自适应的EEMD残余相关基线校正算法

基线校正是光谱分析的重要环节,现有算法通常需要设定关键参数,不具备自适应性。根据总体平均经验模态分解(ensemble empirical mode decomposition,EEMD)残余量特点,提出用残余量拟合光谱基线。通过残余量与信号相关性、残余量自相关和互相关性(称为残余相关准则)判断残余量是否是基线组成部分,以此为基础提出一种自适应的EEMD残余相关基线校正算法。对叠加曲线背景和线性背景的模拟光谱数据进行实验,结果显示在已知基线数学假设情况下,EEMD残余相关法逊于多项式拟合,同非线性拟合相差不多,优于小波分解。在没有光谱背景知识情况下,对真实拉曼光谱数据进行试验。经过上述方法预处理过的玉米叶片光谱采用3层BP神经网络建立与叶绿素之间预测模型,经过残余相关基线校正的模型具有最大校正相关系数和预测相关系数,最小交叉验证标准差和相对分析误差。各种基线校正方法中,残余相关基线校正对特征峰峰位、峰强和峰宽影响最小。实验表明,该算法可用于拉曼谱图基线校正,无需分析样品成分的先验知识,无需选择合适的拟合函数、拟合数据点、拟合阶次以及基函数和分解层数,也无需基线信号分布的数学假设,自适应性很强。     

Crystal → nematic phase transition in the liquid crystalline system 1‐isothiocyanato‐4‐(trans‐4‐propylcyclohexyl) benzene (3CHBT) probed by temperature‐dependent micro‐Raman study and DFT calculations

Raman spectra of 3CHBT in unoriented form were recorded at 14 different temperature measurements in the range 25–55 °C, which covers the crystal → nematic (N) phase transition, and the Raman signatures of the phase transition were identified. The wavenumber shifts and linewidth changes of Raman marker bands with varying temperature were determined. The assignments of important vibrational modes of 3CHBT were also made using the experimentally observed Raman and infrared spectra, calculated wavenumbers, and potential energy distribution. The DFT calculations using the B3LYP method employing 6‐31G functional were performed for geometry optimization and vibrational spectra of monomer and dimer of 3CHBT. The analysis of the vibrational bands, especially the variation of their peak position as a function of temperature in two different spectral regions, 1150–1275 cm⁻¹ and 1950–2300 cm⁻¹, is discussed in detail. Both the linewidth and peak position of the ( C H ) in‐plane bending and ν(NCS) modes, which give Raman signatures of the crystal → N phase transition, are discussed in detail. The molecular dynamics of this transition has also been discussed. We propose the co‐existence of two types of dimers, one in parallel and the other in antiparallel arrangement, while going to the nematic phase. The structure of the nematic phase in bulk has also been proposed in terms of these dimers. The red shift of the ν(NCS) band and blue shift of almost all other ring modes show increased intermolecular interaction between the aromatic rings and decreased intermolecular interaction between two  NCS groups in the nematic phase. Copyright © 2009 John Wiley & Sons, Ltd.     

Deconvolution of pure component FT‐Raman spectra from thermal emission of barium sulfate and graphite samples using the BTEM algorithm

Band‐target entropy minimization (BTEM) was applied for the extraction of pure component Raman spectra from samples exhibiting a significant thermal background due to sample emission. In this method, singular value decomposition was first used to calculate the right singular vectors of the spectroscopic data matrix. Then the non‐noise right singular vectors were examined for localized spectral features, which were subsequently used for spectral recovery. These local features were targeted with the BTEM algorithm to recover the pure component Raman spectra. Accordingly, the interfering thermal background was removed. This method of analysis is applied to graphite and barium sulfate Raman spectroscopic data. Copyright © 2006 John Wiley & Sons, Ltd.     

