Cardiovascular disease remains a major cause of death among women, while coronary artery calcification often appears later in women than in men and is frequently linked to worse outcomes after myocardial infarction. Breast arterial calcifications are common incidental findings on routine screening mammograms and usually appear as linear calcifications in mammary arteries. A 2026 analysis published in the Journal of Breast Imaging examined whether visual scoring of breast arterial calcifications on mammography could help assess coronary artery calcifications seen on cardiac CT. The evaluation compared five manual visual scoring systems in 90 women who had both mammography and cardiac CT, focusing on reproducibility, diagnostic accuracy and the additional reading time required in routine imaging practice.
Visual Scoring Systems Compared
The cohort included adult female patients from University Hospital Zurich who had cardiac CT for coronary calcium assessment and routine screening mammography. One patient from the initial group was excluded because CT data were incomplete. Most women had mammography of both breasts, while a small number had unilateral imaging because the other breast had been removed. Standard mammographic views were assessed, while tomosynthesis and digitally reconstructed mammography were not included.
Five visual approaches were compared. The simplest method classified breast arterial calcification as either present or absent. A vessel count scale graded cases according to the number of affected vessels. A severity scale graded calcifications from absent or minimal findings to more pronounced disease. Another scale combined severity with vessel count. The most complex system incorporated the number of vessels involved, calcification length and density.
The aim was not only to test whether breast arterial calcification related to coronary calcium, but also whether each method could be applied consistently and efficiently. This distinction is important for clinical feasibility, as a scoring system must be reliable between readers and practical enough for mammography workflows.
Must Read: Mammography-Linked Lung Screening Increases Uptake
Correlation With Coronary Calcium
Breast arterial calcification was present in a minority of women. Coronary calcification was more common among women with breast arterial calcification than among those without it, showing a trend toward an association. Women with coronary calcification were also older at mammography than those without coronary calcification. Cardiac CT showed differences between women with and without breast arterial calcification in the overall coronary calcium burden and in the number of affected coronary vessels.
Several scoring systems showed a statistically significant relationship with coronary calcium measured by the Agatston score. The vessel count scale had the strongest positive correlation, although the association remained modest. The simple present-or-absent method, the severity scale and the more complex system using vessel count, length and density also showed significant correlations. The combined severity and vessel count scale did not show a significant correlation.
Diagnostic accuracy was limited across all methods. The severity-based scale achieved the highest overall performance, but differences between the scoring systems were not statistically significant. Sensitivity was generally low, while specificity was higher. This pattern suggests that visual breast arterial calcification scoring may identify some women with coronary calcification, but it cannot be treated as a sensitive standalone method for detecting coronary disease.
Workflow Feasibility and Reader Agreement
Reader agreement was high across the evaluated systems. The combined severity and vessel count scale showed the highest agreement, closely followed by the vessel count scale and the more complex system. The severity-only scale also performed strongly. Even the simplest present-or-absent method produced substantial agreement, although it provided less detail than the multilevel systems.
Reading time increased as scoring became more complex. The present-or-absent approach required the least time. The severity-only scale also remained quick. The vessel count scale took longer but still remained practical, with an average reading time of about half a minute per patient. The combined scale required a similar amount of time. The system incorporating vessel count, length and density required the most time and added complexity without a clear performance advantage.
The vessel count scale offered the most favourable balance. It combined strong reader agreement, the clearest correlation with coronary calcium and manageable reading time. Its simplicity may make it more suitable for routine mammography than more detailed models, while still providing more information than a binary assessment. The findings also indicate that structured classification can be applied consistently by readers with different levels of experience.
Visual scoring of breast arterial calcifications on mammography is feasible and reproducible across several manual systems. Overall diagnostic performance remains modest, with low sensitivity and comparatively higher specificity. The four-level vessel count scale appears to offer the best balance between accuracy, reproducibility and practicality. Interpretation remains limited by the retrospective single-institution design, small sample size, limited number of women with breast arterial calcification, time interval between imaging tests and absence of recorded major cardiac events during follow-up. Further evaluation is needed before structured reporting can support mammography-based cardiovascular risk assessment.
Source: Journal of Breast Imaging
Image Credit: iStock
References:
Saenger JA, Happe J, Kerber B et al. (2026) Breast Arterial Calcifications in Mammography: Evaluating Diagnostic Accuracy, Reproducibility, and Clinical Feasibility of Visual Scoring Systems. Journal of Breast Imaging, 8(2):149–159.
