An Efficient Extraction Approach of Heart Rate Variability from Real-time Smartphone Videos

摘要

Heart rate variability(HRV) is widely used to autonomic nerves system assessment and cardiac disease diagnosis in clinic. The traditional HRV analysis is based on Electrocardiogram (ECG). However, it is more convenient to measure HRV by Photoplethysmography (PPG) in most wearable applications, such as watches. This paper proposed a novel method called Temporal Difference Interval Pixels(TDIP) method to obtain PPG signal through smartphone camera. In the experiment, we compared the cost time of obtaining PPG between traditional Temporal Difference(TD) method and TDIP method. We used these two methods to obtain PPG from 5 minutes' smartphone videos, which was acquired from 10 subjects. The average cost time of these two methods are 152.72 seconds and 82.38 seconds respectively. Then, we compared the time domain parameters the standard deviation of NN intervals (SDNN)and the Root mean square of successive differences between NN intervals (RMSSD) obtained by the proposed method with that obtained by ECG. The Pearson coefficient is 0.91 for SDNN and 0.95 for RMSSD respectively. The result can meet the HRV accuracy requirement, and it also significantly reduces the amount of calculation to improve the real-time performance.