Our aim was to evaluate the performance of machine learning (ML), integrating clinical parameters with coronary artery calcium (CAC), and automated epicardial adipose tissue (EAT) quantification, for the prediction of long-term risk of myocardial infarction (MI) and cardiac death in asymptomatic subjects. Our study included 1912 asymptomatic subjects [1117 (58.4%) male, age: 55.8 9.1 years] from the prospective EISNER trial with long-term follow-up after CAC scoring. EAT volume and density were quantified using a fully automated deep learning method. ML extreme gradient boosting was trained using clinical co-variates, plasma lipid panel measurements, risk factors, CAC, aortic calcium, and automated EAT measures, and validated using repeated 10-fold cross validation. During mean follow-up of 14.5 2 years, 76 events of MI and/or cardiac death occurred. ML obtained a significantly higher AUC than atherosclerotic cardiovascular disease (ASCVD) risk and CAC score for predicting events (ML: 0.82; ASCVD: 0.77; CAC: 0.77, P < 0.05 for all). Subjects with a higher ML score (by Youden’s index) had high hazard of suffering events (HR: 10.38, P < 0.001); the relationships persisted in multivariable analysis including ASCVD-risk and CAC measures (HR: 2.94, P = 0.005). Age, ASCVD-risk, and CAC were prognostically important for both genders. Systolic blood pressure was more important than cholesterol in women, and the opposite in men. In this prospective study, machine learning used to integrate clinical and quantitative imaging-based variables significantly improves prediction of MI and cardiac death compared with standard clinical risk assessment. Following further validation, such a personalized paradigm could potentially be used to improve cardiovascular risk assessment.
1 March 2016 in Journal of Cardiovascular Computed Tomography
There are significant sex-specific differences in left ventricular ejection fraction (LVEF), with a higher LVEF being observed in women. We sought to assess the clinical relevance of an increased LVEF in women and men. A total of 4632 patients from the CONFIRM (COronary CT Angiography EvaluatioN For Clinical Outcomes: An InteRnational Multicenter) registry (44.8% women; mean age 58.7 ± 13.2 years in men and 59.5 ± 13.3 years in women, P = 0.05), in whom LVEF was measured by cardiac computed tomography, were categorized according to LVEF (low 65%). The prevalence of high LVEF was similar in both sexes (33.5% in women and 32.5% in men, P = 0.46). After 6 years of follow-up, no difference in mortality was observed in patients with high LVEF in the overall cohort (P = 0.41). When data were stratified by sex, women with high LVEF died more often from any cause as compared to women with normal LVEF (8.6% vs. 7.1%, log rank P = 0.032), while an opposite trend was observed in men (5.8% vs. 6.8% in normal LVEF, log rank P = 0.89). Accordingly, a first order interaction term of male sex and high LVEF was significant (hazard ratios 0.63, 95% confidence intervals 0.41–0.98, P = 0.043) in a Cox regression model of all-cause mortality adjusted for age, cardiovascular risk factors, and severity of coronary artery disease (CAD). Increased LVEF is highly prevalent in patients referred for evaluation of CAD and is associated with an increased risk of death in women, but not in men. Differentiating between normal and hyperdynamic left ventricles might improve risk stratification in women with CAD. https://clinicaltrials.gov/ct2/show/NCT01443637.