Session S72.6

Heart Rate Detection in Highly Noisy Handgrip Electrocardiograms

CC Lin, WC Hu*, CM Chen, CH Weng

National Chin-Yi University of Technology
Taichung, Taiwan

The electrocardiogram recorded by handgrip of holding on to a pair of metal electrodes has the advantage of convenient measurement of heart beat at low cost for sport utility. The heart rate detection using the handgrip electrocardiogram has been applied in treadmills to monitor the instantaneous heart rate that adjusting the volume of exercise automatically. However, the major limitation of the handgrip ECG monitor was the highly noisy. The noise level that induced by the handgrip measurement during the exercise has degraded the waveform of QRS complex. It is caused by the poor and unstable contact of electrodes from the motion during exercise. This limited the sensitive of R-peak detection and increased the difficulty of heart rate detection.
This study is to develop the handgrip heart rate detection methodology based on template matching algorithm. A sixth-order bandpass Butterworth filter with a passband of 1 to 30 Hz was used to filter out baseline shift and noise above 30 Hz including power line noise. The resulting QRS waveform will be processed by the matching filter and ready for heart beat detection. Finally, the result of detected heart beat will be processed by a weighting function to further minimize the false heartbeat detection. The performance of heart rate detection was tested under resting and exercising. Five Taiwanese male subjects averaged age of 25 were tested. The experimental results showed that the error rates were 0.78% (total of 766 heart beats) at resting condition and 3.83% (total of 3077 heart beats) during exercising. The application of this result from the handgrip heart rate detection can be applied in the heart rate variability analysis.

(Abstract Control Number: 195)