Mammography AI Spots Vascular Signals - Summary - MDSpire

Mammography AI Spots Vascular Signals

  • By

  • Kathryn Wighton

  • March 24, 2026

  • 3 min

Share

Objective:

To evaluate the prognostic value of automated quantification of breast arterial calcification (BAC) on screening mammography for predicting major adverse cardiovascular events (MACE) and mortality.

Approach:
    Key Findings:
    • BAC detected in 16.1% of Emory cohort and 20.6% of Mayo cohort.
    • Event rates for MACE increased with BAC severity: 5.96 per 1,000 person-years (no BAC) to 48.89 (severe BAC).
    • Each 1 mm² increase in BAC area correlated with a 1% to 2% increase in cardiovascular risk.
    • BAC remained an independent predictor of MACE after adjusting for the PREVENT score.
    • Adding BAC to the PREVENT model improved discrimination in both cohorts.
    Interpretation:

    Automated BAC quantification from routine mammography may serve as an effective cardiovascular risk assessment tool for women, enhancing existing risk models without additional radiation exposure.

    Limitations:
    • Study was retrospective and dependent on electronic health record data.
    • Missing variables such as menopause status and reproductive history.
    • PREVENT scores could only be calculated for a subset of patients due to incomplete clinical data.
    Conclusion:

    Automated BAC quantification could provide significant prognostic value in assessing cardiovascular risk in women undergoing routine mammography.

    Sources:

Original Source(s)

Related Content