Characterization and validation of forced input method for respiratory impedance measurement.

We have studied the effect of spontaneous breathing on the measurement of total respiratory impedance by the oscillation method, that is using forced random pressure input. Free breathing of the subject induces parasite signals on the pressure and flow measured at the mouth. These induced random signals are correlated through a loudspeaker box with side tubing and flowmeter impedance. The respiratory impedance of the subject, calculated by the cross-power spectrum method, is shown to be systematically biased. This bias is obvious in a calculation model, experimentally verified by an analogue electrical experiment. The validity of the method depends on the respiratory noise level which affects the correlation between the forced pressure input and related flow output. The impedance of the apparatus should thus be minimized in order to decrease the respiratory noise. In this manner, the bias would be shifted to lower frequencies and the accuracy of the measurement near the resonant frequency of the respiratory system would be improved.