Segment Anything in Medical Images

TL;DR

MedSAM is a large-scale foundation model for universal medical image segmentation, trained on over 1.5 million image-mask pairs across multiple modalities and cancer types, showing superior accuracy and robustness.

Contribution

This work introduces MedSAM, a novel foundation model that generalizes across diverse medical imaging modalities and diseases, unlike prior specialized segmentation methods.

Findings

Outperforms modality-specific models in accuracy and robustness

Demonstrates versatility across 86 internal and 60 external validation tasks

Enables faster, more personalized diagnostic and treatment planning

Abstract

Medical image segmentation is a critical component in clinical practice, facilitating accurate diagnosis, treatment planning, and disease monitoring. However, existing methods, often tailored to specific modalities or disease types, lack generalizability across the diverse spectrum of medical image segmentation tasks. Here we present MedSAM, a foundation model designed for bridging this gap by enabling universal medical image segmentation. The model is developed on a large-scale medical image dataset with 1,570,263 image-mask pairs, covering 10 imaging modalities and over 30 cancer types. We conduct a comprehensive evaluation on 86 internal validation tasks and 60 external validation tasks, demonstrating better accuracy and robustness than modality-wise specialist models. By delivering accurate and efficient segmentation across a wide spectrum of tasks, MedSAM holds significant potential to expedite the evolution of diagnostic tools and the personalization of treatment plans.