CN112712108A - Raman spectrum multivariate data analysis method - Google Patents
Raman spectrum multivariate data analysis method Download PDFInfo
- Publication number
- CN112712108A CN112712108A CN202011482853.5A CN202011482853A CN112712108A CN 112712108 A CN112712108 A CN 112712108A CN 202011482853 A CN202011482853 A CN 202011482853A CN 112712108 A CN112712108 A CN 112712108A
- Authority
- CN
- China
- Prior art keywords
- raman spectrum
- spectral
- data
- classification
- analysis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001237 Raman spectrum Methods 0.000 title claims abstract description 49
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000000491 multivariate analysis Methods 0.000 title claims abstract description 17
- 230000003595 spectral effect Effects 0.000 claims abstract description 36
- 238000013145 classification model Methods 0.000 claims abstract description 24
- 238000010606 normalization Methods 0.000 claims abstract description 21
- 238000000513 principal component analysis Methods 0.000 claims abstract description 20
- 238000007781 pre-processing Methods 0.000 claims abstract description 18
- 238000004458 analytical method Methods 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 15
- 238000002790 cross-validation Methods 0.000 claims abstract description 13
- 238000001069 Raman spectroscopy Methods 0.000 claims abstract description 12
- 230000035945 sensitivity Effects 0.000 claims abstract description 11
- 238000001514 detection method Methods 0.000 claims abstract description 6
- 238000012360 testing method Methods 0.000 claims abstract description 6
- 238000001543 one-way ANOVA Methods 0.000 claims abstract description 3
- 239000011159 matrix material Substances 0.000 claims description 21
- 238000012706 support-vector machine Methods 0.000 claims description 17
- 238000001228 spectrum Methods 0.000 claims description 16
- 238000010239 partial least squares discriminant analysis Methods 0.000 claims description 13
- JJWKPURADFRFRB-UHFFFAOYSA-N carbonyl sulfide Chemical compound O=C=S JJWKPURADFRFRB-UHFFFAOYSA-N 0.000 claims description 6
- 229910010272 inorganic material Inorganic materials 0.000 claims description 6
- 239000011147 inorganic material Substances 0.000 claims description 6
- 239000011368 organic material Substances 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000009499 grossing Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 238000000354 decomposition reaction Methods 0.000 claims description 3
- 230000010365 information processing Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 5
- 238000000605 extraction Methods 0.000 description 4
- 239000000126 substance Substances 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 2
- 238000000556 factor analysis Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- -1 chemical engineering Substances 0.000 description 1
- 238000003889 chemical engineering Methods 0.000 description 1
- 230000035572 chemosensitivity Effects 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 238000004624 confocal microscopy Methods 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000002329 infrared spectrum Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 229920002521 macromolecule Polymers 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004001 molecular interaction Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2132—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on discrimination criteria, e.g. discriminant analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
本发明涉及一种拉曼光谱多元数据分析方法，具体包括以下步骤：利用拉曼光谱检测仪器采集各种样品的原始拉曼光谱数据；通过自主开发分析方法对原始数据进行预处理；在光谱数据预处理基础上开展归一化和均值中心化数据处理；针对预处理后的数据选择使用主成分分析或最小二乘判别分析提取光谱特征数据，并分别利用单因素方差分析和交叉验证选取所述光谱数据中的显著特征成分；结合分类模型进行光谱识别，使用无偏留一法交叉验证来评估各分类模型可靠性；选取剩余样本进行测试，得到样本分类的准确率，敏感性，特异性，以及模型的受试工作者特征曲线评估系统的识别性能。本发明可广泛应用于拉曼光谱信息处理及其光谱特征计算机识别领域。
The invention relates to a Raman spectrum multivariate data analysis method, which specifically includes the following steps: collecting original Raman spectrum data of various samples by using a Raman spectrum detection instrument; preprocessing the original data by independently developing an analysis method; On the basis of preprocessing, normalization and mean centering data processing are carried out; for the preprocessed data, principal component analysis or least squares discriminant analysis is used to extract spectral characteristic data, and one-way analysis of variance and cross-validation are used to select the Significant feature components in spectral data; spectral identification combined with classification models, and unbiased leave-one-out cross-validation to evaluate the reliability of each classification model; select remaining samples for testing to obtain sample classification accuracy, sensitivity, specificity, As well as the subject-worker characteristic curve of the model to evaluate the recognition performance of the system. The invention can be widely used in the fields of Raman spectral information processing and computer identification of spectral characteristics.
Description
Technical Field
The invention relates to the field of Raman spectrum information processing and spectral feature computer identification, in particular to a Raman spectrum multivariate data analysis method.
