AI Trained on 400,000 Abdominal CTs Boosts Radiologist Sensitivity by ~10%

A vision-language AI trained on over 400,000 abdominal CTs increased diagnostic sensitivity of 26 radiologists by ~10% in a reader study, achieving high performance across 146 imaging findings.

• Trained on >400,000 CTs and 15 million image-text pairs without manual annotation.
• Evaluated across 18 anatomical structures and 146 imaging findings with robust generalization.
• Internal and external validation across multiple centers and clinical scenarios.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.