Multinational study of AI-assisted versus AI-unassisted identification in magnetic resonance images

Artificial intelligence (AI) assistance in magnetic resonace imaging (MRI) assessment for prostate cancer shows promise for improved diagnostic accuracy but lacks large-scale observational evidence. A diagnostic study at 53 centers across 17 countries included 61 readers and 360 prostate MRI examinations among 360 men with suspicion of prostate cancer. AI assistance was associated with a statistically significant improvement in detecting clinically significant prostate cancer (csPCa), increasing the area under the receiver operating characteristic curve, sensitivity, and specificity compared with unassisted readings. The authors conclude that these findings suggest a potential added value of AI assistance during radiologic assessments to improve the diagnostic accuracy of csPCa.