Raman labeled nanoparticles: characterization of variability and improved method for unmixing

Raman spectroscopy can differentiate the spectral fingerprints of many molecules, resulting in potentially high multiplexing capabilities of Raman‐tagged nanoparticles. However, an accurate quantitative unmixing of Raman spectra is challenging because of potential overlaps between Raman peaks from each molecule, as well as slight variations in the location, height, and width of very narrow peaks. If not accounted for properly, even minor fluctuations in the spectra may produce significant error that will ultimately result in poor unmixing accuracy. The objective of our study was to develop and validate a mathematical model of the Raman spectra of nanoparticles to unmix the contributions from each nanoparticle allowing simultaneous quantitation of several nanoparticle concentrations during sample characterization. We developed and evaluated an algorithm for quantitative unmixing of the spectra called narrow peak spectral algorithm (NPSA). Using NPSA, we were able to successfully unmix Raman spectra of up to seven Raman nanoparticles after correcting for spectral variations of 30% intensity and shifts in peak locations of up to 10 cm⁻¹, which is equivalent to 50% of the full width at half maximum (FWHM). We compared the performance of NPSA to the conventional least squares (LS) analysis. Error in the NPSA is approximately 50% lower than in the LS. The error in estimating the relative contributions of each nanoparticle with the use of the NPSA are in the range of 10–16% for equal ratios and 13–19% for unequal ratios for the unmixing of seven composite organic–inorganic nanoparticles (COINs); whereas, the errors from using the traditional LS approach were in the range of 25–38% for equal ratios and 45–68% for unequal ratios. Here, we report for the first time the quantitative unmixing of seven nanoparticles with a maximum root mean square of the percentage error (RMS%) error of less than 20%. Copyright © 2012 John Wiley & Sons, Ltd.     

A new general-purpose fully automatic baseline-correction procedure for 1D and 2D NMR data

A new procedure for automatic baseline correction of NMR data sets is presented. It is based on an improved automatic recognition of signal-free regions that uses a Continuous Wavelet transform derivative calculation, followed by a baseline modelling procedure based on the Whittaker smoother algorithm. The method has been proven to automatically flatten 1D and 2D NMR spectra with large baseline distortions arising from different sources, is tolerant to low signal-to-noise ratio spectra, and to signals of varying widths in a single spectrum. Even though this procedure has so far only been applied to NMR spectra, we believe it to also be applicable to other spectroscopies having relatively narrow peaks (e.g., mass spectrometry), and potentially to those with broad peaks (e.g., near infrared or ultraviolet).     

拉曼光谱的局域动态移动平均全自动基线校准算法

基线校准是极其重要的光谱预处理步骤,能够显著提高后续光谱分析算法的准确性。目前基线校准算法大多数都是手动或半自动的,手动基线校准算法完全依赖于用户的经验,个人主观因素会严重影响基线校准的准确性,半自动基线校准需要针对不同的拉曼光谱设置不同的优化参数,使用不便。提出了一种局域动态移动平均(LDMA)全自动基线校准算法,并且详细阐明了该算法的基本思想和具体算法步骤。该算法采用了改进移动平均算法(MMA)实现拉曼光谱峰的逐渐剥离,通过自动识别原始拉曼光谱的基线子区间来将整个拉曼光谱区间自动分割为多个拉曼峰子区间,从而实现了在每个拉曼峰子区间中动态改变MMA窗口半宽度和控制平滑迭代次数,最大程度地避免了基线校准过度和基线欠校准现象。无论对于凸形基线、指数形基线、反曲线形基线模拟拉曼光谱,还是真实物质的拉曼光谱,LDMA全自动基线校准算法都取得了很好的基线校准效果。     

基于共焦显微拉曼的柑橘黄龙病无损检测研究

黄龙病危害柑橘果树日益严重,对柑橘黄龙病进行快速检测研究具有重大意义。采用拉曼光谱技术,结合偏最小二乘判别分析(PLS-DA)方法探讨快速诊断柑橘黄龙病及病情类别的可行性。获取柑橘叶片拉曼光谱并进行普通PCR鉴别分为轻度、中度、重度、缺素和正常5类。在715～1 639.5 cm-1范围内采用一阶导,基线校正(Baseline)和多项式拟合三种方法扣除光谱背景,突显叶片拉曼光谱特征峰。多项式拟合方法分别进行了2次,3次和4次拟合,与一阶导和基线校正两种扣除背景方法进行比较,结合最小二乘支持向量机(LS-SVM)和偏最小二乘判别分析(PLS-DA)建立判别模型。经比较发现,多项式拟合方法扣除光谱背景效果均好于另外两种方法,其中用2次多项式拟合的PLS-DA模型的效果最好,预测相关系数(R_P)为0.98,预测均方根误差(RMSEP)为0.67,总误判率最小为0。基线校正扣除光谱背景的LS-SVM模型效果最差,总误判率最大为40%。研究结果表明,利用拉曼光谱技术对柑橘黄龙病进行快速识别研究具有一定的可行性,为柑橘黄龙病无损检测研究提供一种新途径。     

