AI Model Using Mammograms Shows Higher Accuracy for 10-Year Breast Cancer Risk

An artificial intelligence model that processes digital mammograms produced more accurate 10-year breast cancer risk estimates than existing clinical models and the Mirai tool. Researchers tested the model on separate cohorts in the United States and Sweden. The results indicated stronger performance for identifying individuals who could benefit from primary prevention measures.

The model was applied to routine screening mammograms already collected during standard care. No additional imaging or patient procedures were required for the risk calculation. Performance comparisons showed higher accuracy than current clinical risk calculators and the earlier Mirai algorithm across both populations.

Health systems could use the model to flag patients for earlier or more intensive prevention steps based on calculated risk levels. Further validation studies would be needed before the model could be integrated into routine screening programs.