Volatilome and Single-Lead Electrocardiogram Optimize Ischemic Heart Disease Diagnosis Using Machine Learning Models

This is a prospective, case-control, single-center, observational, non-randomized study. It is designed to evaluate the diagnostic accuracy of functional tests involving physical exertion monitored via a 12-lead ECG, combined with analysis of exhaled breath volatile organic compounds (VOCs) and single-lead ECG parameters.

The planned number of participants to include in the study is 80, admitted to the University Clinical Hospitals No. 1, at the I.M. Sechenov First Moscow State Medical University of the Ministry of Health of the Russian Federation (Sechenov University). The study includes the following stages: 1. Participants will be selected according to inclusion and exclusion criteria; 2. Work with medical documentation; 3. Instrumental and laboratory examinations of the participants: 3.1. Analysis of exhaled air will be carried out with the Compact PTR-MS instrument manufactured by Ionicon (Austria) (analytical device), registration certificate No. (C16)07/C05. 3.2. All the participants will undergo a single blood sampling during the day of performing the study, a blood test, 10 ml from a peripheral vein to determine the level of total cholesterol, low-density lipoprotein (LDL), very low-density lipoprotein (VLDL), high-density lipoprotein (HDL), triglycerides, C-reactive protein (CRP), lipoprotein a, apolipoprotein B, and interleukin-6 (IL-6). 3.3. Both groups will perform a bicycle ergometry test (on a SCHILLER c200 device) to evaluate the response to physical activity. 3.4. Before and immediately after the exercise test, all patients are scheduled to record a single-lead ECG and pulse wave, using a portable single-lead recorder (Cardio-Qvark) (Russia, Moscow). 4.5. Stress computed tomography myocardial perfusion imaging (CTP) with a vasodilation test using adenosine triphosphate on a CT device with 640 slices (Canon; Aquilion One Genesis) will be performed. After completion of the instrumental and laboratory analysis, a statistical analysis will be conducted using classical statistics and machine learning methods, including gradient boosting.