AI on mammograms flags stroke, hypertension, and ischemic heart disease risk

Deep learning on routine mammograms detected stroke, hypertension, and ischemic heart disease in 30,000 women: diagnostic performance was highest for stroke (0.86), followed by hypertension (0.79) and ischemic heart disease (0.78), independent of age and cancer status.

• Trained on mammographic features associated with each condition using clinical and imaging data.
• Results remained consistent after accounting for cancer status and age.
• Could enable opportunistic cardiovascular risk assessment during routine breast cancer screening without additional imaging.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.