Identification of transducer defect in respiratory impedance measurements by forced random noise. Correction of experimental data.

The measurement of respiratory impedance by means of the forced oscillations technique depends on the frequency response of various measurement devices: pressure transducers, flowmeter, filters, etc. In the range of frequencies where measurements were done (0--100 Hz) it is difficult to have the same response (amplitude and phase) for mouth pressure and flow transducers. It is possible to identify the relative frequency response of both devices. This can be done by comparison of measurements made on a well-defined mechanical system with a theoretical model representing this system. The method also avoids separate frequency calibrations of pressure and flow. Its validity depends only on the accuracy of the model of impedance of the standard system used.