Breast Cancer Detection At Air Force Medical Facilities Using Deep Learning Software

More than 39 million mammograms are performed every year to achieve early detection of breast cancer and have helped improving its mortality and morbidity. However, interpretation of images remains challenging, which affects diagnosis. We have developed an AI-based software that aids radiologists during interpretation and enables high quality screening to be achieved at any clinic, and it is now in evaluation use at non-defense sites. By automatically detecting the presence, or absence, of suspicious lesions in 2D and 3D mammograms, our software can help improve cancer detection rates, decrease unnecessary recalls, and catch cancers earlier. We seek to identify pilot Air Force collaboration sites for this solution.