Remote photoplethysmography imaging (rPPGc) is a new method measuring essential parameters, such as heart rate (HR), which uses a video camera during teleconsultation. Our work aimed to evaluate the accuracy of such remote measurements compared with existing contact point measurement methods in real-life clinical settings. The prospective hospital-based study recruited 1045 patients who required a pulmonary function test. For each patient, measurements of HR using a standard electrocardiogram acquisition system (gold standard) were carried out concomitantly with the measurements made by the rPPGc system (Caducy v1.0.0) taken within a 60 s timeframe. Age, gender, and skin phototype were collected. We performed an intraclass coefficient correlation (ICC) and Bland-Altman plotting to determine the accuracy and precision of the rPPGc algorithm readings. We achieved measurement of HR using the two methods in 963 patients. The ICC measured at a 60 s timeframe, and when we compared the rPPGc with the gold standard, it had a 95% confidence interval (CI95) value of 0.886 [0.871:0.899]. In all, 94.6% (n = 911) patients showed promising results with a CI95 in Bland-Altman plotting. Fifty-two measurements were discordant, and further analysis established the method's accuracy at 96.2%. Our results described a good accuracy and correlation between the rPPGc system and the gold standard, thus paving the way for more precise care via telemedicine.
Keywords: heart rate; remote photoplethysmography; vital signs.