State of abdominal CT datasets: A critical review of bias, clinical relevance, and real-world applicability

Résumé fourni par la source

This review systematically searched publicly available abdominal CT datasets and critically evaluates suitability for artificial intelligence (AI) applications in clinical settings. We examined 45 publicly available abdominal CT datasets (47,049 studies). Across all 45 datasets, we found substantial redundancy (51% case reuse) and a Western/geographic skew (75.3% from North America and Europe). A bias assessment was performed on the 22 datasets with more than 100 cases; within this subset, the most prevalent high-risk categories were racial bias (with a score of 16 out of 22) and selection bias (with a score of 15 out of 22), both of which may undermine model generalizability across diverse healthcare environments—particularly in resource-limited settings. To address these challenges, we propose targeted strategies for dataset improvement, including multi-institutional collaboration, adoption of standardized protocols, and deliberate inclusion of diverse patient populations and imaging technologies. These efforts are crucial in supporting the development of more equitable and clinically robust AI models for abdominal imaging.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.