Analysis of TeagerKaiser Energy Operator Pertinence for R Peak Detection in ECG Recordings on a Smartphone

Continuous health monitoring holds promise for early detection of health status impairment and plays significant role in forward-looking applications and services related to fitness, well-being, chronic diseases treatment and independent living for elderly.

A number of arrhythmia detection algorithms are being developed within a CardiaCare project that is aimed at continuous monitoring of heart function in real-time and analyzing electrocardiograms on a smartphone. Arrhythmia detection algorithms are heavily rely on features extracted from electrocardiogram recordings. In particular, robust detection of so-called R peak is important. There is a variety of approaches to extract the R peaks. Nevertheless, widespread algorithms often demand of heavy computations. If these algorithms were run in mobile devices, the battery would quickly drained. Since this situation is not acceptable in continuous health monitoring, appropriate R peak detector should be carefully selected.

In proposed approach to addressing the requirement of low power consumption, fast R peak detection algorithm based on Teager-Kaiser energy operator (TKEO) is utilized. The algorithm have been thoroughly verified using real electrocardiograms with abnormalities annotated by experts that are fetched from open databases. Analysis of TKEO-based R peak extraction algorithms showed high detection performance indicators (sensitivity, specificity and accuracy) comparable to widespread Pan-Tompkins algorithm.