Performance of an Artificial Intelligence Support System on Screening Mammography Cases Proceeding to Stereotactic Biopsy

Simple Summary Digital breast tomosynthesis (DBT) for screening mammography is widely adopted in the United States and deep learning-based artificial intelligence (AI) algorithms have been developed to increase accuracy and efficiency in breast cancer detection. Few studies have evaluated the performance of AI decision support in real-world settings, especially on DBT exams. Transpara AI decision support was recently integrated into the clinical workflow for screening mammography at our institution and our aim was to assess its standalone performance in detecting breast cancer on DBT cases proceeding to stereotactic biopsy. This article presents performance metrics for Transpara AI along with representative cases of AI misses and false positives to provide insights into how radiologist–AI synergy can improve patient care.