Analysis of a Pulse Rate Variability Measurement Using a Smartphone Camera

Figures

• 

  Figure 1. Connection between HRV and PRV analysis. From the ECG signal, the NN intervals (time durations) are determined, with the corresponding timestamps. The timestamps are needed when spectral analysis is applied to the NN time serial. The data for PRV is similarly obtained from the PP durations between the consecutive maximum values in the PPG signal.

• 

  Figure 2. Results got from applying our peak detection method to an ECG signal.

• 

  Figure 3. Experimental arrangement. The subject is sitting in a resting position, the electrodes of the ECG device are connected to the limbs, and the smartphone is held in the subject's palm.

• 

  Figure 4. A screenshot of the PPG measurement application during a recording.

• 

  Figure 5. Illustration of the results of the synchronization process, with the ECG signal shown in blue and the PPG signal shown in red.

• 

  Figure 6. Delta RR duration series computed on ECG (blue) and PPG (red) signals.

• 

  Figure 7. Plots of PRV indices related to HRV indices (horizontal axis) with R2 and linear regression.

• 

  Figure 8. Bland-Altman plots for PRV and HRV indices with limits of agreement (blue dashed lines), bias (black lines), and acceptance limits (red dotted lines).

• 

  Figure 9. Plots of HRV indices calculated for ECG channel 1 and ECG channel 2 with R2 and linear regression.

• 

  Figure 10. Bland-Altman plots of HRV indices calculated for ECG channel 1 and ECG channel 2 with limits of agreement (blue dashed lines), bias (black lines), and acceptance limits (red dotted lines).

• 

  Figure 11. Plots of HRV indices calculated for ECG channel 1 and ECG channel 3 with R2 and linear regression.

• 

  Figure 12. Bland-Altman plots of HRV indices calculated for ECG channel 1 and ECG channel 3 with limits of agreement (blue dashed lines), bias (black lines), and acceptance limits (red dotted lines).