Extremity fracture AI yields 99.1% PPV for high-confidence detections in 2,508 patients

Artificial intelligence (AI) detected extremity fractures in 2,508 patients with 92.7% sensitivity and 94.4% specificity. High-confidence detections had 99.1% PPV vs 59.1% for low-confidence; children had higher PPV (91.7%) than adults (86.1%).

• Per-fracture sensitivity 92.7% (95% CI 90.9–94.5), specificity 94.4% (93.2–95.4); NPV 96.6%.
• PPV was 99.1% for high-confidence vs 59.1% for low-confidence detections (p<0.001).
• Children had higher per-fracture PPV than adults (91.7% vs 86.1%, p=0.01), with no significant difference in sensitivity.

Automated summary

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