基于脉搏特征参数和面部温度分布特性的女性气血病理的决策分析

发布时间：2018-01-22 19:55

本文关键词： 脉搏波 AR模型谱估计三阶微分红外热图气血理论　出处：《昆明理工大学》2016年硕士论文　论文类型：学位论文

【摘要】：传统中医理论中,认为气血是生命的根本,人体最需要的就是气和血,尤其对女性来说,气血影响着身体的方方面面。脉搏信号可以很好地表征出人体的脏腑信息、血管信息乃至气血信息等各种重要的生理病理信息,这些信息均反映在波形轮廓的变化之中,这即是中医脉诊的理论基础。随着科技的进步,中医脉诊的客观化、数字化越来越多地应用在疾病诊断方面,人们往往通过提取脉搏波形的变化信息,将其与不同病理对应,从而辅助医生进行疾病诊断。面部黄褐斑是中医气血病理方面的一个典型病症。医学上,对黄褐斑的诊断有其相应的标准,包括表示面积大小及颜色深浅的分级和表示发生部位的分型这两个方面。然而,诊断结果依赖于医生的主观判断,缺乏客观统一标准,故此,本课题提出通过对拍摄到的人体面部红外热图进行分析,获得面部温度分布,进而对病人面部黄褐斑的情况进行辅助判断。黄褐斑病人常常伴有月经不调的症状,它们的病因病机存在着相关性。月经不调的病症对应着不同的中医证型。本课题从脉搏信号时域特征点的提取入手,对气血病理典型病症的证型与女性患者的生理病理特点进行总结整理,探讨脉搏信号特征参数与女性气血病理相关病症证型、病症表现的对应关系,从而辅助医生对本文研究的女性疾病的具体病因作出正确的判断。本课题的研究内容主要包括以下几个方面：(1)脉搏波的产生机理以及脉搏信号特征与人体生理病理的关系。脉搏为人体血液循环系统形成,其波形受到各种身体状态的影响,与人体的生理信息息息相关,对于脉搏波特征与疾病对应关系的研究对于临床诊断及治疗都具有很重要的意义。(2)脉搏信号的采集。本课题选用HK-2000C脉搏传感器来采集人体的脉搏信号,所有脉搏数据均在云南省中医院妇科门诊部实际采集而来。(3)脉搏信号的预处理。实际采集的脉搏信号包含着各种噪声和无效成分。本课题提出了一种基于AR模型谱估计的方法来去除脉搏信号中的随机噪声,采用改进的LIP算法对整个脉搏信号进行周期划分,采用三次样条差值的方法去除基线漂移,采用非匹配延拓算法对信号中的无效成分从后往前依次进行切除,并提取信号的平均波形,对信号在单周期内进行归一化处理。(4)脉搏波特征信息的提取。本课题选择提取脉搏信号的时域特征信息,具体为用特征值法来提取脉搏信号的六个时域特征点。在总结其他特征点提取算法的基础上,分析了六个特征点在波形形态上的特殊意义之后,提出了一种改进的微分算法,该算法针对微分法对噪声的鲁棒性差的缺点,将AR模型去噪、小波单支重构算法与一阶微分、三阶微分相结合,使得对于特征点的提取更准确有效。(5)面部黄褐斑分布情况的判断。本课题提出了两种方法对某红外热像仪拍摄到的红外热图进行人体面部温度分布的提取,将提取出的温度数据按照原图像阵列的顺序保存在excel表格中,设计了一个界面,实现对某个温度值或温度范围内温度数据位置的确定,帮助我们分析研究黄褐斑的分布情况。(6)分析脉搏信号特征信息与病理信息的对应关系。首先根据患者的病例信息,对脉搏信号特征信息与黄褐斑、月经不调这两种气血病理相关病症的病理数据作统计归类,分析总结脉搏信号特征、月经不调的中医证型以及黄褐斑分级分型情况的相关性。
[Abstract]:The traditional Chinese medicine theory, that blood is essential for life, the human body is most in need of Qi and blood, especially for women, and affects all aspects of the body. The pulse signal can well characterize the body organs, blood vessels and blood physiological and pathological information information information and other important information, these are reflect changes in the waveform profile, which is the theoretical basis of TCM. With the progress of science and technology, the objectification of pulse diagnosis digitization, more and more applications in disease diagnosis, the change information extraction pulse waveform, and the corresponding different pathology, to assist doctors to diagnose disease face. Chloasma is a typical disorder of Qi and blood in TCM Pathology. The medicine, with corresponding diagnostic criteria of chloasma, including grading size and color depth and table In place of the type of the two aspects. However, the diagnosis results depend on the subjective judgment of the doctor, the lack of objective standard, therefore, this paper based on the human face of infrared shot to get facial analysis, temperature distribution, and then the patients with facial chloasma chloasma patients for auxiliary judgment. Often accompanied by irregular menstruation symptoms, etiology and pathogenesis of their correlation exists. Menstruation disorders correspond to different TCM type. This topic from the feature point extraction in time domain pulse signal with the physiological and pathological characteristics and syndromes of female patients with typical symptoms of pathological blood were summarized and finishing of pulse the signal characteristic parameters associated with female blood pathological disorders syndrome, corresponding relationship between the symptoms of the disease, the specific cause of this research to assist doctors in women's diseases to make the right judgments. The main research contents of this dissertation include the following aspects: (1) the pulse wave generation mechanism and characteristics of pulse signals and the physiological and pathological relationship. The pulse for human blood circulation system is formed, the waveform is affected by all kinds of the state of the body, and is closely related to human physiological information, research for the corresponding wave characteristics and the relationship between pulse disease it has a very important significance for clinical diagnosis and treatment. (2) the pulse signal acquisition. The project uses HK-2000C pulse sensor to collect the pulse signal of human body, all the data are in the outpatient department of gynecological pulse of traditional Chinese medicine hospital of Yunnan province and the actual acquisition. (3) pretreatment pulse signal. The pulse signal actually collected contains all kinds of noise and invalid components. This paper presents a method of AR model based on spectrum estimation to remove the random noise of the pulse signal, using the improved LIP algorithm Cycle of the whole pulse signal, remove the baseline drift using three spline interpolation method, a non matching extension algorithm for resection of the invalid components in the signal from the back after the turn, and extract the average waveform signal, the signal was normalized in a single cycle. (4) the characteristics of pulse wave information extraction. The extraction time domain characteristic information of pulse signal, specifically to extract six time domain features of pulse signal with the characteristic value method. On the basis of other feature extraction algorithm, after analyzing the special significance of the six feature points in the wave form, and proposes an improved differential the proposed algorithm, the differential method for noise robustness is poor, the AR model of wavelet denoising, single branch reconstruction algorithm and first-order differential, three order differential combination makes the feature extraction more accurately and effectively. (5 Facial chloasma) distribution of judgment. This paper puts forward two methods of infrared thermography infrared imager captured by extracting the human face temperature distribution, the temperature data will be extracted is stored in the excel table in accordance with the array of original image sequence, designed an interface, realize the determination of a temperature the value of temperature or temperature range data location, help us analyze the distribution of chloasma. (6) to analyze the corresponding relationship between information and pathological information characteristics of pulse signals. According to the cases of information of the pulse signal feature information and chloasma, irregular menstruation of the two pathological conditions associated with pathological data of Qi and blood for statistical classification, analyze and summarize the characteristics of pulse signals, the correlation between TCM syndrome types of irregular menstruation and chloasma classified situation.

【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：R241

【相似文献】