a2z-1 for Multi-Disease Detection in Abdomen-Pelvis CT: External Validation and Performance Analysis Across 21 Conditions
This work addresses the need for automated multi-disease detection in medical imaging to assist radiologists, but it is incremental as it applies an existing AI method to new data and validation scenarios.
The study tackled the problem of detecting 21 time-sensitive and actionable findings in abdomen-pelvis CT scans using the AI model a2z-1, achieving an average AUC of 0.931 across conditions and 0.923 in external validation, with high performance for critical findings like small bowel obstruction (AUC 0.958) and acute pancreatitis (AUC 0.961).
We present a comprehensive evaluation of a2z-1, an artificial intelligence (AI) model designed to analyze abdomen-pelvis CT scans for 21 time-sensitive and actionable findings. Our study focuses on rigorous assessment of the model's performance and generalizability. Large-scale retrospective analysis demonstrates an average AUC of 0.931 across 21 conditions. External validation across two distinct health systems confirms consistent performance (AUC 0.923), establishing generalizability to different evaluation scenarios, with notable performance in critical findings such as small bowel obstruction (AUC 0.958) and acute pancreatitis (AUC 0.961). Subgroup analysis shows consistent accuracy across patient sex, age groups, and varied imaging protocols, including different slice thicknesses and contrast administration types. Comparison of high-confidence model outputs to radiologist reports reveals instances where a2z-1 identified overlooked findings, suggesting potential for quality assurance applications.