Real Time Analysis of ECG Signal for Baseline Wander Removal and R Peak Detection Using Discrete Wavelet Transform

Abstract

Inspiration of this project is from the need to find an efficient method for analysis of ECG signal which needs to be simple, have good accuracy and have less computational time. As the main cause of death around the globe is being led by heart related diseases, hence a recent study shows that in the working age group of around 24-65 years, the death percentage of 25 is only because of the various heart related diseases. Hence ECG is the most important signal for consideration and observation to prevent these issues. ECG is a bio medical signal which can be detected by using electrodes placed in the particular locations of human body. In my research work there are two stages for the efficient analysis of ECG signal. Primary stage is the enhancement of ECG signal, that is removal of noise. It is done by extracting the required cardiac components by rejecting the background noises mainly Baseline Wander noise which is a low frequency noise and generally lies below 0.5 Hz. This is done by using a traditional MA filter method and by using wavelet theory. This thesis work focuses on the algorithms proposed for enhancing for both methods and finding out the correlation for the enhancement. Secondary stage is to determine the R peak using DWT and perform the signal reconstruction. This work also follows a proposed algorithm along with the comparison for getting a better detection which can be used for the required operations in the typical ECG signal. The simulation is done in the MATLAB environment. The results are shown in the view to compare the different methods to denoise signal with baseline wander and also the comparison of different threshold methods for appropriate detection of R peaks for fast and accurate computation.