NONLINEAR HEART RATE VARIABILITY FEATURES IN DEPICTING THE EFFICACY OF RUTIN UNDER STREPTOZOTOCIN-INDUCED DIABETES MODEL WITH SUPPORT VECTOR MACHINE

Abstract

Diabetes mellitus (DM) is a multifaceted disease that leads to higher cardiovascular events with neuronal damage, inflammation, and oxidative stress in subjects. It also causes an autonomic imbalance with the onset of the disease which disturbs the cardiac dynamics. This work demonstrates the rutin in treating the inflammation caused by hyperglycemia through nonlinear heart rate variability features in predicting diabetes using a support vector machine (SVM). The lead-I electrocardiogram was acquired from the control, experimental, and treated group of the male Wister rats ([Formula: see text] gm and age 10–12 weeks). A dataset of 669 samples was obtained from the recorded ECG signal and taken as input vectors to the SVM. The observed results presented an accuracy of 92.9% in classifying the control and experimental group. Further, the same model with the treated group dataset showed an accuracy of 7.7% (samples nearer to the experimental group) while 92.3% of samples were close to the control group. The findings suggested the efficacy of rutin drugs in restoring the blood sugar level and the sympathovagal balance. The usefulness of the non-invasive technique in the prognosis of the disease gives direction in the design and development of the computer-aided cost-effective wearable system. However, the need for expert clinicians cannot be ignored.