Statistical analysis of heart rate and heart rate variability monitoring through the use of smart phone cameras

Video recordings of finger tips made using a smartphone camera contain a pulsatile component caused by the cardiac pulse equivalent to that present in a photoplethysmographic signal. By performing peak detection on the pulsatile signal it is possible to extract a continuous heart rate signal. We performed direct comparisons between 5-lead electrocardiogram based heart rate variability measurements and those obtained from an iPhone 4s and Motorola Droid derived pulsatile signal to determine the accuracy of heart rate variability measurements obtained from the smart phones. Monitoring was performed in the supine and tilt positions for independent iPhone 4s (2 min recordings, n=9) and Droid (5 min recordings, n=13) experiments, and the following heart rate and heart rate variability parameters were estimated: heart rate, low frequency power, high frequency power, ratio of low to high frequency power, standard deviation of the RR intervals, and root mean square of successive RR-differences. Results demonstrate that accurate heart rate variability parameters can be obtained from smart phone based measurements.