Investigation on Complexity in Variability of ECG Time Segments for Diagnosis of Cardiac Autonomic Neuropathy

Abstract

Cardiac autonomic neuropathy (CAN) disrupts autonomic nerves controlling heart and blood vessels, impacting heart rate, blood pressure, and overall function. Linked to diabetes, CAN poses major health risks. The research investigates the self-similar nature of the ECG waveform, a focal point in numerous studies. Given the susceptibility of biomedical signals to recording noise, the study initially employs basic filters for artifact removal. Extracting ECG features such as QRS complex, P, and T wave detection follows artifact removal. Specific time segments (PR, QT, RR, and ST) are chosen due to their physiological relevance in CAN representation, with their time intervals analyzed to explore CAN significance. newlineA novel fractal-based approach is adopted to examine self-similarity changes in the specific time segments of ECG. Fractal analysis yields mean and standard deviation (SD) values for different time segments. While mean value variation among groups in investigated segments is comparable, early diagnosis demands clear differentiation between normal and early groups. However, mean and SD values extracted from segment data fail to exhibit significant intergroup differences, substantiated by statistical segment interval analysis newline