多光谱及漫射层析双模式成像技术用于中医舌诊客观化研究

Tongue diagnosis plays an important role in the syndrome differentiation of traditional Chinese medicine(TCM) diagnose, however,it is difficult to meet the clinical needs with single indicator at present.Dual-mode imaging technology which combined multi-spectra with diffuse optical tomography is intended to be used in this project.Research on tongue optical characteristics by means of building tongue tissue optical parameters and reflect multi-spectral images, and provide quantitative indicators to TCM tongue diagnosis;grouping the collected samples according to age,gender,TCM consititution,TCM syndrome, and hemorheology, and then use multivariate statistical analysis method to find the possible factors which cause changes in tongue optical characteristics,and then establish the basis and characteristics database;using data mining technology like neural net works,partial least squares and support vector machines to build the relationship between the ingredients,microcirculation and TCM consititution,TCM syndrome.This method may provide an objective standard for TCM diagnose and efficacy evaluation, and do try to make further improve of tongue diagnose.

舌诊作为中医望诊的一部分，对于中医诊断辨证起着非常重要的作用，然而现阶段所采用的中医舌诊客观化方法指标单一，难以满足临床需要。本项目拟结合多光谱和漫射层析双模式成像技术，通过构建舌体内部组织光学参数以及构建舌表面反射多光谱图像，研究舌体组织的光学特征参数表达形式，为中医舌诊提供量化指标；对采集样本根据年龄、性别、中医体质、证侯、血液流变学等进行分组，运用多元统计方法进行分析，获取引起舌体光学特征参数变化的影响因素，并建立基础与特征数据库；采用神经网络、偏最小二乘、支撑向量机等数据挖掘的手段，初步建立舌组织、成分、微循环以及中医体质、证型之间的关系，为中医辨证分型以及中医药疗效评估提供客观依据，也为进一步完善发展中医舌诊作出尝试。

针对现有中医舌诊客观化研究中指标单一，信息量少，难以满足中医临床需要的问题，本研究将多光谱技术和漫射层析成像技术引入到舌诊客观化的研究中来，通过搭建多光谱采集平台，获取舌体表面多光谱信息，对舌苔光谱信息进行分析，通过实验发现舌质与舌苔的光谱差异存在一定的规律性，利用该规律，提取特征波长，构建特征指数，为中医舌诊提供定量的客观指标。为了更进一步探究舌体内部组织结构信息，基于漫射层析成像理论，构建舌体三维平板模型，验证了漫射层析成像用于舌体组织内部信息探究的可行性，并针对现有的漫射层析成像技术探测位置不准确的问题以及皮下脂肪层厚度的差异对探测结果影响的问题，提出了Monte Carlo优化方法。本课题在项目进度和经费使用上按照计划进行，完成了预定的研究内容，达到了预期目标。