Quantitative,comparable coherent anti‐Stokes Raman scattering (CARS) spectroscopy: correcting errors in phase retrieval

Coherent anti‐Stokes Raman scattering (CARS) microspectroscopy has demonstrated significant potential for biological and materials imaging. To date, however, the primary mechanism of disseminating CARS spectroscopic information is through pseudocolor imagery, which explicitly neglects a vast majority of the hyperspectral data. Furthermore, current paradigms in CARS spectral processing do not lend themselves to quantitative sample‐to‐sample comparability. The primary limitation stems from the need to accurately measure the so‐called nonresonant background (NRB) that is used to extract the chemically sensitive Raman information from the raw spectra. Measurement of the NRB on a pixel‐by‐pixel basis is a nontrivial task; thus, surrogate NRB from glass or water is typically utilized, resulting in error between the actual and estimated amplitude and phase. In this paper, we present a new methodology for extracting the Raman spectral features that significantly suppresses these errors through phase detrending and scaling. Classic methods of error correction, such as baseline detrending, are demonstrated to be inaccurate and to simply mask the underlying errors. The theoretical justification is presented by re‐developing the theory of phase retrieval via the Kramers–Kronig relation, and we demonstrate that these results are also applicable to maximum entropy method‐based phase retrieval. This new error‐correction approach is experimentally applied to glycerol spectra and tissue images, demonstrating marked consistency between spectra obtained using different NRB estimates and between spectra obtained on different instruments. Additionally, in order to facilitate implementation of these approaches, we have made many of the tools described herein available free for download. Published 2015. This article is a U.S. Government work and is in the public domain in the USA.     

Two-dimensional correlation analysis of Raman optical activity data on the α-helix-to-β-sheet transition in poly(L-lysine)

Raman optical activity (ROA) has evolved into an incisive probe of structure and conformational transitions in polypeptides and proteins revealing many signal patterns characteristic of specific secondary structural elements. In order to further facilitate analysis of ROA spectral intensity variations, two-dimensional correlation methods are applied to ROA and Raman spectra monitoring the α-helix-to-β-sheet transition in poly(L-lysine) as a function of temperature. Pretreatment of data using background subtraction, normalization and gentle smoothing is essential for the successful generation of 2D ROA correlations, 2D Raman correlations and 2D Raman/ROA heterocorrelations. The pseudoscalar nature of ROA spectra results in detailed 2D correlation analyses providing extensive interpretation of spectral intensity variations. Synchronous plots indicate band assignments consistent with established assignments in poly(L-lysine) together with possible new assignments. Corresponding asynchronous plots probe the temporal sequence of the conformational transition indicating distinct temporal phases while monitoring aggregation through a small amount of β-structure present at the start of the experiment ahead of α-helix unfolding. This study demonstrates the potential of 2D correlation analysis as a valuable technique for the extraction of detailed information about aggregation and conformational transitions in polypeptides and proteins from associated ROA and Raman spectra. Results indicate that aggregation of poly(L-lysine) monomers precedes intramolecular conversion of α-helix to β-sheet, which is then followed by fibril formation.     

