Denoising and features extraction of ECG Signals using Unbiased FIR estimation techniques

Abstract

The electrocardiogram (ECG) signals bear fundamental information for deciding about heart diseases. So the scientific community has been performing many efforts during decades to extract features of heartbeats via ECG records with high accuracy and efficiency using different strategies and methods. However, the noise and artifacts provided by external factors avoid significant patterns associated with the ECG signals. These patterns play an important role to find specific abnormalities in ECG signals. Hence, techniques based on unbiased FIR (UFIR) filtering promises better results. In this dissertation, we have applied a model based on UFIR to ECG signals. Hence, we compare the proposed technique with traditional method such as predictors, standard filters (e.g. low-pass filter) wavelet filters, Savitsky-Golay filter. The UFIR method outperforms other studied techniques for ECG signals.