Analysis of biofluids in aqueous environment based on mid-infrared spectroscopy
Analysis of biofluids in aqueous environment based on mid-infrared spectroscopy
复制标题
DOI：
10.1117/1.1917844
复制
发表时间：
2005-05-01
影响因子：
3.5
通讯作者：
Naumann, D
中科院分区：
文献类型：
--
作者：
Fabian, H;Lasch, P;Naumann, D
在这项研究中，我们描述了一种半自动傅里叶变换红外光谱分析方法的液体血清样品，它结合了简单的样品引入高的样品吞吐量。这种新的红外技术的适用性，从一个队列的牛自然感染牛海绵状脑病和对照组的液体血清样品的分析进行了探讨，比较传统的方法的基础上的透射红外光谱干燥的血清膜。对红外数据进行神经网络分析，以区分牛海绵状脑病阴性对照和疾病晚期动物。在训练人工神经网络分类器后，对来自独立外部验证数据集的血清红外光谱进行了分析。通过这种方式，根据数据收集和数据分析的策略，分别实现了90%至96%的灵敏度和84%至92%的特异性。基于这些结果，液体样品技术和干膜方法的生物流体的常规分析的优点和局限性进行了讨论。(c)2005年，由光学仪器工程师学会（Society of Photo-Optical Instrumentation Engineers）主办。
In this study we describe a semiautomatic Fourier transform infrared spectroscopic methodology for the analysis of liquid serum samples, which combines simple sample introduction with high sample throughput. The applicability of this new infrared technology to the analysis of liquid serum samples from a cohort of cattle naturally infected with bovine spongiform encephalopathy and from controls was explored in comparison to the conventional approach based on transmission infrared spectroscopy of dried serum films. Artifical neural network analysis of the infrared data was performed to differentiate between bovine spongiform encephalopathy-negative controls and animals in the late stage of the disease. After training of artifical neural network classifiers, infrared spectra of sera from an independent external validation data set were analyzed. In this way, sensitivities between 90 and 96% and specificities between 84 and 92% were achieved, respectively, depending upon the strategy of data collection and data analysis. Based on these results, the advantages and limitations of the liquid sample technique and the dried film approach for routine analysis of biofluids are discussed. (c) 2005 Society of Photo-Optical Instrumentation Engineers.