The paper presents a procedure for correction of the error of an ECG signal, introduced by the skin-electrode interface. This procedure involves three main measuring-calculating stages: parametrical identification of the mathematical model of the interface, realized directly before the diagnostic measurements, registration of the signal at the output of electrodes as well as reconstruction of the input signal of the interface.

The first two stages are realized in the on-line mode, whereas the operation of signal reconstruction presents a numerical task of digital signal processing and is realized in the off-line mode through deconvolution of the registered signal with the transfer function of the skin-electrode interface.

The aim of the paper is to discuss in detail the procedure of parametric identification of the skin-electrode interface with the use of a computer system equipped with a DAQ card and LabVIEW software. The algorithm for error correction introduced by this interface is also presented.

The paper presents major issues associated with the problem of modelling change output accelerometers. The presented solutions are based on the weighted least squares (WLS) method using transformation of the complex frequency response of the sensors. The main assumptions of the WLS method and a mathematical model of charge output accelerometers are presented in first two sections of this paper. In the next sections applying the WLS method to estimation of the accelerometer model parameters is discussed and the associated uncertainties are determined. Finally, the results of modelling a PCB357B73 charge output accelerometer are analysed in the last section of this paper. All calculations were executed using the MathCad software program. The main stages of these calculations are presented in Appendices A−E.