AI Should Read Mammograms Only When Confident: A Hybrid Breast Cancer Screening Reading Strategy

Sarah D. Verboom, Jaap Kroes, Santiago Pires, Mireille J. M. Broeders, Ioannis Sechopoulos · Radiology · 2025

Leveraging the prediction uncertainty of an artificial intelligence mammography interpretation model can reduce the workload in screen reading without decreasing performance by considering the model’s interpretations only when it is confident.

Read the paper · More papers on PaperTik