Multifactor Regression Analysis for Cardiac Diagnosis Prediction Based on Amplitude Variability Function

Multifactor regression model for automatic cardiac diagnosis prediction based on statistical characteristics of the amplitude variability function of electrocardiographic signals is proposed. The model employs seven statistically significant predictors (mean, median, mode, standard deviation, sample variance, kurtosis, and skewness) selected by means of stepwise regression with forward variable selection from the initial set of 13 parameters. The predictor inclusion criterion established a significance level of p < 0.05. The model for prediction of cardiac diagnosis, based on the function of the amplitude variability takes into consideration weight coefficients for each predictor, determined by the least squares method. Central tendency measurements make the greatest contribution into diagnostic outcome formation (β = 201.78 for mean and β = 68.69 for median). The model provides effective differentiation among three clinical states: conditional norm, cardiac rhythm disorders in the form of extrasystole, and morphological conduction abnormalities represented by incomplete left bundle branch block, achieving a Nagelkerke coefficient of determination R² = 0.991. Normal conditions are characterized by minimal amplitude variability values (mathematical expectation 0.00003…0.00064 mV), while extrasystole demonstrates an increase of 3—4 orders of magnitude. Model validation on a representative sample of 204 electrocardiographic signals (102 normal, 51 extrasystole, 51 block) confirmed its high statistical significance (Fisher's F-criterion 953.93, p < 0.001) and compliance with fundamental regression analysis assumptions. Residual analysis demonstrated normal distribution and homoscedasticity, confirming model adequacy. The proposed approach combines the advantages of high interpretability of classical statistical methods with innovative application of the amplitude variability function for comprehensive analysis of morphological and rhythmic characteristics of cardiac signals. The practical significance lies in creating mathematical tools for automated cardiovascular disease diagnostic systems and clinical decision support systems.