Signal source impedance of implanted pacemaker electrodes estimated from the spectral ratio between loaded and unloaded electrograms in man

Abstract

The signal source impedance Z of implanted pacemaker electrodes plays an important role for adequate sensing of electrogram signals and has so far been little explored. A frequency-domain method for thein vivo calculation of Z is described. Electrogram signals picked up between two electrodes were recorded unloaded, and loaded with resistors and capacitors, then amplified, and stored on magnetic tape. After time sampling and a.d. conversion, the signals were read into a digital computer as discrete time vectors with 2^P samples. Each vector was subjected to a fast Fourier transform (f.f.t.), and the ratio H(jω_i) between the transformed loaded and unloaded vectors was calculated. For the electrode impedance Z a linear model was chosen that consisted of the tissue and electrode resistance R_T, in series with a parallel coupling between Faraday resistance R_F, and Helmholtz capacitance C_H. By nonlinear regression analysis the parameters (∧) in this model were estimated from the ratios H(jjω_i) obtained in patients with permanent pacemaker electrodes, and in patients with temporary multicore wire electrodes after open heart surgery. As an estimate for R_T was also used the voltage/current ratio R′_T, 90 μs into the stimulation pulse from an external variable parameter pulse generator. The 3-component impedance model described the essential features of the electrode impedance in the frequency range up to 100 Hz. It was found that Ĉ_H, decreased slightly with increasing frequency, and that Ř_T in general was larger than R′_T. A modified model with frequency dependent parameters was found to obtain a better approach to the experimental data.