Researchers have developed an artificial intelligence (AI) model that can analyze mammograms to identify women at higher risk of heart disease, strokes, and high blood pressure, potentially transforming routine breast cancer screenings into dual-purpose cardiovascular assessments. The findings were presented in August 2026 at the European Society of Cardiology annual meeting in Munich by scientists from Tel Aviv University.
The study involved retrospective analysis of 97,364 mammogram scans from 29,364 women, with an average age of 54. Medical records indicated that 16 percent of these women had hypertension, while 2.5 percent each had coronary heart disease or had experienced a stroke. The AI model was trained to detect indicators of these conditions through mammographic images alone.
According to the researchers, the AI successfully assigned a high probability of hypertension and coronary heart disease approximately 79 percent of the time and stroke 86 percent of the time. The research team emphasized ongoing efforts to improve the model’s accuracy with the aim of enabling earlier diagnosis of cardiovascular diseases in women.
Coronary heart disease remains the leading cause of death among women worldwide. In Australia, for example, it accounted for 6,122 female deaths in 2024—almost twice the number of deaths from breast cancer in the same year. Despite its prevalence, cardiovascular disease in women is frequently underrecognized and undertreated, with diagnoses often occurring at later stages compared to men.
Experts note that women have historically been underrepresented in cardiology research and face challenges in timely detection and treatment of heart-related conditions. Sonya Babu-Narayan, clinical director at the British Heart Foundation, highlighted that the persistence of the misconception that heart disease is primarily a men’s issue contributes to this disparity. She described the new AI approach as a promising development that could improve cardiovascular screening by integrating heart health assessments into established breast cancer screening programs.
Viana Copeland of Tel Aviv University, who presented the study, underscored that many women currently living with undiagnosed cardiovascular conditions might benefit from this non-invasive, scalable screening method. Using mammograms, which millions of women undergo routinely, for heart disease detection could help address the gap in early diagnosis without the need for additional imaging.
While the results are encouraging, researchers stress that further validation and refinement are necessary before the AI model can be implemented widely in clinical settings. If successful, this technology could enhance preventive care by identifying at-risk women earlier, thereby improving treatment outcomes for cardiovascular disease as well as breast cancer.
