基于变分模态分解的直肠压力信号预处理研究

Preprocessing of Rectal Pressure Signal Based on Variational Mode Decomposition

针对依托于人工肛门括约肌系统的直肠功能诊断模型中采集信号存在大量干扰噪声的问题,提出了一种基于变分模态分解(VMD)算法与小波加权平均阈值去噪结合的预处理方法。利用VMD算法对原始直肠压力信号进行分解,对各模态分量进行小波阈值去噪,提取出有用信号进行重构。通过仿真比较分析该方法与EMD、小波阈值去噪等方法的去噪效果。实验结果表明,该方法在不同噪声水平下均显著提高输出信号的信噪比,同时避免原始信号中有用信息的丢失,具有良好的去噪效果,对直肠功能诊断的准确性具有重要意义。

In view of the problems of large amount of interference noise in the collected signal in the rectal function diagnosis model based on the artificial anal sphincter system,a preprocessing method based on variational mode decomposition(VMD)and wavelet weighted average threshold denoising is proposed in this paper.The VMD is used to decompose the original rectal pressure signal,denoise the modal components with wavelet threshold,and extract useful signals for reconstruction.The denoising effect of this method is compared with that of EMD and wavelet threshold denoising.