无创血糖监测是指导糖尿病治疗、重症病人监护以及反映人体糖代谢能力的有效手段，对生命科学研究及临床医学有重要意义。多种无创监测方法中，近红外光谱法占据主导地位。然而受人体组织背景、血液中不同生化成分光谱重叠等干扰的影响，近红外无创血糖监测技术还远没有达到应有的检测精度。我们根据临床动态血糖监测的需求，打破以往常规，提出血糖日变化规律概念及监测方法。该方法针对单体检测，消除了个体差异影响。而且在单日内，人体表皮状态和血液中其它成分变化较小，有利于降低背景干扰。同时由于测量位置固定，温度、压力等因素相对而言也更容易控制。本项目将深入研究无创伤血糖日变化规律监测面临的方法及技术问题，研制一种基于波长渐变滤光片和多元阵列探测器的便于固定的、小型的、能随身佩戴的无创伤血糖监测装置。在此基础上，建立血糖日变化基准模型和子模型，结合临床实验研究近红外光谱无创伤血糖日变化规律监测的可行性及分析精度。

Non-invasive blood glucose monitoring is significant to life science research and clinical medicine. It's an effective means to guide the treatment of diabetes and ICU patient monitoring, as well as to reflect the ability of glucose metabolism in the human body. The near-infrared spectroscopy dominates among a variety of non-invasive monitoring methods. However, influenced by human tissue backgrounds and spectral overlap of different biochemical compositions in blood, the near-infrared noninvasive blood glucose monitoring technology is still far from the required detection accuracy. Based on clinical needs of dynamic glucose monitoring, we put forward the concept of diurnal variation of blood glucose and the monitoring method, which break the normal procedure previously. This method is aimed at individual detection that eliminate the influence of individual difference; In a single day, the status of human skin and other components in blood change a little that avail to reduce the interference of background; because the measurement position is fixed, the temperature , pressure and other factors are also easier to control relatively. This project is going to research the methods and technical matters which are the diurnal variation of non-invasive blood glucose monitoring faced with. Meanwhile, based on wavelength wedge filter and multiple array detectors, we plan to develop a non-invasive blood glucose monitoring device which is small, fixed easily and able to wear. On this foundation, we will establish the diurnal variation of blood glucose baseline model and sub-model, research the feasibility and analysis accuracy of diurnal variation of non-invasive blood glucose monitoring by near-infrared spectrum.