In addition to searching for breast cancer, artificial intelligence (AI) used in mammography reads may help to detect common cardiovascular diseases, according to study findings presented at the ESC Congress 2026.
“Because mammography is already widely used, analyzing the same images for cardiovascular information could potentially offer a scalable approach without requiring an additional imaging examination. Mammography also reaches many women in midlife, an important period for recognizing and addressing cardiovascular risk,” stated presenting author Viana Copeland, MBBS, of Chaim Sheba Medical Center, Tel Aviv University in Ramat Gan, Israel.
Study Methods
Researchers conducted a retrospective cohort study of women who had undergone at least one mammogram between 2011 and 2025 at a tertiary referral center (n = 29,921).
They developed a deep learning–based algorithm for detecting cardiovascular diseases from mammography issues and explored its diagnostic performance for detecting hypertension, ischemic heart disease, and cerebrovascular accident, which were defined using diagnoses extracted from electronic health records in addition to prescriptions, procedural findings, imaging findings, and in-hospital measurements.
“Despite being the leading cause of death in women worldwide, cardiovascular disease is consistently underdiagnosed and undertreated. A common finding in our medical center, and around the world, is that when women do seek medical help, their cardiovascular disease is already advanced. On the other hand, many women do attend routine breast cancer screening, even when they haven't sought care for cardiovascular symptoms. We investigated whether AI could help mammography serve an additional purpose in this group—the early detection of cardiovascular disease—enabling preventive strategies to be implemented,” Dr. Copeland said.
The model architecture consisted of a convolutional neural network that predicted the presence of hypertension, ischemic heart disease, or cerebrovascular accident.
Performance of the AI model was tested using area under the curve and receiver operating characteristic curves for each of the cardiovascular diseases.
Key Findings
Among the women included in the analysis, 18% had breast cancer. Patients were followed for a median of 7.3 years (interquartile range = 4.0–11.0 years).
The AI model detected hypertension in 16% of women, ischemic heart disease in 2.5%, and cerebrovascular accident in 2.5%.
The algorithm achieved an area under the curve of 0.79 for hypertension detection, 0.78 for ischemic heart disease detection, and 0.86 for cerebrovascular accident detection.
In sensitivity analyses, results were consistent, but showed an improved performance for mediolateral oblique views, with areas under the curve improving to 0.80 for both hypertension and ischemic heart disease and 0.88 for cerebrovascular accident.
Going forward, the researchers are planning to improve the model's accuracy and reduce the rates of false positives and false negatives. They also plan to explore if mammograms may be able to also detect other cardiovascular conditions.
DISCLOSURES: For full disclosures of the study authors, visit esc365.escardio.org.
