拉曼光谱技术结合主成分分析-支持向量机对砷类矿物药的分类识别研究

为了快速检测砷类矿物药,建立了拉曼光谱技术结合主成分分析(PCA)-支持向量机(SVM)分类识别砷类矿物药的新方法.首先对雄黄、雌黄、信石和砒霜的拉曼光谱进行了归属,并进行了 PCA-SVM分类识别.然后分别测定了五个批次雄黄的拉曼光谱,并进行了 PCA-SVM分类识别,最后对其拉曼光谱的差异进行了分析.该方法快速、方...     

基于拉曼光谱技术的联苯苄唑药物有效成分的快速鉴定和生产溯源研究

有效成分的定性和定量分析是药物分析的核心。实际样品测定中,通常需要进行样品前处理,以消除复杂辅料成分的干扰,此过程耗时长,且难度较大。因此,发展快速定性筛查技术,提高药物分析的工作效率非常必要。本研究以联苯苄唑药品作为模型药物,将拉曼光谱技术和主成分分析(Principal components analysis, PCA)和支持向量机(Support vector machine, SVM)等机器学习算法联用,探讨了其用于药物有效成分快速定性分析的可行性。针对不同药品中辅料成分复杂、组成不一,并且其拉曼光谱谱图与低含量(1%)的联苯苄唑分子的谱图存在明显交叠的问题,以原料药的拉曼光谱为基准,利用PCA方法精准定位和提取成品药拉曼谱图中联苯苄唑分子的特征拉曼光谱信息,实现了准确定性分析。本研究无需获得每一种辅料的特征拉曼光谱谱图,根据拉曼谱图中1600和1650 cm^-1等处谱峰的细微差异,采用PCA方法结合SVM分类器策略,实现了药物生产溯源的准确区分和鉴定。本研究为药物分析研究提供了一种无损快速溯源的分析方法。     

基于主成分分析-Fisher判别分析的食品类塑料瓶物证差分拉曼光谱分类

食品类塑料瓶物证携带许多潜在证据信息,目前针对此类物证的检验研究尚处于探索阶段。利用差分拉曼光谱对46个食品类塑料瓶样品进行检验,依据样品材质及光谱特征峰可将样品分为三类。利用主成分分析（Principal component analysis, PCA）-Fisher判别分析,绘制主成分得分图,构建判别函数,建立分类模型。结果表明,食品类塑料瓶样品具有明显的聚类关系,原始分类与交叉验证分类准确率达到100 %。差分拉曼光谱结合PCA-Fisher判别分析,检验鉴别食品类塑料瓶物证具有一定的科学性。     

基于逆检索-非负最小二乘法的拉曼混合物分析方法研究

拉曼光谱因它可以提供非常丰富的结构信息而被看作是一项"指纹"技术,因此拉曼光谱可以被用作物质的定性识别。并且拉曼光谱具有制样简单,不破坏样品,在几乎所有的环境下都可以采集。通常认为拉曼光谱只能提供纯物质的结构信息,故利用拉曼光谱分析混合物的成分是有难度的。在便携式拉曼光谱仪、光谱数据库和化学计量学的基础上,开发了一种快速的混合物鉴别方法。根据基于小波域的自动精确峰值检测拉曼光谱的特点,对经典的逆搜索过程进行了改进。匹配质量可以用计算混合物和数据库中相减光谱中的负比率(按最小比例计算反向匹配峰的比值),提出一种基于改进的逆搜索和非负最小二乘法(Reverse searching and non-negative least squares,RSearch-NNLS),用于混合物分析。方法包括以下步骤:1)通过Whittaker平滑、ariPLS基线校正以及连续小波变换建立纯物质的拉曼光谱库;2)通过逆检索法对采集到的混合物拉曼光谱进行定性分析;3)根据第2步的结果,使用非负最小二乘法对候选化合物进行比例估算。方法是一种鉴别混合物的方法,具有一定的应用前景。     

