Calculation of Heartbeat Rate and SpO2 Parameters Using a Smartphone Camera: Analysis and Testing

Figures

• 

  Figure 1. Flowchart of the methodology used to obtain pulse rate and ratio-of-ratios.

• 

  Figure 2. (Left): Mean intensity followed by a moving-average filter with several sliding window lengths vs. frame number. (Right): shows, in magnification, the region around the peak at frame 880.

• 

  Figure 3. (Left): signal intensity obtained with the moving-average filter applied with and without the Savitzky–Golay filter (with several sliding window lengths). The red circled peak at frame 915 is magnified on the right. The red circled peak at frame 915 shows that a sliding window of length 20 across neighboring data elements gives smoother results.

• 

  Figure 4. Typical FFT spectrum from a smoothed mean intensity signal (red channel). The estimated pulse rate is 67 bpm (1.1195 Hz).

• 

  Figure 5. Schematic diagram representing a random frame window used to calculate the AC and DC components of the signal.

• 

  Figure 6. Normalized pulse rate (PR) of all participating volunteers; the video recording time in these graphs corresponds to 45 s.

• 

  Figure 7. Normalized pulse rate (PR) of all participating volunteers; the video recording time in these graphs corresponds to 30 s.

• 

  Figure 8. Normalized pulse rate (PR) of all participating volunteers; the video recording time in these graphs corresponds to 15 s.

• 

  Figure 9. Ratio-of-ratios for a group of 19 healthy individuals calculated using the FFT method and the local maximum method.