| Summary: | This research work explores the feasibility of using frequency domain analysis in the study of arrhythmias. The research involves the application of spectrum analysis to obtain the dominant frequency (DF) of atrial electrograms (AE) at different sites in the atria. It is an alternative way of interpreting the chaotic electrical activity seen during AF and reveals critical sites to guide ablation. As longer ablation procedure time implies higher risk to the patient, DF estimation needs to be obtained as quickly as possible. Four techniques (FFT, Blackman-Tukey, Autoregressive and Multiple Signal Classification) were used to compare the computation times taken for spectrum estimation analysis. The FFT technique produces an accurate DF result with the shortest time. DF analysis was first used for ventricular fibrillation with data from the surface of the left ventricle (in animal studies). It was found that spectrograms show the DF drifting along time and with significant changes in power. This approach was then applied for bipolar AF signals (in human studies). The changes of the frequency along time were observed when the stimulation was given, either using high frequency stimulation or drug infusion. We have developed a novel technique for the removal of ventricular signals from virtual AE. The surface ECG is used to identify ventricular activity. A band pass filter (8 Hz to 20 Hz) followed by rectification and then a low pass filter (6 Hz) are used for QRS detection. QRST subtraction was performed using three different approaches: flat, linear and spline interpolation. QRST subtraction affects the power of the signals but not the DF. We also developed an adaptive power threshold tool to observe the distribution of the DFs with an adjustable power threshold setting. Using this tool the 3D maps can display the evolution of the DFs within a chosen threshold power bracket.
|
|---|
