Description and validation of an ECG removal procedure for EMGdi power spectrum analysis

We describe a cross-correlation procedure for removing contaminating electrocardiogram (ECG) complexes from the diaphragmatic electromyogram (EMGdi). First, the operator selects ECG templates from the EMGdi signal during expiratory intervals. Second, these templates are used to locate ECG complexes occurring during inspiratory EMGdi activity. Third, at the point of maximum correlation between the template and these ECG complexes, the template is adjusted in size and offset to 'match' the ECG complex, and adjustments are determined by the linear regression coefficients. Finally, the modified template is subtracted from the EMGdi signal. To evaluate our method, we compared the power spectral density (PSD) obtained from processing EMGdi signals by our method with those obtained from the EMGdi signal in which ECG complexes had been removed by gating. Our results indicate that PSD obtained by these two different methods shows no statistically significant differences with respect to the following features: centroid frequency, median frequency, total power, standard deviation, skewness, and kurtosis.