The Rise of Risk-Reading Mammograms

Clairity Breast uses routine scans to forecast five-year cancer risk as hospitals assess cost, workflow, and clinical impact

25 Feb 2026

Artificial intelligence is moving beyond detecting disease to forecasting it, as US hospitals begin deploying tools that estimate future cancer risk from routine scans.

Beth Israel Deaconess Medical Center has started clinical use of Clairity Breast, an AI platform designed to predict a woman’s likelihood of developing breast cancer within five years using a standard mammogram. The system was trained on more than 400,000 mammograms linked to five-year outcomes, one of the largest datasets used for a predictive imaging model in this field.

Unlike conventional software that identifies visible tumours, Clairity analyses subtle pixel-level patterns that may indicate elevated future risk. The technology is integrated into existing imaging systems and does not require new hardware or additional appointments. Risk scores are produced during routine screenings, allowing clinicians to incorporate predictive data into established workflows.

The platform received De Novo authorisation from the US Food and Drug Administration, creating a new regulatory category for AI tools focused on risk prediction rather than diagnosis. The clearance marks a regulatory milestone and may reduce uncertainty for hospitals considering adoption.

More than 35mn mammograms are performed annually in the US. If predictive scoring proves reliable at scale, it could influence how health systems allocate advanced imaging, preventive therapies and specialist referrals. Even modest gains in identifying higher-risk patients may affect screening intervals and resource planning.

However, financial and operational questions remain. Reimbursement pathways for predictive AI tools are still developing, and hospital uptake is likely to depend on clinical validation as well as workflow adjustments, patient communication standards and alignment with insurers.

The emergence of risk-based imaging reflects a broader shift in AI-enabled diagnostics, from identifying existing disease to stratifying future risk. How quickly such systems become part of routine care will depend on regulatory clarity, payer support and evidence that early risk identification improves long-term outcomes.