Dynamic frequency spectrum analysis: application in organized and disorganized electrocardiographic signals

A digital spectrum-analysis technique is employed to compute the power of a specific frequency band (FB) of interest in continuous data segments. Only integrated power values of each frequency band are taken into account. To evaluate the electrophysiological characteristics of disorganized ECG signals, the power of each FB is plotted against elapsed time. Sustained ventricular fibrillation (VF) in 15 anesthesized dogs was characterized as follows: (1) The major power was contributed by the FB from 1 to 25 Hz; (2) the power of the low FB (1-25 Hz) and the high FB (>151 Hz) remained relatively constant during sustained VF; (3) the power between 51 and 125 Hz showed a decrease in the first 134±25 s (p<0.0001) and a plateau thereafter. The method eliminates the need for QRS-onset detection in frequency analysis of ECG signals. Although the method also detects high-frequency changes previously described using signal-averaging techniques, it is particularly useful in disorganized ECG signals and is particularly applicable to the analysis of ECG during VF