Heart Side-Channel: Estimation of Cardiovascular Signal Waveforms Through Skin Vibration Sensing

In this letter, we investigated the estimation of cardiovascular signals, electrocardiogram (ECG) and photoplethysmogram (PPG) waveforms through skin vibrations as piezoelectric plethysmogram (PEPG) signals. ECG and PPG measurements are important for biomedical and security applications, and PEPG signals may be used to estimate ECG and PPG waveforms based on physiological activities in the cardiovascular system. We investigated the estimation procedure for ECG and PPG signal waveforms from PEPG signals using a frequency analysis. To evaluate estimation performance, we conducted an experiment using a dataset that included ECG, PPG, and PEPG signals. The results indicate that the ECG and PPG waveforms were successfully estimated from the PEPG signals, with errors of less than 15% and 10%, respectively.