NONLINEAR PULSE RATE VARIABILITY ANALYSIS FOR DIABETES IN ASSOCIATION WITH CORONARY ARTERY DISEASE PATIENTS

This study investigates the effect of coronary artery disease (CAD) and Type2 Diabetes Mellitus (T2DM) on the measurement of nonlinear Pulse Rate Variability (PRV) parameters with respect to the control group. The effect of coronary artery disease with and without DM on nonlinear PRV parameters is aimed to compare with the control group including evaluation of any differences in PRV features between CAD groups with and without diabetes. Patients clinically diagnosed with CAD were recruited for this study. These were sub-grouped as (i) CAD with DM ([Formula: see text] = 56); (ii) CAD without DM ([Formula: see text] = 9); (iii) DM ([Formula: see text] = 109); (iv) Control group ([Formula: see text] = 115). The study was done by recording and analyzing 5 min of Pulse-Plethysmogram (PPG) recordings from left-hand index finger in sitting position. Nonlinear PRV analysis was performed using Poincaré plot analysis, Approximate Entropy (ApEn), Sample Entropy (SampEn), and Detrended Fluctuation Analysis (DFA) along with statistical analysis using student’s [Formula: see text] -test at [Formula: see text] -value < 0.05. Major PRV features measuring heart rate complexity showed significantly lower values in diseased conditions (DM, CAD with and without DM) compared to the control group. No significant difference in any of the nonlinear parameters was found between CAD with and without DM groups. The result suggests that nonlinear PRV analysis using PPG signal recording could be very useful in the classification, measurement and prognosis of diseases due to cardiac-autonomic nervous system abnormalities like CAD and diabetes accurately using robust and reliable machine learning algorithms.