节点文献
近红外无创血液成分测量—动态光谱检测理论及信号提取方法研究
Study on the Detect Theories and Signal Acquisition for Dynamic Spectrum
【作者】 王焱;
【导师】 李刚;
【作者基本信息】 天津大学 , 生物医学工程, 2007, 博士
【摘要】 血液中各种化学物质的成分,是临床诊断中常用的重要生理指标。采用近红外光谱法进行无创血液成分检测已经成为生物医学工程领域的研究热点之一。本文将差分光谱概念引入近红外光谱无创人体成分检测技术中,力图解决个体差异以及某些测量条件对检测结果的影响。对利用循环系统中血液的脉动特性采集动脉血液差分光谱的动态光谱检测技术进行了深入的研究,完善了动态光谱信号提取方式,并对动态光谱检测的两个主要干扰因素——接触压力以及散射影响进行了研究。对混浊介质中的透射差分光谱进行了研究。结果显示,只要在检测过程中维持组织初始厚度不变,并获得差分前后的组织厚度变化量。差分光谱由组织光学参数决定,而后者反映了组织成分的浓度信息。文中给出了三种可应用于人体组织成分检测的差分光谱检测方式,其中最为典型的为利用人体血液的脉动现象,从多波长光电容积脉搏波中提取动脉血液差分光谱的动态光谱检测。根据动态光谱的定义,给出了三种动态光谱检测方法。提出在多波长动态光谱检测中存在相移波形误差,并基于误差分析结果,对动态光谱信号检测和提取技术中的检测频率、信号截取位置、信号截取长度等进行了多方面的研究。提出基于多波长动态光谱幅值比的奇异值剔除方法,该方法更适合于在含有基线波动的光电容积脉搏波中,可以很好地区别由于基线漂移引起脉搏波绝对幅值的变化,和基于由于单一光谱幅值异动产生的奇异点。通过实验方法,测定了接触压力的改变对光电容积脉搏波交流分量、直流分量、各波长交流分量比值等特征参数的影响。结果显示,接触压力对光电容积脉搏波的直流分量、交流分量均有明显的影响,但对各波长交流分量间的比值,即归一化动态光谱的影响很小。分别研究了薄层脉动动脉血液层以及各层静态组织的散射作用对动态光谱检测结果的影响。结果显示静态组织对动态光谱检测有一定的影响,且脉动动脉血液中的等效光程长具有一定的波长特性,这些因素在检测中必须予以考虑。动态光谱为近红外光谱技术在无创人体成分检测技术中提供了一个全新的思路,具有广阔的发展前景。本文的研究成果为该方法的成功实施奠定了理论和实验基础。
【Abstract】 The various chemical components present in human body carry important information about health status. Such chemical information also serves as an important indicator to a number of clinical diagnostics and therapeutic effects. Near-infrared spectroscopy receives global attention as a principal non-invasive diagnostic means. But there is still a long way to go for clinic application, because the near infrared (NIR) spectrum technology is disturbed by many factors, and the precision of the final result is restricted. The difference detect method is employed to the non-invasive blood component concentration sensing in this paper, trying to eliminate the influence of the skin and other tissues. Dynamic Spectrum (DS) which obtain the difference spectrum of artery blood from the pulse wave of blood is focused on in this paper.Firstly, the transmittance difference spectrum is studied. Results show that when the basic thickness of tissues and the additional thickness of the interested tissue are known, the difference spectrum is a function of the optical parameters of the tissues, and the latter is relative with the component concentration of the tissues. Three kinds of application of difference spectrum in the non-invasive sensor are proposed in this paper including the DS.Based on the definition of DS, three kinds of dynamic spectroscopy system are described. The phase shift error of the dynamic spectrum among different pulse wave is discussed. The detect frequency, data intercept position and the data length are studied to perfect the DS data processing.A new method to detect outliers of DS is proposed. This method can distinguish the base line wave of the pulse waves and the abnormity change of signal pulse wave, and is fit for the DS abstract.The influence of contact force is studied by experiment. Result shows that the contact force between the photoplethysmographic sensor and the measurement site influenced the photoplethysmographic signal recorded on the fingertip from young healthy subjects, but it has little influence on the rate between the normalized DS.Based on the Monte Carlo simulation, the influence of scatter in the DS detect is researched. The simulation model of signal layer blood and multilayer tissue of finger tip are founded. The normal thickness of blood layer in the models is searched by Monte Carlo simulation based on the statistical result of pulse signal. Results of the simulations show that the absolute magnitudes and relative differences in path-length factors is a function of wavelength, and the signal of DS is relative with the basic thickness of the stable tissues. The first one can be minified evidently by using the subsection equivalent path-length, and the second one indicates that the DS must be normalized before it is used.The dynamic spectrum method proposed a new idea to the NIR non-invasive detection. It has great value in this field. Theory and experimental results disclosed by this dissertation laid solid basis for the successful application of this method.
【Key words】 Near-Infrared Spectroscopy; Non-invasive; Blood Component Concentration; Data Procession; Contact Force; Scatter;