基于小波变换的全方向M型心动图运动曲线滤噪

摘要

A method based on wavelet for denoising for omnidirectional M-mode cardiogram movement curve, is proposed omni-direc-tional M-mode echocardiography is a ultrasound image sequences obtained by B-mode ultrasonography. Due to various disturbances and staircase effect caused by image digitization in the process of gathering, the Omni-directional M-mode Echocardiography movement curve exsits noises and quantization effect. In order to eliminate noise and quantization effects in the high-frequency components, and ensure that the signal component of medical value is not lost too, according to the traditional wavelet threshold de-nosing method, the author put forward a imroved wavelet threshold method. Expermental results show that it can erase the interfering noise and weaken the influence of quantization effect,and making the movement curve become smoother, using the same method to smooth the speed curve and acceleration curve, it can provide more accurate information for the doctor' s diagnosis.%提出一种基于小波滤噪的全方向M型心动图运动曲线滤噪方法,全方向M型心动图是一种由B超得到的超声序列图像,由于在采集的过程中受到各种干扰以及由于图像数字化造成的阶梯效应,使得从全方向M型心动图中提取出的运动曲线存在噪声和量化效应. 为了在消除处于高频分量的噪声和量化效应的同时,保证具有医学价值的信号分量不丢失,本文在传统小波阈值去噪的基础上,采用改进的小波阈值去噪方法来处理全方向M型心动图运动曲线,实验结果表明,该算法能够极大地消除心动图运动曲线中的噪声和量化效应,使得运动曲线较为平滑,采用同样的方法对速度曲线和加速度曲线进行平滑处理后,能够为医生的诊断提供更精确的信息.