Several methods are proposed in the literature to detect ballistocardiogram (BCG) peaks. There is a need to narrow these methods down in terms of their performance under similar conditions. This study reports early results from a systematic performance evaluation. To date, we have replicated three methods from the literature and compared their performance using data from five volunteers. A basic cross-correlation approach was also included as a baseline level of performance. The best-performing method had an average peak-detection success rate of 95.0{%, associated with 0.1090 average false alarms per second and 0.0078 s mean standard deviation between real and detected peaks.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1109/EMBC.2018.8512317 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!