Acquisition and Processing of ECG and PPG Signals Using Face-Worn Sensors for Extracting the Cardio-Respiratory Parameters

This work aims to evaluate the feasibility of electrocardiographic (ECG) and photoplethysmographic (PPG) signal detection from the face area, ensuring good signal quality to derive the biophysical parameters (HR, SpO
2
, HRV, breathing rate, etc.). The derived hardware and firmware solutions will be integrated into a sensorized sleep mask. The experimental tests demonstrated that the optimal zones for acquiring the ECG and PPG were the mastoid and cheek areas, respectively, as the mastoid region contains highly conductive muscular tissue, whereas the cheek is significantly perfused with blood from the numerous vessels in this area. The test results indicated that processed signals’ features could be correctly visually discerned, enabling extracting the parameters; in particular, the mastoid ECG achieved an SNR of 15.65 dB. In addition, inertial signals were detected and processed to compute Euler's angles (i.e., roll, pitch, and yaw), used to recognize different head movement patterns. Lastly, the developed firmware solutions are not computationally demanding as they occupy just 107.06, 93.76, and 329.39 kB for ECG, PPG, and inertial signals processing, respectively, ensuring applicability in resource-limited platforms.