Publications

Abstract : This paper investigates the feasibility of using compressed electrocardiogram (ECG) sensing for R-peak detection in conjunction with entropy-based non-linear heart rate variability (NLHRV) analysis to assess the complexity and irregularity of the heart rate variability (HRV) signal. Non-linear measures offer valuable insights beyond conventional time-domain and frequency-domain features, making them instrumental in understanding these intricate dynamics. In the context of resource-constrained wearable internet of things (IoT) devices, computing HRV parameters from high-resolution (HR) ECG signals presents computational and energy consumption challenges. To address this, compressed ECG sensing for R-peak detection emerges as a promising approach for expedited processing. This study focuses on the application of entropy-based NLHRV analysis to evaluate various clinical and functional conditions using ultra-short-term (1 minute) segments of ECG data. A comparative analysis is presented, employing both compressed ECG sensing for R-peak detection and conventional HR R-peak detection. Specifically, 48 ECG records from the MIT-BIH arrhythmia database (MIT-BIH-AD) and 18 ECG records from the MIT-BIH normal sinus rhythm database (MIT-BIH-NSRD) are used for the analysis. The 7 out of 10 entropy indices have presented more than 85% similar results. These findings signify the applicability of compressed ECG sensing-based entropy measures through NLHRV analysis for rapid health monitoring using ultra-short-term ECG signals.