Model‐based pre‐processing in Raman spectroscopy of biological samples

Model‐based pre‐processing has become wide spread in spectroscopy and is the standard procedure in Fourier‐transform infrared spectroscopy. It has also been shown to give valuable contributions in Raman spectroscopy. Extended multiplicative signal correction is flexible enough to handle varying fluorescence background and take into account individual variations in baselines while still keeping enough rigidity through reference spectra and model fitting to avoid degenerate solutions and overfitting, when used correctly. We demonstrate the basic extended multiplicative signal correction method and some extensions, including a novel shift correction, on real Raman data to demonstrate effects on visual appearance, replicate variation and prediction. Comparisons with other standard correction methods are also shown and discussed. © 2016 The Authors. Journal of Raman Spectroscopy Published by John Wiley & Sons, Ltd.     

Understanding tip‐enhanced Raman spectra of biological molecules: a combined Raman,SERS and TERS study

Although conventional Raman, surface‐enhanced Raman (SERS) and tip‐enhanced Raman spectroscopy (TERS) have been known for a long time, a direct, thorough comparison of these three methods has never been carried out. In this paper, spectra that were obtained by conventional Raman, SERS (on gold and silver substrates) and TERS (in ‘gap mode’ with silver tips and gold substrates) are compared to learn from their differences and similarities. Because the investigation of biological samples by TERS has recently become a hot topic, this work focuses on biologically relevant substances. Starting from the TER spectra of bovine serum albumin as an example for a protein, the dipeptides Phe–Phe and Tyr–Tyr and the tripeptide Tyr–Tyr–Tyr were investigated. The major findings were as follows. (1) We show that the widely used assumption that spectral bands do not shift when comparing SER, TER and conventional Raman spectra (except due to binding to the metal surface in SERS or TERS) is valid. However, band intensity ratios can differ significantly between these three methods. (2) Marker bands can be assigned, which should allow one to identify and localize proteins in complex biological environments in future investigations. From our results, general guidelines for the interpretation of TER spectra are proposed. Copyright © 2012 John Wiley & Sons, Ltd.     

Tip‐enhanced Raman spectroscopy – an interlaboratory reproducibility and comparison study

Since its first experimental realization, tip‐enhanced Raman spectroscopy (TERS) has emerged as a potentially powerful nanochemical analysis tool. However, questions about the comparability and reproducibility of TERS data have emerged. This interlaboratory comparison study addresses these issues by bringing together different TERS groups to perform TERS measurements on nominally identical samples. Based on the spectra obtained, the absolute and relative peak positions, number of bands, peak intensity ratios, and comparability to reference Raman and surface‐enhanced Raman spectroscopy (SERS) data are discussed. Our general findings are that all research groups obtained similar spectral patterns, irrespective of the setup or tip that was used. The TERS (and SERS) spectra consistently showed fewer bands than the conventional Raman spectrum. When comparing these three methods, the spectral pattern match and substance identification is readily possible. Absolute and relative peak positions of the three major signals of thiophenol scattered by 19 and 9 cm⁻¹, respectively, which can probably be attributed to different spectrometer calibrations. However, within the same group (but between different tips), the signals only scattered by 3 cm⁻¹ on average. This study demonstrated the suitability of TERS as an analytical tool and brings TERS a big step forward to becoming a routine technique. Copyright © 2014 John Wiley & Sons, Ltd.     

Depth‐resolved in vivo micro‐Raman spectroscopy of a murine skin tumor model reveals cancer‐specific spectral biomarkers

We have developed a micro‐Raman spectrometer system for use to differentiate tumor lesions from normal skin using an in vivo animal model. A study of 494 Raman spectra from 24 mice revealed different spectral patterns at different depths and between normal and tumor‐bearing skin sites. A peak at 899 cm⁻¹ (possibly from proline or fatty acids) and one with higher intensity in the 1325–1330 cm⁻¹ range (assigned to nucleic acids) were correlated with the presence of tumors, which can potentially be used as biomarkers for skin cancer detection. Spectral diagnosis performed on the murine tumor model achieved a diagnostic sensitivity of 95.8% and specificity of 93.8%. These results encourage us to develop further the use of confocal Raman spectroscopy as a clinical tool for noninvasive human skin biochemical analysis, particularly in relation to skin cancer. Copyright © 2010 John Wiley & Sons, Ltd.     

