Wavelet based R-peak detection for heart rate variability studies

Abstract

Detection of QRS complex in electrocardiogram (ECG) signals is of immense importance in cardiac health prognosis. In this paper a new symmetric wavelet for detection of R-peak is presented, which has been designed based on spectral characteristics and morphology of QRS complex. The detection of R-peak was carried out using this designed wavelet, and with existing symmetric wavelets such as db3, db6, haar and bior2.2. The detection accuracy with this wavelet is 99.99%, which is higher than those with existing symmetric wavelets. The algorithm has been tested on standard databases such as Fantasia database of normal and healthy subjects, MIT/BIH (Massachusetts Institute of Technology/Beth Israel Hospital) arrhythmia database, and on self-recorded electrocardiograms of normal subjects and patients under diseased stress. The study of heart rate variability (HRV) through computation of RR-tachogram using the new wavelet has proved to be effective in reliably evaluating HRV parameters.

Keywords:

• QRS complex
• New wavelet
• ECG
• RR-tachogram
• Heart rate variability

Acknowledgements

The authors thank the Electrical Engineering Department and the administration of the Institute of Technology, Roorkee, India for their administrative assistance in providing the various facilities for this work. The authors greatly appreciate the clinical and administrative help regarding ethical clearance and subsequent data recording and interpretation of results rendered by the medical fraternity and associated staff of Cardiology Department, Himalayan Institute and Hospital Trust, Dehradun (Uttrakhand), India and the Chief Medical Officer, Dr M.R. Kural and associated staff at Indian Institute of Technology Roorkee Hospital. The technical inputs of Dr Dilbagh Singh, Department of Instrumentation & Control Engineering, Dr B.R. Ambedkar, National Institute of Technology, Jalandhar have been of immense help to the authors. Ramesh Kumar Sunkaria is also grateful to Prof. (Dr) Rakesh K. Sarin, Head, Department of Electronics & Communication Engineering and Prof. (Dr) Moinuddin, Director, Dr B.R. Ambedkar, National Institute of Technology, Jalandhar for providing him an opportunity to pursue his Ph.D. programme at the Indian Institute of Technology Roorkee, India.

Declaration of interest: The author declares no conflict of interest. The above work was carried out in biomedical engineering lab of the Indian Institute of Technology Roorkee, India. We have not received any funding for this work from any agency.