University of Washington’s Explainable AI Model Outperforms Benchmarks in Breast MRI Tumor Localization

Its pixel-level anomaly maps have enhanced radiologist interpretability in diverse patient cohorts