关键词: 心电信号, 降噪, 小波变换, 阈值选取, 信噪比

Abstract: Multiresolution has strong dependence on the choice of wavelet threshold and wavelet function design, and the difference in the effect of noise reduction for different individuals’ ECG signals is much obviously. In this work, an adaptive wavelet threshold selection method based on heuristic threshold optimization is proposed, in which the wavelet decomposition layer number and level influence factor are incorporated. More reasonable signal decomposition and noise reduction are realized by dynamically adjusting the threshold calculation function of each layer for wavelet coefficients. The experimental results show that the proposed algorithm achieves better performance for reducing the noise of ECG to meet the needs of clinical application.

Key words: electrocardiogram, denoising, wavelet transform, threshold selection, signal-to-noise ratio