing SAM-based methods across multiple benchmarks, highlighting its effectiveness in enhancing segmentation quality in ultrasound imaging.
EP-SAM outperforms traditional SAM methods by effectively addressing contour ambiguity in ultrasound images through innovative edge-aware supervision.
Ultrasound image segmentation is essential for delineating anatomical structures and lesions, providing the foundation for accurate diagnosis. While the Segment Anything Model (SAM) has demonstrated remarkable success on natural images, its performance on ultrasound data is often hindered by poor boundary delineation. To address this limitation, we propose EP-SAM, an edge-aware and prompt-enhanced adaptation of SAM. Specifically, we leverage multi-block feature extraction from the image encoder to enrich coarse-to-fine semantic representations, while edge-aware supervision of the image encoder improves robustness to contour ambiguity and speckle noise. By integrating these complementary cues, EP-SAM generates high-quality prompts that effectively guide the model toward target regions of interest. Experimental results on multiple benchmarks demonstrate that EP-SAM consistently outperforms existing SAM-based methods.
