SIA OpenIR  > 机器人学研究室
基于小波变换的胃电信号处理方法研究
Alternative TitleEGG Signal Processing Method based on Wavelet Transform
丁伟
Department机器人学研究室
Thesis Advisor李洪谊
ClassificationR573
Keyword胃电 小波变换 数字信号处理 小波基函数优化 医疗摩胃机构
Call NumberR573/D58/2012
Pages103页
Degree Discipline模式识别与智能系统
Degree Name博士
2012-03-12
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract随着数字信号处理技术的飞速发展,小波变换的信号处理技术在数字图像处理、工业自动化信号处理、机器人技术以及生物信号处理等领域得到了广泛的应用。基于小波变换(Wavelet Transform,WT)的胃电(Electrogastrogram,EGG)信号采集系统是小波变换在数字信号处理领域中一个新的应用,其特点是利用了小波变换的去噪和分频的功能来解决胃电信号幅值小、频率低的问题,这种解决方案可以充分利用小波变换的特点予以实现。这类系统较之于传统的胃电系统的优点主要体现在两个方面:一方面提高了信噪比,降低了胃电信号的高频噪声,使更多有价值的胃电信息加以突出;另一方面实现了精确的快、慢波频段的滤波,使得胃电信号的快、慢波频段信号得以真实还原。胃电图的使用为胃病的科学诊断开辟了新的可靠的解决途径,但同时也带来了新的挑战,胃电数据的主要特征——幅值小、频率较低、信噪比低等——严重降低了胃电的实际使用价值,甚至导致胃电的不可靠及误诊。传统的胃电信号处理方法和滤波算法效果并不理想,而且滤波效果不够精确,存在大量噪声和干扰。另外,传统的滤波算法并没有考虑快、慢波分别滤波及叠加等问题。一个理想的胃电信号采集系统是对所有特征频段信号进行叠加或加权叠加以合理反映胃电的真实特征。基于以上原则,本文提出了基于小波变换的胃电信号处理方法,并将其应用于医疗摩胃机构控制系统的分析和设计中,针对以上方法及应用在以下几个方面进行了深入研究:1. 针对现有的胃电信号采集和信号处理系统存在的问题,提出了基于小波变换的胃电采集和滤波算法。该信号处理和滤波算法充分考虑了胃电的幅值特性和频率特性,实现了系统胃电快、慢波的滤波和叠加,为胃电系统奠定了基础。2. 基于小波变换理论,提出了新的信号处理和滤波模型,并以狗胃电采集实验问题为例说明了基于小波变换的胃电信号处理和滤波算法。分析小波变换的双尺度分解方法及其特征,并与传统的信号处理和滤波算法进行对比分析,实验结果表明,基于小波变换的信号处理和滤波算法是有效的。3. 在基于小波变换的胃电信号处理系统中,小波基函数的种类繁多,其滤波效果差别很大,需要进行量化衡量和优化。因此,提出了基于信息熵、信号能量和联合熵的小波基函数的选择方法来进行小波基函数的优化。通过对小波基函数的定义及其构造的分析,利用胃电的频率和幅值特征作为约束条件,进行小波基函数的构造和小波提升,实现小波基函数的优化选择。并且通过人胃电的采集和分析证明了小波基函数选择方法的有效性。4. 针对糖尿病性胃轻瘫病人,在分析了糖尿病性胃轻瘫疾病的基本病理的基础上,通过对糖尿病性胃轻瘫的常规药物治疗的分析得知,常规药物治疗对病人的肾脏等器官有很大伤害,并且药物治疗主要是以增强胃动力为目的。针对该问题,本文提出了基于胃电信号处理的医疗摩胃机构系统。通过胃起搏器的电穴位刺激以及实时跟踪胃电慢波的摩胃机构以模拟增强胃动力,通过电信号和机械信号来增强胃动力。通过理论和实验分析,基于胃电信号处理的医疗摩胃机构可以起到增强胃动力的效果。基于小波变换的胃电信号采集系统克服了传统的胃电采集系统的缺点。通过小波双尺度变换和傅里叶变换实现了胃电快、慢波的滤波和叠加。这种滤波算法较之于传统的滤波算法在降低各频段噪声及低频滤波方面具有很大的优势。对于提高信噪比和观察快、慢波特征具有现实意义。
Other AbstractWith rapid development of digital signal processing technology, wavelet transform of signal processing techniques has been widely used in digital image processing, signal processing of industrial automation, robotics and bio-signal processing and etc. Wavelet transform based electrogastrogram(EGG)signal acquisition system, which are characterized by use of the wavelet transform denoising and halve frequency to solve these problems of small amplitude and low frequency of EGG, is a new application in digital signal processing, and this solution can be realized by taking advantage of the characteristics of wavelet transform. Such system has the advantage mainly in two aspects compared to conventional EGG system, one is that it improves the signal to noise ratio, and reduce the high-frequency noise of gastric signals, so that more valuable information can be obvious. The other is to achieve precise spike and slow wave bandwidth filters, and making the gastric signals the spike, slow wave signal to be renewed. The use of EGG opened up a new scientific and reliable solution for the diagnosis of stomach, but it also brings a new challenge, the main features of EGG data - small amplitude, low frequency, low signal to noise ratio - severely reduced the practical value of EGG, even lead to unreliable EGG and mistaken diagnosis. Traditional signal processing methods and filtering algorithms are not ideal, filtering results are not precise enough, and there are a lot of noise and interferences. In addition, the traditional filtering algorithms do not consider filtering the spike and slow wave respectively and addition issues. An ideal signal acquisition system is a system with all the features of band signals superimposed or weighted superimposed to reflect true characteristics of EGG reasonably. Based on these principles, this paper propose the EGG (electrogastrogram) signal processing method based on WT (wavelet transform), and applied it in control system analysis and design of medical gastric massage mechanism, aim at methods and applications mentioned above, this paper does some research in the following areas, 1.      