Assessment of the Severity of Breast Artery Calcification on a Mammogram: Intraoperator and Interoperator Reproducibility

Abstract

Purpose. To study approaches to the quantitative assessment of breast arterial calcification (BAC) – a new surrogate marker of high cardiovascular risk in women, to determine the most detailed way to quantify BAC and to assess the reproducibility of these parameters.Material and methods. Mammograms of 1,078 women were analyzed for the presence of BAC. The intraoperator reproducibility of the assessment of the severity of BAC using a 12-point scale (Margolies L et al., 2016) was studied by analyzing 20 mammograms by the same operator twice with an interval of at least 2 weeks. Inter-operator reproducibility was studied by analyzing 99 mammograms by two independent operators.Results. When assessing the intraoperative reproducibility of the total score for each mammary gland, the exact coincidence of the results was noted in 70% (95% confidence interval [CI] 53.5-83.4), in cases of difference of no more than 1 point – in 27.5% (95%CI 14.6-43.9), only in 1 case the difference in assessments was 2 points. No systematic error was found between the two measurements (p=1.0), the correlation coefficient was rs=0.973. The assessment of inter-operator reproducibility showed that the exact coincidence of indicators was present in 48.5% (95%CI 41.3-55.7), in 91.4% (95% CI 86.6-94.9) cases, the total score for each the mammary gland differed by no more than 1 point. There was no systematic error between the measurements of the two experts (p=0.438), the correlation coefficient was rs=0.942.Conclusion. A good intraoperator and interoperator reproducibility of indicators of the severity of BAC on a 12-point scale has been shown, which makes it possible to recommend it for use in science and practice.