光在多成分介质中传输仿真及光谱测量数据处理新方法
发布时间：2018-10-29 08:57
【摘要】：光谱法无创检测多成分介质特定成分含量是检测领域一直以来的研究热点,光在组织介质中的传输和血液成分含量等组织成分检测是生物医学检测领域的前沿热点课题。对于组织中的光传输模型有待继续完善,同时由于测量信号微弱、测量条件和个体差异等影响,近红外光谱检测进入临床实用具有很大的难度。本课题改进的光在多成分介质中的传输仿真模型以及相关的光谱测量数据处理的新方法有助于推进光谱无创检测技术的发展。本文综述了光在组织中的传输的研究成果及光谱测量各个环节中数据处理的相关方法,说明了生物医学光子学测量中需要解决的问题:如组织光传输模型的精确构建;组织吸收光谱的重叠性;测量环境的影响;光谱分析的模型的建立等。本文针对这些问题的解决,做出的主要工作和创新点有:从蒙特卡罗模型及其光在组织中传输的仿真模型出发,针对传统蒙特卡罗模型的应用局限性,改进了模型,进行了光在多成分组织中传输的研究,并做了实际仿真实验验证和应用举例。这个模型后期可用于血液成分检测的精确理论仿真。将经验模态分解应用于光谱测量数据预处理中,用于提高光谱测量的精度。通过介绍了动态光谱血液成分消除个体差异的检测原理和经验模态分解(EMD)用于滤波的基本原理,将EMD引入光谱数据预处理过程中,用来提高光谱数据的性噪比;以动态光谱无创测量血红蛋白浓度中的应用说明了EMD在光谱测量数据处理中的可行性和有效性,为光谱数据处理及其产品化的设计应用提供了新方法。将方差分析应用于在光谱测量数据处理中,用来判定光谱数据及其模型建立的有效性。介绍了试验方差分析(ANOVA)的基本概念以及应用于光谱数据中进行有效性判定的原理和依据,并以ANOVA的F值和p值作为评价指标,在不同噪声水平下分析了光谱数据的仿真模型。以血氧测量作为一个应用举例,说明了ANOVA应用于血氧光谱无创测量中双波长选取的原理,并做了内部机理分析。通过方差分析的方法可以判定噪声水平及评价光谱数据处理分析方法的有效性,进而能提高物质成分定量分析的可靠性。
[Abstract]:Non-invasive detection of specific components in multi-component media by spectral method has been a hot research topic in the field of detection. The transmission of light in tissue media and the detection of tissue components such as blood content are hot topics in the field of biomedical detection. It is necessary to improve the optical transmission model in tissue. At the same time, it is very difficult for near infrared spectroscopy to be used in clinical practice because of the weak signal, measurement conditions and individual differences. The improved simulation model of optical transmission in multi-component medium and the new method of spectral measurement data processing are helpful to promote the development of spectral non-invasive detection technology. In this paper, the research results of optical transmission in tissue and the related methods of data processing in various links of spectral measurement are reviewed. The problems that need to be solved in biomedical photonics measurement are described, such as the accurate construction of tissue optical transmission model; Overlapping of tissue absorption spectrum; influence of measuring environment; establishment of spectral analysis model. In order to solve these problems, the main work and innovation of this paper are as follows: starting from Monte Carlo model and its simulation model of light transmission in organization, aiming at the limitation of application of traditional Monte Carlo model, this paper improves the model. The transmission of light in multi-component tissue is studied, and the simulation results and application examples are given. This model can be used for accurate theoretical simulation of blood component detection in later stage. The empirical mode decomposition (EMD) is applied to the pretreatment of spectral measurement data to improve the precision of spectral measurement. This paper introduces the detection principle of eliminating individual differences in dynamic spectral blood components and the basic principle of empirical mode decomposition (EMD) for filtering. EMD is introduced into the process of spectral data preprocessing to improve the ratio of spectral data to noise. The feasibility and effectiveness of EMD in spectral data processing are illustrated by the application of dynamic spectrum noninvasive measurement of hemoglobin concentration, which provides a new method for the design and application of spectral data processing and its production. Variance analysis is applied to spectral data processing to determine the validity of spectral data and its modeling. This paper introduces the basic concept of test variance analysis (ANOVA) and the principle and basis of validity evaluation applied in spectral data. The F value and p value of ANOVA are taken as evaluation indexes. The simulation model of spectral data is analyzed under different noise levels. Taking the measurement of oxygen in blood as an example, the principle of selecting two wavelengths in non-invasive measurement of blood oxygen spectrum by ANOVA is explained, and the internal mechanism is analyzed. The method of variance analysis can be used to judge the noise level and evaluate the validity of the spectral data processing and analysis method, thus improving the reliability of the quantitative analysis of material composition.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R318.51
