FDA Approves AI Tool for Breast Cancer Risk Prediction Using Mammograms

The U.S. Food and Drug Administration (FDA) has recently approved the first artificial intelligence (AI) tool designed to predict breast cancer risk, developed by the digital health technology company Clairity. This innovative platform, known as Clairity Breast, utilizes images from routine screening mammograms to forecast a woman’s five-year risk of developing breast cancer. Clairity has plans to introduce this AI technology across healthcare systems by 2025. Traditionally, breast cancer risk assessments have heavily relied on factors such as age and family history.

However, statistics reveal that about 85% of breast cancer cases occur in women who have no familial ties to the disease. This is often attributed to genetic mutations related to aging. Additionally, existing risk models have predominantly focused on data from European Caucasian populations, limiting their applicability to a more diverse demographic. The Clairity Breast platform analyzes subtle image features from mammograms that correlate with breast cancer risk and produces a validated five-year risk score for healthcare providers.

Dr. Connie Lehman, the founder of Clairity and a breast imaging expert, emphasized the significance of mammograms in early cancer detection and remarked on the power of AI and computer vision to reveal insights that are often invisible to the human eye. Experts from the American Cancer Society, including Dr. Robert A. Smith, have highlighted the importance of personalized, risk-based screening to enhance breast cancer outcomes. With the rise of breast cancer cases, particularly among younger women, AI-tools like Clairity’s are expected to play a crucial role in ensuring timely and accurate care. Dr. Marc Siegel, a senior medical analyst, praised Clairity’s development as a significant advancement in the field, noting that AI can enhance the accuracy of mammogram evaluations, especially in underserved areas.

He emphasized that while AI is becoming integral to cancer detection, it is not intended to replace human expertise.