基于显微共聚焦拉曼技术对三种食源性致病菌的快速鉴别研究

采用显微共聚焦拉曼技术,建立了对三种常见食源性致病菌快速鉴别的检测方法。使用XploRA PLUS共聚焦拉曼光谱仪,在激光功率为5 mW、积分时间为30s、积分次数为1次的条件下,对德尔卑沙门氏菌、副溶血性弧菌和金黄色葡萄球菌进行了拉曼光谱数据的采集。对拉曼光谱采用多项式平滑算法和荧光背底扣除后,采用主成分分析法(PCA)对预处理后的数据进行降维,提取出前三个主成分的累计方差贡献率达到了95.4％,样本明显的聚为了3类。同时结合Fisher判别分析法(FLD)构建分类模型,对三种样本进行交叉验证,分类准确率达到了100％。结果表明,采用显微共聚焦拉曼技术与PCA-FLD方法结合可实现对三种食源性致病菌的快速准确鉴别且模型检测精度高,方法具有一定的实用性及参考价值。     

基于近红外光谱技术与化学计量学的绿茶无损鉴别方法研究

该文利用近红外光谱技术结合化学计量学方法开发了不同品种绿茶的无损鉴别方法。通过近红外光谱技术得到了8个品种绿茶样品的近红外光谱,比较了单一以及优化组合光谱预处理方法对光谱的影响,利用无监督的主成分分析(PCA)与有监督的线性判别分析方法(LDA)分别构建了茶叶品种鉴别模型。结果表明:对比单一预处理方法,优化组合预处理具有更优的鉴别准确性。标准正态变量变换预处理消除了茶叶样品大小不均造成的光谱散射影响,一阶导数预处理实现了变动背景的消除,减少了基线漂移的影响,突出了图谱中的有效信息,采用二者相结合的预处理方式并结合无监督的主成分分析法可实现较为准确的绿茶样品种类鉴别分析,准确率达75.0%。此外,采用有监督的线性判别分析方法处理原始光谱数据,可达到100%的鉴别准确率,但该方法需提供类别的先验知识。因此,采用近红外光谱技术和化学计量学相结合的手段可实现不同品种绿茶的快速无损鉴别。     

基于表面增强拉曼光谱的市售氢溴酸右美沙芬止咳糖浆的生产溯源和产品质控研究

药物生产过程中的质量控制与市售成品药物的生产溯源是保障消费者用药安全的关键环节。目前常见的药物分析方法存在操作繁琐或抗干扰能力差等问题,实现高效快速的药物分析仍是尚未解决的难题。基于表面增强拉曼光谱（SERS）的指纹图谱高分辨能力和痕量检测灵敏度,结合主成分分析（PCA）方法对弱信号的准确识别和分类能力,该文开展了氢溴酸右美沙芬止咳糖浆成品药的产地溯源（不同厂家）和产品质控（某一厂家）的可行性研究。通过支持向量机（SVM）分类算法验证,SERS-PCA方法对不同厂家分类的准确度为100%。研究结果为推动“SERS+化学计量学”在药物分析领域的实际应用提供了一种切实可行的参考方案。     

基于近红外光谱和OPLS-DA的不同牌号卷烟分类识别方法研究

为了对卷烟牌号进行准确分类鉴别,提出了一种基于近红外光谱(NIRS)分析技术结合有监督的模式识别快速鉴别卷烟牌号的新方法。利用标准正态变量变换(SNV)、多元散射校正(MSC)、一阶导数(FD)、二阶导数(SD)和Savitzky-Golay平滑(SG)及其相结合的光谱预处理方法对烟丝光谱进行预处理,通过近红外光谱结合主成分分析(PCA)、偏最小二乘判别分析(PLS-DA)和正交偏最小二乘判别分析(OPLS-DA) 3种模式识别方法对不同牌号烟丝进行分类识别研究,并采用分类识别正确率作为评价指标。实验结果表明:(1)烟丝近红外光谱主成分得分图交叉重叠,区分不明显,PCA无法识别出5种牌号的成品烟丝;(2)烟丝光谱经MSC+FD预处理后的PLS-DA模型可得到较好的识别效果,校正集和测试集的分类识别正确率分别为100%和98.3%;(3)烟丝光谱经MSC+SD预处理后的OPLS-DA模型的模式识别效果最好,模型对自变量拟合指数(R²X),因变量的拟合指数(R²Y)和模型预测指数(Q²)分别为0.485、0.907和0.74...     

