Analysis of cardiac abnormalities using Hilbert-Huang transform

This paper investigates the feasibility of using Hilbert-Huang Transform (HHT) so as to enhance the diagnostic procedures in cardiac abnormalities. HHT which embraces empirical mode decomposition and Hilbert spectral analysis is a newly developed adaptive signal analysis method and has been extensively used in the analysis of non-stationary signals. Empirical Mode Decomposition (EMD) algorithm, the key feature of HHT, decomposes a signal into so-called Intrinsic Mode Functions (IMFs). Furthermore, the Hilbert spectral analysis of IMFs provide frequency information evolving with time and quantises the amount of variation due to oscillation at different time scales and time locations. Normal and abnormal ECG signals from MIT-BIH data base are used to verify the effectiveness of the proposed method and the results demonstrate the superiority of proposed technique compared to other wellknown techniques for the analysis of cardiac abnormalities.