利用多光程光谱法检测血液多种成分含量的研究

为研究利用血清多光程光谱信息检测血液成分含量的可行性,通过设计自动微位移测量装置实现了对200份血清样品0～4.0 mm(间隔0.2 mm)近红外多光程光谱的测量。利用多光程光谱的非线性特性,使用血清多光程光谱共同参与建模的方法,对血清样品的葡萄糖(GLU)、总胆固醇(TC)、总蛋白(TP)和白蛋白(ALB)四种成分含量进行定量分析。采用偏最小二乘法(PLS)对160个血清样本的多光程近红外吸收谱与全自动生化分析仪得到的生化分析结果建立校正模型,并对预测集40个样本进行预测。４种血液成分预测值与生化分析结果的相关系数(r)分别为0.932 0,0.971 2,0.946 2和0.948 3。研究结果证明多光程光谱法建模方法用于血液成分含量分析的可行性,为利用多光程光谱法实现方便、快捷的检测血液以及其他液体成分含量奠定了基础。

Research on the Blood Components Detecting by Multi-Optical Path Length Spectroscopy Technique

To discuss the feasibility of using the serum’s multi-optical path length spectroscopy information for measuring the concentration of the human blood components, the automatic micro-displacement measuring device was designed, which can obtain the near-infrared multi-optical path length from 0 to 4.0 mm (interval is 0.2 mm) spectra of 200 serum samples with multi-optical path length spectrum of serum participated in building the quantitative analysis model of four components of the human blood: glucose (GLU),total cholesterol (TC), total protein (TP) and albumin (ALB ), by mean of the significant non-linear spectral characteristic of blood. Partial least square (PLS) was used to set up the calibration models of the multi-optical path length near-infrared absorption spectrum of 160 experimental samples against the biochemical analysis results of them. The blood components of another 40 samples were predicted according to the model. The prediction effect of four blood components was favorable, and the correlation coefficient (r) of predictive value and biochemical analysis value were 0.932 0, 0.971 2, 0.946 2 and 0.948 3, respectively. All of the results proved the feasibility of the multi-optical path length spectroscopy technique for blood components analysis. And this technique established the foundation of detecting the components of blood and other liquid conveniently and rapidly.