Derivation of respiration from ECG measured without subject´s awareness using wavelet transform

We developed a system to measure ECG without subject´s awareness and derived respiration from the ECG using the wavelet transform. The bed sheet electrodes consisted of three pieces of conductive textiles. Respiratory signals (EDR) were derived by reconstructing the detail signal of the 9th decomposition from the wavelet transform. The respiration periods were calculated by detecting the zero-crossing of the respiratory signals. Correlations between series of respiration periods extracted from EDR and airflow were high for all subjects.