Higher order spectral analysis methods, such as Bispectrum and Wigner-Ville Distribution, were applied to ECG recordings to determine signal abnormalities in ventricular arrhythmias.
Does high-order spectral analysis using the MATLAB HOSA toolbox identify signal abnormalities in ventricular arrhythmic ECG signals?
Higher-order spectral analysis methods can be applied to ECG signals to process phase information and identify ventricular arrhythmias.
In signal processing methods using second order statistics and / or power spectrum, phase relationships between frequency components are not considered. For this reason these methods are blind to the phase. Second order statistics and power spectrum are also not sufficient to statistically define the non-gaussian processes. In recent years, random processes have been defined as more statistically more sensitive, and high-order statistics (greater than two) and spectrum studies have been carried out to process phase information. In this study, high-order spectral analyzes were performed on ECG recordings in MIT-BIH Malignant Ventricular Arrhythmia and MIT-BIH Normal Sinus Rhythm databases. The obtained results were compared and tried to determine signal abnormalities with the help of higher order spectral analysis methods (Bispectrum, Wigner-Ville Distribution etc.). Analyzes were performed using the MATLAB HOSA toolbox.
Sayın et al. (Mon,) conducted a other in Ventricular Arrhythmia. Higher order spectral analysis (Bispectrum, Wigner-Ville Distribution) vs. Normal Sinus Rhythm was evaluated on Determination of signal abnormalities. Higher order spectral analysis methods, such as Bispectrum and Wigner-Ville Distribution, were applied to ECG recordings to determine signal abnormalities in ventricular arrhythmias.