EndoDINO: A Foundation Model for GI Endoscopy

TL;DR

EndoDINO is a large-scale foundation model for gastrointestinal endoscopy that demonstrates strong generalization and state-of-the-art performance across multiple clinical tasks by pre-training on extensive curated image datasets.

Contribution

This work introduces EndoDINO, a foundation model pre-trained on the largest GI endoscopy image dataset, achieving high performance with simple decoder heads.

Findings

State-of-the-art performance in anatomical landmark classification

Superior results in polyp segmentation

Effective Mayo endoscopic scoring with minimal additional training

Abstract

In this work, we present EndoDINO, a foundation model for GI endoscopy tasks that achieves strong generalizability by pre-training on a well-curated image dataset sampled from the largest known GI endoscopy video dataset in the literature. Specifically, we pre-trained ViT models with 1B, 307M, and 86M parameters using datasets ranging from 100K to 10M curated images. Using EndoDINO as a frozen feature encoder, we achieved state-of-the-art performance in anatomical landmark classification, polyp segmentation, and Mayo endoscopic scoring (MES) for ulcerative colitis with only simple decoder heads.