Rapid screening of sildenafil and tadalafil adulterated in healthcare products by Micro‐Raman spectroscopy

Sildenafil and tadalafil are inhibitors of phosphodiesterase type 5, which are frequently added into healthcare products. The objective of this study was to evaluate the possibility of using micro‐Raman spectroscopy as a non‐destructive technique to screen for sildenafil and tadalafil in adulterated healthcare products. Using a viewing microscope, the suspect area of healthcare products was selected, which had a discernable crystal form or shape from the surrounding zone. Optimization of instrumental parameters of the Raman spectrometer was chosen to reduce the background fluorescence, and the Raman spectra were collected. The spectra collected were compared with the standard Raman spectra of pure sildenafil and tadalafil. Samples with an identifiable Raman signature to that of sildenafil or tadalafil could be confirmed using liquid chromatography–mass spectrometry (LC/MS). Additionally, wavelet denoising combined with similarity calculation was used to establish an automated approach for discrimination of adulterated healthcare products. Correlation coefficient was chosen for similarity calculation based on the spectra collected and the standard Raman spectra of pure sildenafil and tadalafil. We compared ten samples, secured by administrative authorities in Shanghai, to analyse and demonstrate the capabilities of our proposed method. We established six samples containing sildenafil or tadalafil warranting analysis using LC/MS. Thus, the use of micro Raman spectroscopy provides a quick, convenient and non‐destructive method for screening adulterated chemicals in healthcare products. Raman spectroscopy combined with similarity calculation requires little training after spectra library is developed, thus showing great promise to identify the adulterated healthcare products in the future. Copyright © 2012 John Wiley & Sons, Ltd.     

An intelligent background‐correction algorithm for highly fluorescent samples in Raman spectroscopy

Fluorescent background is a major problem in recoding the Raman spectra of many samples, which swamps or obscures the Raman signals. The background should be suppressed in order to perform further qualitative or quantitative analysis of the spectra. For this purpose, an intelligent background‐correction algorithm is developed, which simulates manual background‐correction procedure intelligently. It basically consists of three aspects: (1) accurate peak position detection in the Raman spectrum by continuous wavelet transform (CWT) with the Mexican Hat wavelet as the mother wavelet; (2) peak‐width estimation by signal‐to‐noise ratio (SNR) enhancing derivative calculation based on CWT but with the Haar wavelet as the mother wavelet; and (3) background fitting using penalized least squares with binary masks. This algorithm does not require any preprocessing step for transforming the spectrum into the wavelet space and can suppress the fluorescent background of Raman spectra intelligently and validly. The algorithm is implemented in R language and available as open source software ( http://code.google.com/p/baselinewavelet ). Copyright © 2009 John Wiley & Sons, Ltd.     

Polarization‐resolved high‐resolution Raman spectroscopy with a light‐emitting diode

We demonstrate the use of a light‐emitting diode (LED) based experimental setup for collecting polarization‐resolved Raman spectra with good spectral resolution. The combination of a commercial red LED (630 nm), a 1‐nm bandwidth laser‐line filter, and a polarizing prism is used as a light source. Polarization‐resolved spectra in dimethyl sulfoxide are recorded and compared with the corresponding laser‐Raman spectra. The LED‐excited spectra exhibit a resolution slightly lower than those in the laser case but still close to the resolution of the spectrometer. All relevant spectral features of dimethyl sulfoxide including the symmetric and antisymmetric stretching modes of the CSC moiety are resolved with the experimental setup providing a spectral resolution of approximately 20 cm⁻¹. Copyright © 2013 John Wiley & Sons, Ltd.     

A non‐destructive approach for doping profiles characterization by micro‐Raman spectroscopy: the case of B‐implanted Ge

B‐implanted Ge samples have been investigated by micro‐Raman spectroscopy under different excitation wavelengths, with the aim of gaining insights about the B distribution at different depths beneath the sample surface. The intensities, observed under the different excitation wavelengths, of the B–Ge Raman peak at about 545 cm⁻¹, which is due to the local vibrational mode of the substitutional B atoms in the Ge matrix, have been used to calibrate the optical absorption lengths in B‐implanted Ge. Then, by using these calibrated values, a very sharp correlation between the spectral features of the Ge–Ge Raman peak at ~300 cm⁻¹ and the content of substitutional B atoms has been derived. Accordingly, a non‐destructive approach, based on micro‐Raman spectroscopy under different excitation wavelengths, is presented to estimate, at least at the lowest depths, the carrier concentration profiles from the spectral features of the Ge–Ge Raman peak. Copyright © 2014 John Wiley & Sons, Ltd.     

