A new parametric frequency-wavenumber spectrum estimation algorithm and its application to the analysis of 3-dimensional epicardial ECG signals.

Abstract This paper presents a 3-dimensional (3-D) frequency-wavenumber spectrum estimation (FWSE) approach to the analysis of ECG signals. This approach treats the data as ‘wavefronts plus noise’ and provides a means of estimating key parameters associated with propagating wavefronts. A high resolution technique based on minimum variance representations of 3-D data fields (3-D CLS technique) is employed to obtain the FWSE. Computer simulation results that demonstrate the high resolution property of the technique when compared with the maximum-likelihood method of Capon are presented. Results of application of the technique to epicardial ECG data collected from a sensor array are also presented and discussed.