晚期糖基化终末产物（AGEs）是一种结构多样的化合物, 在人体血糖高于正常范围时, 会大量产生且不能通过自身代谢降解, 具有血糖长期异常的记忆作用。 研究表明AGEs是引起糖尿病及其并发症的重要因素之一, 通过检测体内AGEs的积累情况可以预测糖尿病及其并发症的发生和发展进程。 现有的离体AGEs检测方法存在操作复杂、 时间较长、 成本较高和不易推广等问题; 在体AGEs检测方法存在皮肤色素、 年龄和血红蛋白干扰等问题。 为此, 基于角膜良好的光学特性和AGEs的自体荧光特性, 提出一种角膜晚期糖基化终末产物荧光光谱检测方法。 构建了一套角膜AGEs荧光光谱检测系统, 系统由微型光纤光谱仪、 集成LED激发光源、 Y型12+1光纤和PC端光谱处理显示软件组成。 荧光光谱检测系统采用激发光源波长分别为370和395 nm在暗室条件下对17名志愿者（男性9人, 女性8人, 糖尿病患者4人, 年龄最小15周岁, 最大81周岁）进行数据采集, 得到激发光波长分别为370和395 nm的荧光光谱数据。 为了准确识别荧光光谱中的有用信息, 先截取需要的荧光光谱数据段（450~700 nm）, 然后对其进行去除背景噪声、 归一化、 小波变化等方法处理, 可以将荧光光谱中不明显的荧光峰值进行放大和识别。 实验结果发现, 采用波长为370和395 nm的LED作为激发光源, 检测到角膜发射的荧光光谱范围在420~600 nm内, 并且都分别在450~500, 500~550和550~600 nm三个范围内存在光谱峰值。 根据荧光性物质的荧光峰值与激发光波长无关的原理, 表明两种不同波长的激发光所得到的荧光光谱都是由同一种物质AGEs产生。 对糖尿病患者和正常人的荧光峰值强度进行分析, 显示糖尿病患者的荧光强度明显高于正常人, 表明本研究通过荧光光谱法检测角膜晚期糖基化终末产物具有可行性。

Advanced glycation end products (AGEs) are a kind of compounds with various structures. When the blood sugar is higher than a normal value, it will be produced in large quantities and cannot be metabolized by the body’s own metabolism, and has a memory function of long-term abnormal blood sugar. Studies have shown that AGEs is one of the important factors causing diabetes and its complications. By detecting the accumulation of AGEs in vivo, the occurrence and development of diabetes and its complications can be predicted. The existing in vitro AGEs detection methods have problems of complicated operation, long detection time, high cost and difficulty in promotion; The in vivo AGEs detection methods have problems such as skin pigmentation, age, and hemoglobin interference. Therefore, based on the good optical properties of the cornea and the autofluorescence characteristics of AGEs, a fluorescence spectroscopic detection method for advanced glycation end products of the cornea was proposed. A set of corneal AGEs fluorescence spectrum detection system was constructed. The system consisted of microfiber spectrometer, integrated LED excitation light source, Y-type 12+1 fiber and PC-side spectral processing display software. The fluorescence spectrum detection system was used to collect data from 17 volunteers (9 males, 8 females, 4 diabetics, the youngest 15 years old and the oldest 81 years old) in darkroom conditions. The fluorescence spectrum data that excitation light central wavelengths were 370 and 395 nm were obtained. In order to accurately identify the useful information of fluorescence spectrum, the required fluorescence spectrum data segments (450~700 nm) were intercepted, and then processed them by removing background noise, normalization, wavelet transform, and so on. The above methods could amplify and identify the non-obvious fluorescence peaks in the fluorescence spectrum. The experimental results show that the fluorescence spectrum of corneal is detected within 420~600 nm when the LEDs with wavelengths of 370 and 395 nm are used as excitation sources. Moreover, the fluorescence spectra have peaks in the range of 450~500, 500~550 and 550~600 nm, respectively. According to the principle that the fluorescence peak of fluorescent substances is independent of the excitation wavelength, it is shown that the fluorescence spectra of two different excitation wavelengths are all produced by AGEs. The peak fluorescence intensity of diabetes mellitus patients and normal people were analyzed. The results showed that the fluorescence intensity of diabetes mellitus patients was significantly higher than that of normal people, which indicated that it was feasible to detect advanced glycation end products of the cornea by fluorescence spectroscopy.