基于自动线性拟合的快速拉曼基线校正算法

近年来拉曼光谱以其无创、灵敏度高等众多优点在化学表征、生物医药、材料等领域引起广泛关注,而基线漂移的存在为后续的定性定量分析带来严重困扰,因此设计高性能的基线校准算法以提高分析结果的有效性及准确性具有重要意义。针对传统算法在批量拉曼光谱数据基线校正方面的不足,基于自动线性拟合算法提出一种快速基线校正算法以校正具有相似背景的批量拉曼光谱数据并详细阐述了该算法的核心思想以及算法实现流程。该算法首先从批量拉曼光谱数据中自动选择一条拉曼光谱数据作为基准光谱,使用自动线性拟合算法对其进行基线校准得到其基线以及分段标记点,然后利用标记点快速计算出组内其他与基准光谱具有较高相关性的拉曼光谱数据的基线,对于组内与基准光谱相关性不满足阈值要求的拉曼光谱则使用自动线性拟合算法对其进行单独基线校正,这使得算法具有具有较强的鲁棒性,可以适应复杂的拉曼光谱基线校正情形。分别使用快速基线校正算法与单独基线校正算法对多组实际拉曼光谱数据进行基线校正以对比分析算法基线校正效果,结果表明该算法可以实现对批量拉曼光谱数据的快速校正,基线校正效果良好,并且相较于单独进行基线校正算法耗时减少了30%以上,算法无参,简单易行,无需额外人工干预,是一种切实可行的批量拉曼数据自动基线校正算法。     

Near‐infrared Raman spectroscopy for gastric precancer diagnosis

Raman spectroscopy is a molecular vibrational spectroscopic technique that is capable of optically probing the biomolecular changes associated with neoplastic transformation. The purpose of this study was to apply near‐infrared (NIR) Raman spectroscopy for differentiating dysplasia from normal gastric mucosa tissue. A total of 65 gastric mucosa tissues (44 normal and 21 dysplasia) were obtained from 35 patients who underwent endoscopy investigation or gastrectomy operation for this study. A rapid NIR Raman system was utilized for tissue Raman spectroscopic measurements at 785‐nm laser excitation. High‐quality Raman spectra in the range of 800–1800 cm⁻¹ can be acquired from gastric mucosa tissue within 5 s. Raman spectra showed significant differences between normal and dysplastic tissue, particularly in the spectral ranges of 850–1150, 1200–1500 and 1600–1750 cm⁻¹, which contained signals related to proteins, nucleic acids and lipids. The diagnostic decision algorithm based on the combination of Raman peak intensity ratios of I₈₇₅/I₁₄₅₀ and I₁₂₀₈/I₁₆₅₅ and the logistic regression analysis yielded a diagnostic sensitivity of 90.5% and specificity of 90.9% for identification of gastric dysplasia tissue. This work demonstrates that NIR Raman spectroscopy in conjunction with intensity ratio algorithms has the potential for the noninvasive diagnosis and detection of precancer in the stomach at the molecular level. Copyright © 2009 John Wiley & Sons, Ltd.     

Simple and inexpensive instrument for deep‐UV Raman spectroscopy

Deep‐UV Raman spectroscopy is a powerful way to collect chemically specific information about complex samples. The availability of inexpensive and reliable light sources in the spectral region below 250 nm has been always considered a major bottleneck problem on the way of a widespread of this powerful spectroscopic technique. We report on the efficient fourth‐harmonic generation of a low‐power microchip Nd:YAG laser operating at 946 nm. High‐quality deep‐UV Raman spectra were collected using a newly developed laser source. Copyright © 2013 John Wiley & Sons, Ltd.