Improved HRV Analysis in ECG Data: A Comparative Study Using MATLAB Code, Kubios, and gHRV

Abstract

Heart Rate Variability (HRV) analysis is a vital tool in assessing autonomic nervous system regulation and cardiovascular
health. This study explores improved HRV analysis techniques by employing MATLAB code and comparing its performance with
widely used software tools, Kubios and gHRV. Electrocardiogram (ECG) data from ten subjects under four distinct conditions -
baseline, rest, Stroop color task, and meditation - were collected and analyzed. The study focuses on developing and implementing
novel algorithms in MATLAB for HRV estimation, providing a comprehensive comparison against existing methods. The study
examines the accuracy and reliability of HRV analysis results obtained through MATLAB implementation in contrast to Kubios
and gHRV. The MATLAB code is optimized for enhanced computational speed and accuracy, allowing for real-time processing
of ECG data. The results indicate significant improvements in HRV analysis using the proposed MATLAB implementation. The
proposed MATLAB code and Kubios have similar accuracy for the High-Frequency feature, with 85% accuracy. gHRV, on the
other hand, has 100% accuracy for PNN50, indicating its high accuracy in matching reference data. The comparative analysis
demonstrates the diverse HRV metrics across different experimental conditions. Additionally, the results highlight the difference
in the study approach between Kubios and gHRV, showcasing its potential for widespread adoption in clinical and research
settings.
This study not only presents an advanced HRV analysis methodology but also provides valuable insights into the reliability of
existing software tools. The findings offer researchers and clinicians an informed choice when selecting HRV analysis tools for
their specific applications, ensuring accurate and efficient assessment of cardiovascular health and autonomic nervous system
function. Further investigations and validations are warranted to establish the robustness and generalizability of the proposed
methodology across diverse populations and experimental paradigms