Abstract:

Abstract An electrocardiogram (ECG) is a recording of the heart's electrical activity. Symptoms of cardiac arrhythmias (CA) include an irregular heart rate. Large noise signal components will affect the acquired raw ECG signal from the MIT-BIH arrhythmia database. An ECG wave undergoes a pre-processing technique to eliminate noise signals and baseline drift. The most prominent characteristics will provide accurate and helpful data on cardiac arrhythmias.  The Pan Tompkins technique is a real-time algorithm developed specifically for detecting QRS complexes in ECG signals. As distortion in the reconstructed signal increases, the PSNR metric's numerical value decreases and is used to rank signals from best to worst. Therefore, when PSNR increases, so does the quality of the compressed or reconstructed signal. The wavelet transform can be used for denoising and pre-processing nonstationary ECG signals. Filtering is a vital stage in the processing of ECG data because the current goal of the healthcare sectors is to preserve essential diagnostic information with little noise. This research demonstrates that the wavelet Thresholding approach has an effect on the quality of reconstructed electrocardiograms (ECGs). In this research, we use a discrete wavelet transform to analyse and de-noise an electrocardiogram (ECG) signal. The results of a comparative analysis were used to evaluate the efficacy of the third-level decomposition wavelet Denoising technique in reducing errors. According to the findings of the tests, this technique generates signals that are cleaner and smoother while keeping the essential elements.

Keywords: