用小波变换和神经网络检测心室晚电位

Detection of Ventricular Late Potentials Using Wavelet Transform and Neural Networks

提出了一种基于小波变换和神经网络的晚电位时域检测改进方法 .对矢量幅值波进行连续小波变换 ,然后从矢量幅值波的时频能量分布中提取心室晚电位的 8个特征值 ,送入人工神经网络进行晚电位的自动检测 .实验结果表明 :这种检测方法可有效降低噪声影响 ,提高心室晚电位的检测准确率 .

Based on wavelet transform and neural networks , an improved method is proposed for detecting ventricular late potentials in time domain. Continuous wavelet transform is used to vector magnitude waves. Eight features are extracted from the time frequency representation of the vector magnitude waves, and are inputted into the neural networks to detect ventricular late potentials. The results show that the proposed method may effectively reduce the noise affection and improve the detection accuracy of ventricular late potentials.