RESUMEN
Based on the fact that the respiratory component modulates the cardiac cycle component in the ballistocardiogram (BCG) signal, we propose a method that detects respiratory with time-frequency analysis for the sitting ballistocardiography system. Firstly, we demodulated the BCG signal by using the variable frequency complex demodulation (VFCDM) to obtain the output for different center frequency of interest. Then we calculated the instantaneous frequencies and the instantaneous amplitudes by the time-frequency representation. We reconstructed the time-domain waveform of respiratory at last. In order to verify the feasibility and accuracy of this method, we applied wavelet transform and nasal thermistor signal to compare qualitatively and quantitatively. The simulation results showed that the proposed method could detect the respiratory rate from BCG signal more accurately, which provided meaningful attempt for monitoring the multiple physiological parameters synchronously and unconsciously.