S Blok1, M A Piek2, I I Tulevski2, G A Somsen2, M M Winter3. 1. Department of Cardiology, Cardiology Centers of the Netherlands, Karel du Jardinstraat 61, Amsterdam, the Netherlands; Department of Vascular Medicine, Internal Medicine, Amsterdam UMC, University of Amsterdam, Amsterdam Cardiovascular Sciences, Meibergdreef 9, Amsterdam, the Netherlands. Electronic address: s.blok@cardiologiecentra.nl. 2. Department of Cardiology, Cardiology Centers of the Netherlands, Karel du Jardinstraat 61, Amsterdam, the Netherlands. 3. Department of Cardiology, Cardiology Centers of the Netherlands, Karel du Jardinstraat 61, Amsterdam, the Netherlands; Department of Cardiology, Amsterdam UMC, University of Amsterdam, Amsterdam Cardiovascular Sciences, Meibergdreef 9, Amsterdam, the Netherlands.
Abstract
INTRODUCTION: Photoplethysmography (PPG) in wearable sensors potentially plays an important role in accessible heart rhythm monitoring. We investigated the accuracy of a state-of-the-art bracelet (Corsano 287) for heartbeat detection in cardiac patients and evaluated the efficacy of a signal qualifier in identifying medically useful signals. METHODS: Patients from an outpatient cardiology clinic underwent a simultaneous resting ECG and PPG recording, which we compared to determine accuracy of the PPG sensor for detecting heartbeats within 100 and 50 ms of the ECG-detected heart beats and correlation and Limits of Agreement for heartrate (HR) and RR-intervals. We defined subgroups for skin type, hair density, age, BMI and gender and applied a previously described signal qualifier. RESULTS: In 180 patients 7914 ECG-, and 7880 (99%) PPG-heartbeats were recorded. The PPG-accuracy within 100 ms was 94.6% (95% CI 94.1-95.1) and 89.2% (95% CI 88.5-89.9) within 50 ms. Correlation was high for HR (R = 0.991 (95% CI 0.988-0.993), n = 180) and RR-intervals (R = 0.891 (95% CI 0.886-0.895), n = 7880). The 95% Limits of Agreement (LoA) were -3.89 to 3.77 (mean bias 0.06) beats per minute for HR and -173 to 171 (mean bias -1) for RR-intervals. Results were comparable across all subgroups. The signal qualifier led to a higher accuracy in a 100 ms range (98.2% (95% CI 97.9-98.5)) (n = 143). CONCLUSION: We showed that the Corsano 287 Bracelet with PPG-technology can determine HR and RR-intervals with high accuracy in cardiovascular at-risk patient population among different subgroups, especially with a signal quality indicator.
INTRODUCTION: Photoplethysmography (PPG) in wearable sensors potentially plays an important role in accessible heart rhythm monitoring. We investigated the accuracy of a state-of-the-art bracelet (Corsano 287) for heartbeat detection in cardiac patients and evaluated the efficacy of a signal qualifier in identifying medically useful signals. METHODS:Patients from an outpatient cardiology clinic underwent a simultaneous resting ECG and PPG recording, which we compared to determine accuracy of the PPG sensor for detecting heartbeats within 100 and 50 ms of the ECG-detected heart beats and correlation and Limits of Agreement for heartrate (HR) and RR-intervals. We defined subgroups for skin type, hair density, age, BMI and gender and applied a previously described signal qualifier. RESULTS: In 180 patients 7914 ECG-, and 7880 (99%) PPG-heartbeats were recorded. The PPG-accuracy within 100 ms was 94.6% (95% CI 94.1-95.1) and 89.2% (95% CI 88.5-89.9) within 50 ms. Correlation was high for HR (R = 0.991 (95% CI 0.988-0.993), n = 180) and RR-intervals (R = 0.891 (95% CI 0.886-0.895), n = 7880). The 95% Limits of Agreement (LoA) were -3.89 to 3.77 (mean bias 0.06) beats per minute for HR and -173 to 171 (mean bias -1) for RR-intervals. Results were comparable across all subgroups. The signal qualifier led to a higher accuracy in a 100 ms range (98.2% (95% CI 97.9-98.5)) (n = 143). CONCLUSION: We showed that the Corsano 287 Bracelet with PPG-technology can determine HR and RR-intervals with high accuracy in cardiovascular at-risk patient population among different subgroups, especially with a signal quality indicator.
Authors: Sanchit Kumar; Angela M Victoria-Castro; Hannah Melchinger; Kyle D O'Connor; Mitchell Psotka; Nihar R Desai; Tariq Ahmad; F Perry Wilson Journal: J Cardiovasc Transl Res Date: 2022-09-09 Impact factor: 3.216