The potential role of radiochemical analysis of CT images of epicardial adipose tissue in the prognosis of acute myocardial infarction

Introduction. Multispiral computed tomography (MSCT) coronary angiography (CAG) is a highly informative method of visualizing atherosclerotic plaques in the coronary arteries and assessing their structure. At the same time, this method has a few significant drawbacks associated with the intravenous administration of iodine-containing radiopaque agents as well as high radiation exposure. The radiomic analysis of contrast-free MSCT images allows calculating many additional quantitative parameters, which can potentially be associated with atherosclerotic plaque instability and the degree of coronary artery stenosis. At the same time, the prognostic and diagnostic value of radiomic characteristics has not been investigated.Aim: To assess whether there is an association between radiomic indexes of EAT on non-contrast MSCT cardiac images with the degree of atherosclerotic coronary artery stenosis in patients with stable CAD, as well as the incidence of acute coronary syndrome (ACS) within 5 years in this category of patients.Material and Methods. We retrospectively reviewed 100 MSCT-CAG studies performed to diagnose coronary heart disease in patients. 39 patients with signs of coronary stenosis up to 50% and registered in Tomsk medical information systems (MIS) for at least 5 years were selected, as well as 15 people without signs of coronary arteries (CA) atherosclerosis as a control group. Epicardial adipose tissue (EAT) volume was assessed and 837 radiomic characteristics were calculated on non-contrasted MSCT cardiac images of all patients (54 people). The presence or absence of ACS within 5 years after MSCT-CAG in each patient was monitored according to Tomsk MIS data. Statistical analysis and comparison of indices were performed in control group (group 2) and study group (group 1), as well as in subgroups of patients who had suffered AMI (group 1a) and those who had not (group 1b).Results. When comparing group 1 with the control group, significant differences (p 0.05), however, 8 out of 837 radiomic parameters differed significantly. Multiple regression analysis has shown that the Size Zone Nonuniformity gray level zone matrix (SZN-GLSZM) and Gray Level Variance (GLCM) gray co-occurrence matrix are independent predictors of the development of ACS within 5 years. According to the results of the ROC analysis, the logistic model with the inclusion of radiomic data showed high sensitivity and specificity in predicting the development of ACS (cut-off point function show_eabstract() { $('#eabstract1').hide(); $('#eabstract2').show(); $('#eabstract_expand').hide(); }