Heart Rate Variability Analysis on Reference Heart Beats and Detected Heart Beats of Smartphone Seismocardiograms
Executive Summary
This study evaluates heart rate variability (HRV) indices derived from smartphone-based seismocardiograms (SCG) using a modified beat detection algorithm. The algorithm achieved high sensitivity (Se=0.957) and positive predictive value (PPV=0.904) overall, with perfect performance (Se=1.000, PPV=1.000) for one subject. HRV indices such as mean NN, RMSSD, and LF/HF showed minimal differences between reference and detected beats, demonstrating the feasibility of smartphone-based SCG for HRV analysis.
Answer Machine Insights
Q: What was the sensitivity and positive predictive value of the beat detection algorithm?
The algorithm achieved an overall sensitivity of 0.957 and a positive predictive value of 0.904.
Overall beat detector performance (for all beats Se=0.957, PPV=0.904) is promising.
Q: Which HRV indices showed minimal differences between reference and detected beats?
Mean NN, RMSSD, pNN50, LF/HF showed minimal differences.
HRV indices calculated for two heart beat detectors are similar, especially for mean inter-beat interval (mean NN), RMSSD, pNN50, LF/HF.
Key Results
Overall sensitivity (Se) of 0.957 and positive predictive value (PPV) of 0.904 for the beat detection algorithm.
Minimal differences in HRV indices such as mean NN, RMSSD, and LF/HF between reference and detected beats.
Clinical Snapshot
Evidence Rating
Relevance
high Priority