Large-scale esophageal cancer screening through noncontrast computed tomography and artificial intelligence

A deep-learning tool named Esophageal AI-Guided malignant Lesion Evaluation (EAGLE) detected esophageal cancer on routine noncontrast computed tomography with high sensitivity and specificity, validated across multiple diverse clinical settings in a large-scale study. Because noncontrast CT is already performed for many other indications, the approach could turn existing scans into an opportunistic screening channel for a cancer usually diagnosed late, without endoscopy or contrast agents. Consistent performance across different hospitals and scanner environments suggests the model generalizes beyond a single center, though prospective evaluation of downstream mortality benefit is still needed.