Analysis of Wavelet Transform Application for Filtering Real ECG Signals from High-Frequency Noise

The application of wavelet transformation for denoising real ECG signals from high-frequency noise is considered. The study is conducted for two types of signals: signals without preprocessing and signals processed by a low-pass Fourier filter. The most common wavelet filters with different decomposition levels are applied: Haar, Daubechies, Symlet, and Coiflet wavelets. Metrics such as MSE, SNR, correlation coefficient, and PRD are used to evaluate the effectiveness of signal filtration. It is concluded that the best high-frequency filter for ECG signals is the second-level Symlet filter.