Ballistocardiogram (BCG):Through the medical-grade piezoelectric thin-film sensor, the body's weak vibration information during the heart's mechanical movement is collected to obtain the BCG signal. Based on this information, the system can obtain the heart rate, body movement and other physiological parameters. Based on the multi-parameter grading model, the system can analyse a whole night's sleep for staging, fall asleep and wakefulness judgement.

Heart rate variability (HRV) analysis technique based on ballistocardiogram (BCG):There is a difference in the time interval between adjacent wave peaks, i.e., heart rate variability (HRV). The change of JJ interval in the BCG is consistent with that of RR interval in the electrocardiogram (EKG). The HRV is caused by the activities of sympathetic and parasympathetic nervous systems; based on the BCG signal, accurate J-wave peak points were obtained by filtering, fast FFT, and autocorrelation feature extraction. Thus, the JJ interval time series were obtained. Based on the continuous monitoring of the vital signs and the calculation of the time domain, frequency domain, and nonlinear features of the JJ interval time series, the HRV characteristic parameters of the whole night were obtained.

BCG-based sleep analysis technique:By constructing a hierarchical sleep staging model based on heart rate, body movement and other parameters. The system can stage the whole night's sleep, determine the time of asleep and wake up, obtain the R.E.M., light sleep, deep sleep duration, and time of wake up in the middle of sleep. Following this, the system will combine these data with personalised characteristic information, tracking the whole night's sleep and evaluating the quality of the entire night's sleep.

Contactless blood pressure measurement based on BCG:Based on the BCG signal, the S1 and S2 signal structure is obtained by filtering and calculating the higher-order differential signals and other pre-processing means, which can effectively obtain the relevant cardiac systole and diastole features. Based on these features, the blood pressure calculation can be performed, and the real-time beat-by-beat blood pressure is obtained by constructing a linear regression equation model.