Automation an Algebraic Algorithm for Solving the Inverse Problem of Electrocardiography

Cardiovascular diseases (CVD) are one of the most common causes of death in Russia. Pathological changes of the myocardium in the early stages are practically no noticeable on the electrocardiogram. The development and implementation of algorithms for solving the inverse problem of electrocardiography (ECG IP) are designed to expand the capabilities of early diagnosis of heart diseases. The paper considers the issues of automatic selection of the regularization coefficient of the algebraic algorithm for solving the ECG IP using multielectrode electrocardiographic signals (ECGS) and their selected high-frequency low-amplitude (HF LA) components. Dependencies have been established by which the regularization coefficient can be calculated, knowing the noise level in the ECG signals and the current cardiocycle interval. This made it possible to automate the algebraic algorithm for solving ECG IP without a significant loss of accuracy and, thereby, reducing the time for reconstruction by eliminating the stage of selecting the regularization coefficient.