Dr. Oluwadamilola Oladeru, Assistant Professor of Radiation Oncology at the University of Florida College of Medicine, recently discussed the significant potential of Artificial Intelligence (AI) to improve breast cancer screening. She highlighted the current limitations of traditional mammography, specifically the occurrence of false positives and negatives. This discussion centered on a deep learning AI model developed by Google Health and tested by the National Cancer Institute (NCI, with its findings published back in 2020, aiming to enhance early detection and address health disparities, particularly among women of color.
The AI algorithm analyzes mammograms to detect subtle signs of cancer, which can reduce the reliance on human interpretation and potentially decrease the need for second opinions and biopsies. The Google Health/NCI study demonstrated promising results, showing a 5.7% reduction in false positives in the U.S. and a 1.2% reduction in the U.K. Furthermore, the model achieved a 9.4% reduction in false negatives in the U.S. and a 2.7% reduction in the U.K., indicating improved accuracy in detecting actual cancer while reducing unnecessary follow-ups.
This advanced AI technology offers benefits such as reduced workload for radiologists and improved overall screening accuracy. Dr. Oladeru specifically underscored its importance for women of color, who disproportionately face disparities in healthcare access and outcomes for breast cancer. The ongoing adoption and integration of such AI tools into clinical practice are expected to contribute significantly to more equitable and effective breast cancer detection.