基于高光谱的黑色签字笔墨水种类鉴别方法研究

该文提出了高光谱成像技术结合机器学习快速无损鉴别黑色签字笔墨水种类的新方法。采集36支不同品牌型号的黑色签字笔笔迹的高光谱图像,对每支签字笔笔迹的高光谱图像选取18个感兴趣区域,共提取648个平均光谱作为样本集。对450～950 nm的原始光谱进行Savitzky－Golay平滑、Z－Score标准化和两种组合方法光谱预处理,使用线性判别分析（LDA）和随机子空间－线性判别分析（RSM－LDA）分别构建黑色签字笔墨水种类鉴别模型。实验结果表明：不同预处理方法对RSM－LDA模型的鉴别准确率影响较小,而对于LDA模型,组合预处理具有更优的鉴别准确率;相比LDA模型,RSM－LDA模型分类效果更佳,训练集的平均分类准确率达100%,交叉验证平均分类准确率达99.09%,测试集的平均分类准确率达90.70%,每类样本的准确率、精准率、召回率均高于LDA模型分类结果,模型的接受者操作特征曲线下方面积（AUC值）达0.998 3,模型性能良好。因此,采用高光谱成像技术结合RSM－LDA可实现不同品牌型号黑色签字笔墨水的快速无损鉴别。     

用PCA-LDA对多环芳烃的致癌性进行分类

本文利用主成分提取-线性判别分析(PCA-LDA)模型对多环芳烃(Polycyclic Aromatic Hydrocarbons,PAHs)的致癌性进行分类,与致癌性有关的多环芳烃的表面积、代谢活性区域中心碳原子的离域能、亲电活性区域中心碳原子的离域能以及分子脱毒区的总数四个参数作为模型的输入,用已知致癌性的67个样本作为训练集建立PCA-LDA模型,对10个预测集样本的致癌性进行预测,结果表明:致癌性按高(h)、低(l)、非(n)分类时预测准确率达100%。     

Discrimination of Human and Nonhuman Blood by Raman Spectroscopy and Partial Least Squares Discriminant Analysis

A novel method is reported to discriminate human and animal blood by Raman spectroscopy without complex sample preparation. A partial least squares discriminant analysis model was constructed from a calibration set of Raman spectra from three species of animal blood using 785-nm laser excitation. The discrimination between human and nonhuman blood was calculated by three sigma. Various performance measures, including a blind test and external validation, confirmed the discriminatory performance of the chemometric model. The model provided 100% accuracy in its differentiation between human and nonhuman blood. These results demonstrate that Raman spectroscopy is a promising tool for the discrimination of animal and human blood.     

Rapid Authentication of Olive Oil by Raman Spectroscopy Using Principal Component Analysis

A method of principal component analysis was employed to authenticate genuine olive oil based on Raman spectroscopy, which can reliably distinguish olive oil from other types of oils and can also accurately identify the level of adulteration in a set of olive oil samples contaminated with 5% or more of other types of oils, such as soybean oil, rapeseed oil, sunflower seed oil, and corn oil. The method is very easy, effective, time-saving, and requires minimal sample preparation. Therefore, the method is a promising technique for the rapid authentication application of olive oil.

[Supplementary materials are available for this article. Go to the publisher's online edition of Analytical Letters for the following free supplemental resource(s): Additional text and table]     

漫反射光谱结合化学计量学方法无损分析彩绘文物颜料

为了对彩绘文物颜料进行无损判别分析,该文采集了12种彩绘常用颜料参考样品的漫反射光谱,根据光谱曲线外形将参考样品分为红黄色系及蓝绿色系两类。使用主成分分析(PCA)分别对两个色系参考样品的光谱数据集进行降维,抽提出最有代表性的3个主成分,利用成分得分散点图确认了颜料间的类间差异,并通过成分矩阵探讨了对该差异贡献率最高的光谱波长区间。在此基础上,采用线性判别分析(LDA)对PCA的分析结果进行建模,拟合判别函数并将其应用于2件颐和园仁寿殿上架彩画文物样品颜料的分析。基于样品判别得分散点图与各类别参考样品间的距离,判定文物使用的颜料分别为铁红和群青。研究表明,该文构建的判别函数能够准确区分颜料的种类,可用于彩绘文物样品的无损分析鉴别。文物样品的污染或老化可能会影响分析,需事先进行表面清理。     

