Arrhythmia ECG signal De-Noising by using Discrete Wavelet Transform based on Genetic Algorithm
ECG is a series of waves and deflection recording the hearts electrical activity from different view, these waves can be used to diagnosis different arrhythmia cardiac disease .ECG is characterized by QRS complex, which provide the fundamentals of the heart electrocardiogram signal so the enhancement of the signal and detection of QRS is necessary for automatic determine the heart conditions, Different research and studies were proposed for this purpose. In this paper , we have proposed an algorithm for ECG signal enhancement ,de-noising and detection of QRS complex based on discrete wavelet transform with different filters, that can be used to convert input signal into different sub-band according to the frequency, each sub-band have coefficient with different significant, which is enhanced with different threshold methods and thresholds. We have proposed anew threshold method based on wavelet level, the value of threshold is change according to the wavelet level. Also genetic algorithm was used to determine the best parameters that are used in the de-noising methods, in which, the binary encoding was used by represent the chromosome with ten bits, two bits to select threshold method, one bit is used to select threshold function, four bit to chose wavelet filter type and three bits to select the best wavelet level . The result are compared with other works, and the comparison show that the proposed work is better.