AI for NEC on Abdominal Radiographs: Meta-Analysis Shows Moderate Accuracy

Artificial intelligence interpretation of abdominal radiographs for necrotising enterocolitis (NEC) achieves pooled sensitivity of 0.78 (95% CI 0.67-0.85) and specificity of 0.84 (95% CI 0.73-0.91). The meta-analysis suggests a role as a decision-support adjunct rather than a st…

• Systematic review and meta-analysis of 10 retrospective studies; 6 were pooled, all using convolutional neural networks with limited external validation.
• Positive and negative likelihood ratios were 4.84 and 0.27, respectively, indicating moderate diagnostic value; explainability analyses highlighted clinically relevant bowel features.
• Key limitation: All included studies were retrospective with substantial heterogeneity and a lack of prospective, external validation.

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