Characterization of Gasoline by Raman Spectroscopy with Chemometric Analysis

A rapid Raman spectroscopy protocol is reported to classify gasoline according to its distributor and to identify and quantify common adulterants. Gasoline from three distributors was collected from 19 stations in São Paulo, Brazil. Principal component analysis (PCA) showed specific clusters for each distributor, and partial least squares discriminant analysis (PLS-DA) correctly identified the origin of the samples. To evaluate the technique for the identification and quantification of the adulterants, authentic samples from each distributor were fortified at levels from 2.5 up to 25.0% (v/v) using ethanol, methanol, toluene, and turpentine to obtain 120 altered samples. PCA showed clear separation among the samples with the adulterants and PLS-DA precisely identified the adulterants (478 in 480 predictions by cross-validation), irrespective of the distributor and the concentration. One classification model was used to characterize all distributors. To quantify the adulterants, 36 multivariate calibration models were constructed using partial least squares (PLS), interval PLS, and PLS genetic algorithm for each distributor and for each adulterant. Cross-validation errors of less than 5.0% were obtained for all adulterants regardless of the distributor. Raman spectroscopy and multivariate analysis were shown to be powerful for rapid and inexpensive for the characterization of gasoline origin and the identification and quantification of common adulterants.     

Discrimination of Osteonecrosis and Normal Tissues by Near-Infrared Spectroscopy and Successive Projections Algorithm-Linear Discriminant Analysis

Osteonecrosis of femoral head (ONFH) is a disease characterized by an impaired blood flow in the bone. The pathogenesis is still unknown, which makes an exact diagnosis troublesome and heavily dependent on experience. Exploring the information of molecular level by modern spectroscopy may help to discover the underlying pathogenesis and find its diagnostic application in clinical medicine. The study focuses on the combination of near-infrared (NIR) spectroscopy and classification models for discriminating ONFH and normal tissues. A total of 128 surgical specimens was prepared and NIR spectra were recorded by an integrating sphere. The experiment data set was divided into three subsets, i.e., the training set, validation set, and test set. Successive projection algorithm-linear discriminant analysis (SPA-LDA) was used to compress variables and build the diagnostic model. Partial least square-discriminant analysis (PLS-DA) was used as the reference. Principal component analysis (PCA) was used for exploratory analysis. The results showed that compared to PLS-DA, SPA-LDA provided a more parsimonious model using only seven variables and achieved better performance, i.e., sensitivity of 90.5 and 85%, and specificity of 100 and 95.5% for the validation and test sets, respectively. It indicated that NIR spectroscopy combined with SPA-LDA algorithm was a feasible aid tool for discriminating ONFH from normal tissue.     

Metabolic Profiling of Human Colorectal Cancer Using High Resolution 1H Nuclear Magnetic Resonance Spectroscopy

Colorectal cancer (CRC) is the third commonest malignancy cancer worldwide. Clear understandings of global metabolic profiling of the normal mucosa and cancer tissues are vitally important to aid optimizing the clinical management strategy and understanding CRC biology. We studied metabolic characteristics of 20 CRC and 20 distant normal mucosa tissues extracts from 20 patients using high resolution ¹H NMR spectroscopy in conjunction with multivariate analyses, such as principal component analysis (PCA). Compared with distant normal mucosa tissues, lactate, taurine, ornithine and polyamine were present at significantly higher levels in CRC tissue extracts whereas myo‐inositol was present at significantly lower level. Two metabolites ratios such as myo‐inositol/taurine and myo‐inositol/(ornithine+polyamine) appear to be the most valuable biomarkers for the differentiation CRC from normal mucosa tissues. Our data suggested that HR ¹H NMR spectroscopy combined with multivariate analyses is a potentially useful technology for detecting malignant changes in the normal mucosa tissues, the technique may be further exploited for future CRC biomarker research or identification of targets for therapeutic manipulations.     

