Wearable technology provides fitness information and, in some cases, continual access to tracking and monitoring of cardiac health to physicians and patients. However, wearable methods suffer from motion artifacts. To address this issue, we have developed a multi-modal system using the combination of (1) PPG with three wavelengths of light (green, red, and IR at 527, 660, and 880 nm), (2) bioimpedance, and (3) single-sided ECG to measure heart rate (HR) on the upper arm. This study investigated measuring HR under conditions of sedentary motion and micro-motion (e.g., typing). We compared the system with signals acquired from commercial wrist and chest devices, verifying the accuracy of the HR measurement for each anatomical location. This multi-modal approach investigated the wavelength of light chosen to provide the most accurate HR measurement and assessed cross-correlation to minimize motion artifacts. Results indicated the green PPG and SS-ECG modalities have the lowest mean absolute error (1.0 and 2.0 bpm, respectively) relative to the chest device during typing conditions compared with the wrist device and other upper arm modalities.
|