本文编号：2297230
[Abstract]:Non-invasive detection of specific components in multi-component media by spectral method has been a hot research topic in the field of detection. The transmission of light in tissue media and the detection of tissue components such as blood content are hot topics in the field of biomedical detection. It is necessary to improve the optical transmission model in tissue. At the same time, it is very difficult for near infrared spectroscopy to be used in clinical practice because of the weak signal, measurement conditions and individual differences. The improved simulation model of optical transmission in multi-component medium and the new method of spectral measurement data processing are helpful to promote the development of spectral non-invasive detection technology. In this paper, the research results of optical transmission in tissue and the related methods of data processing in various links of spectral measurement are reviewed. The problems that need to be solved in biomedical photonics measurement are described, such as the accurate construction of tissue optical transmission model; Overlapping of tissue absorption spectrum; influence of measuring environment; establishment of spectral analysis model. In order to solve these problems, the main work and innovation of this paper are as follows: starting from Monte Carlo model and its simulation model of light transmission in organization, aiming at the limitation of application of traditional Monte Carlo model, this paper improves the model. The transmission of light in multi-component tissue is studied, and the simulation results and application examples are given. This model can be used for accurate theoretical simulation of blood component detection in later stage. The empirical mode decomposition (EMD) is applied to the pretreatment of spectral measurement data to improve the precision of spectral measurement. This paper introduces the detection principle of eliminating individual differences in dynamic spectral blood components and the basic principle of empirical mode decomposition (EMD) for filtering. EMD is introduced into the process of spectral data preprocessing to improve the ratio of spectral data to noise. The feasibility and effectiveness of EMD in spectral data processing are illustrated by the application of dynamic spectrum noninvasive measurement of hemoglobin concentration, which provides a new method for the design and application of spectral data processing and its production. Variance analysis is applied to spectral data processing to determine the validity of spectral data and its modeling. This paper introduces the basic concept of test variance analysis (ANOVA) and the principle and basis of validity evaluation applied in spectral data. The F value and p value of ANOVA are taken as evaluation indexes. The simulation model of spectral data is analyzed under different noise levels. Taking the measurement of oxygen in blood as an example, the principle of selecting two wavelengths in non-invasive measurement of blood oxygen spectrum by ANOVA is explained, and the internal mechanism is analyzed. The method of variance analysis can be used to judge the noise level and evaluate the validity of the spectral data processing and analysis method, thus improving the reliability of the quantitative analysis of material composition.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R318.51
【参考文献】
相关期刊论文 前10条
1 林凌;李威;周梅;曾锐利;李刚;张宝菊;;EMD算法在动态光谱无创测量血红蛋白浓度中的应用[J];光谱学与光谱分析;2014年08期
2 林凌;武若楠;李永城;周梅;李刚;;基于最小二乘法的动态光谱补偿拟合提取[J];光谱学与光谱分析;2014年07期
3 林凌;李威;曾锐利;刘瑞安;李刚;吴晓荣;;基于试验方差分析的血氧测量波长选取的研究[J];光谱学与光谱分析;2014年07期
4 李威;张强武;;用多位置多光程漫透射光谱法分析复杂溶液的Monte Carlo仿真研究[J];红外;2013年10期
5 张婉洁;刘蓉;徐可欣;;基于平行因子分析法提高近红外无创血糖校正模型稳健性的研究[J];化学学报;2013年09期
6 李刚;李哲;李晓霞;林凌;张宝菊;王为;;基于“M+N”理论的光谱分析中光源电压对预测精度影响的研究[J];光谱学与光谱分析;2013年06期
7 张芳君;邵军;钱蔚颖;申会利;白春雄;张少强;;多平面手术联合治疗重度阻塞性睡眠呼吸暂停低通气综合征的临床观察[J];中国耳鼻咽喉颅底外科杂志;2013年01期
8 孙健;;迈瑞PM9000多参数监护仪常见故障及排除方法[J];中国计量;2013年01期
9 张志勇;李刚;林凌;崔新仪;张宝菊;;EMD和SPA算法在光谱法检测面粉过氧化苯甲酰添加量中的应用[J];光谱学与光谱分析;2012年10期
10 熊婵;林凌;王蒙军;李刚;张宝菊;;多维漫透射光谱提高复杂混合溶液分析能力[J];光谱学与光谱分析;2012年08期
，本文编号：2297230
本文链接：https://www.wllwen.com/yixuelunwen/swyx/2297230.html