Characterization of Automobile Plastics by Principal Component Analysis and Near-Infrared Spectroscopy

Based on near-infrared diffuse reflection spectroscopy, multivariate calibration models for discarded automobile plastic were constructed using principal component analysis and clustering analysis to rapidly characterize four widely employed materials: polypropylene, polyethylene, acrylonitrile butadiene styrene, and polymethylmethacrylate with an accuracy rate of 97%. The method was shown to rapidly discriminate waste automobile plastic.     

元素含量分析应用于樱桃产地溯源

利用电感耦合等离子体质谱仪(ICP-MS)和电感耦合等离子体发射光谱仪(ICP-OES)对来自美国、加拿大和中国3个地区的92份樱桃中51种元素含量进行测定,并结合化学计量学中主成分分析(PCA)和偏最小二乘法-判别分析(PLS-DA),建立了基于元素含量的判别模型对3个地区樱桃进行区分。结果表明,PCA和PLS-DA多元统计模型均可区分中国和美国、加拿大和中国来源的樱桃样品,中国樱桃与美国、加拿大两国樱桃样品具有显著聚集特征,但美国与加拿大两国樱桃样品聚集交叉在一起。樱桃样品的PLS-DA产地溯源鉴定模型显示,Pr、Nd、Ce、Y、Co、Mo、Mn、Dy、Gd、Ni、Ho、Sm、Sr、Er、Ga、U、Na、Yb、Be和Zn 20种元素为主要显著元素变量,根据这20种元素含量分布情况即可构建樱桃的产地溯源鉴定模型。PLS-DA产地溯源鉴定模型对3个产地来源樱桃样品识别率分别是美国50%、加拿大83.3%和中国83.3%。     

Analysis of plant genomic DNAs and the genetic relationship among plants by using surface-enhanced Raman spectroscopy

The study aimed to distinguish genomic DNAs from nine species of plants belonging to six families and analyze their genetic relationship by using surface-enhanced Raman scattering (SERS). The silver nano-colloid and excitation wavelength of 785 nm used in this study yielded excellent quality of the SERS spectra. Raman signals were remarkably enhanced. Although the spectra for the nine species of plants appeared very similar, there were significant differences according to the analysis of variance analysis. There were three strong characteristic peaks. The peak at 625 cm⁻¹ was due to the vibration overlap of C3′-endo/anti deoxyribose, cytosine, and guanine; the one at 715 cm⁻¹ was due to the scissoring vibrations of C2N1C6 of adenine; and that at 1011 cm⁻¹ was due to the stretching vibration of the CO bond of deoxyribose and vibrations of cytosine. The SERS data were smoothed and standardized and evaluated using second derivative analysis, principal component analysis, and hierarchical cluster analysis. A model was established using the data from hierarchical cluster analysis and principal components of the second derivative. The clustering result of this model was highly consistent with the traditional classification of plants; all plant species investigated were correctly clustered into classes according to the cluster distance coefficient among them; the accuracy of clustering was 100%. Chinese cabbage (Brassica pekinensis Rupr.) and green cabbage (Brassica chinensis L.) belonging to Cruciferae, maize (Zea mays L.) and bamboo (Sinocalamus affinis McClure) belonging to Gramineae, and magnolia (Magnolia delavayi Franch.) and champaca (Michelia alba DC.) belonging to Magnoliaceae were clustered into three separate classes, and fern (Nephrolepis auriculata L., Nephrolepidaceae), garlic (Allium sativum L., Amaryllidaceae), and ginkgo (Ginkgo biloba L., Ginkgoaceae) were each clustered into separate classes. These findings suggest that the SERS spectra of plant genomic DNAs can be used to classify species and analyze their genetic relationship. It is an effective and perfect supplement to traditional classification and can form the basis for genetic analysis.