To cope with the problems of existing signal acquisition and signal processing system, we propose an EGG acquisition and filtering algorithms architecture based on wavelet transform. The signal processing and filtering algorithm fully take into account the characteristics of gastric electrical amplitude and frequency, and achieve the gastric spike wave superimposed on slow wave, and at last laid the foundation for EGG. 2.      Based on wavelet transform theory, we propose a new signal processing and filtering module, and the problem of dog EGG is adopted as an example to describe the wavelet-based signal processing and filtering algorithms. The method and characteristics of two-scale wavelet decomposition were analyzed, and comparing it with the traditional signal processing and filtering algorithms, the experimental results show that this filter method is effective. 3.      In the wavelet-based signal processing systems, there are a lot of wavelet functions, and the filtering results vary widely. The method needs to be quantitied and optimized. So, we propose selective methods of wavelet basis function based on information entropy, signal energy and joint entropy to optimize wavelet basis functions. Through the analysis of definition of wavelet function and wavelet construction, the characteristics of gastric electrical frequency and amplitude were used as constraints for construction of wavelet function and wavelet lifting to achieve the optimal choice of wavelet basis function. The collection and analysis of human EGG show that the selective methods of wavelet function are effective. 4.      For patients with diabetic gastroparesis, on the basis of the analysis of diabetic gastroparesis basic pathology of the disease diabetic gastroparesis, it is clear that the conventional drug therapy for patients do great harms to kidney and other organs, and the mainly purpose of drug treatment is to enhance gastric motility. For this problem, this paper proposed a medical gastric massage mechanism system based on signal processing of EGG. By electrical acupoint stimulate of gastric pacemaker and real-time tracking gastric slow wave of gastric massage mechanism to simulate the enhanced gastric motility, electrical signals and mechanical signals were used to enhance gastric motility. Through theoretical and experimental analysis, a medical gastric massage mechanism system based on signal processing of EGG is proved that it can enhance gastric motility. Wavelet-based gastric signal acquisition system overcomes the traditional shortcomings of gastric electrical collection system. Through the two-scale wavelet transform and Fourier transform of EGG, spike wave and slow wave filtering and addition were accomplished. This filtering algorithm compared to traditional filtering algorithms has great advantages in reducing noise and filtering low frequency signal. It has great practical significance in improving the signal to noise ratio and observation of characteristics of spike and slow wave.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/9390
Collection机器人学研究室
Recommended Citation
GB/T 7714
丁伟. 基于小波变换的胃电信号处理方法研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2012.
Files in This Item:
File Name/Size DocType Version Access License
基于小波变换的胃电信号处理方法研究.pd(2300KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[丁伟]'s Articles
Baidu academic
Similar articles in Baidu academic
[丁伟]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[丁伟]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.