INTELCAPE: A Deep Learning-Powered System for Automated, High-Accuracy Crohn's Disease Diagnosis via Capsule Endoscopy

This study presents INTELCAPE, a multi-task deep learning system for automated capsule endoscopy analysis in Crohn's disease. Built and evaluated on 872 videos from two Chinese hospitals, the pipeline segments small-intestine regions, detects suspicious lesions, and performs video-level diagnosis with strong cross-center generalizability. INTELCAPE achieved AUCs of 0.982 and 0.984 for Crohn's disease diagnosis, reached 90% diagnostic accuracy comparable to specialists while operating around 10 times faster, and improved clinicians' accuracy from 76.7% to 94.8% while reducing interpretation time from 67.9 to 22.5 minutes. The work highlights the practical value of AI-assisted capsule endoscopy as a decision-support tool for faster and more standardized Crohn's disease diagnosis.