Background
Raman spectroscopy is based on the interaction of light and chemical bonds within materials and is a non-destructive analytical technique that can yield detailed information about the chemical structure, phase and morphology, crystallinity, and molecular interactions of a sample. The raman spectrum can also be used to shift the molecular energy spectrum in the infrared region to the visible region for detection. Therefore, the Raman spectrum is a powerful weapon for researching the structure of molecular substances as a supplement of the infrared spectrum. With the development and progress of science and technology, the Raman spectrum technology is applied to multiple fields such as petroleum, chemical engineering, materials, biology, environmental protection, geology and the like, and provides more information on molecular structures for the development of various industries.
Currently, raman spectroscopy has been developed as one of the most important techniques in the basic and applied scientific research in the field of analytical science. Due to the technical characteristics of molecular sensitivity, easy implementation, water environment applicability and the like, the Raman spectrum analysis technology is also widely applied to other multidisciplinary research fields. Furthermore, recent developments have combined the chemosensitivity and specificity of raman scattering with the high spatial resolution of confocal microscopy to reconstruct image information that yields the biochemical makeup of the sample. Nevertheless, the wide application of raman spectroscopy and its related analysis techniques is limited by some technical difficulties. Firstly, raman scattering is a weak optical phenomenon, and the generated spectral information (i.e., raman spectrum) is easily interfered by the environment and external factors; secondly, in a complex biochemical environment or other systems, different types of biological macromolecules contain similar biochemical components, so that the Raman spectrum has the phenomena of spectral peak position overlapping, spectral peak intensity non-uniformity and spectral peak width (half-height width) extension.
Based on the background, the multivariate data analysis method of the Raman spectrum is provided, and on the basis of realizing the original Raman spectrum pretreatment of different types of samples, the multivariate data analysis method of characteristic extraction and classification and identification is applied to realize the extraction and judgment of the spectral characteristic information of different materials.
Disclosure of Invention
The invention aims to provide a Raman spectrum multivariate data analysis method and a software system, which are applied to Raman spectrum and spectrum data set preprocessing and multivariate analysis of various organic and inorganic materials. And performing feature extraction on the sample spectrum according to the Raman spectrum data set by combining PCA and PLS-DA algorithms, and performing discriminant analysis on the sample features by combining LDA, PLS-DA, SVM and PCA-SVM algorithms.
In order to achieve the purpose, the invention provides the following scheme:
a Raman spectrum multivariate data analysis method comprises the following steps:
s1, measuring by using a Raman spectrum detection instrument to obtain original Raman spectrums and spectrum data sets of various organic and inorganic materials;
s2, preprocessing the obtained Raman spectrum data set by using a Raman spectrum multivariate data analysis software system;
s3, preprocessing the obtained Raman spectrum data set, and then performing normalization and mean value centralization processing on the Raman spectrum data;
s4, extracting Raman spectrum characteristic data by adopting a Principal Component Analysis (PCA) method or a partial least squares-discriminant analysis (PLS-DA) method, and extracting significant characteristic components in the Raman spectrum data by respectively utilizing one-way analysis of variance and cross validation;
s5, respectively establishing the features extracted in the step S4 by combining classification models, and performing classification and identification on spectral information by using the four classification models;
s6, evaluating the reliability of the classification model by using unbiased leave-one method for cross validation;
and S7, selecting the residual data for testing to obtain the accuracy, sensitivity and specificity of sample classification and the characteristic curve of the tested worker of the classification model, and evaluating the performance of the classification model.
Preferably, in step S2, the preprocessing mainly includes: spectral feature range selection, cosmic ray removal, background fluorescence signal processing based on a polynomial fitting method, and spectral smoothing processing based on a Savitzky-Golay convolution method.
Preferably, in step S3, on the basis of the preprocessing, the spectral intensity normalization, the spectral peak area normalization, the peak intensity normalization and the mean centering processing are selected according to requirements.
Preferably, the principal component analysis PCA in step S4 includes:
converting a group of linear correlation variables into linear independent variables through orthogonal transformation, reducing the dimensionality of a spectral data set, and simultaneously extracting a significant feature J in the data set; constructing a sample data set X (I multiplied by J) according to the observed sample number I and the spectral feature number J, carrying out spectral peak area normalization and mean value centralization on the sample data set, and then obtaining a covariance matrix XTX; performing singular value decomposition on the covariance matrix to obtain X ═ P delta QTWhere P is the left singular vector, Q is the right singular vector, and Δ is the diagonal matrix of singular values;
F＝PΔ，F＝PΔ＝PΔQTq is XQ, and the matrix Q gives the coefficients for calculating the linear combination of the factor scores, and is therefore also referred to as a projection matrix, and multiplying X by Q yields the projection F of the observed values on the principal component.
Preferably, the four classification models in step S5 include: the method is based on a classification model established by a linear discriminant analysis method LDA, a partial least square-discriminant analysis method PLS-DA, a support vector machine SVM and a principal component analysis combined support vector machine PCA-SVM algorithm.
Preferably, the remaining data is selected in the step S7 for testing to obtain a characteristic curve ROC of the test worker for each classification model performance index, and the raman spectrum data and biochemical difference are analyzed in combination with the steps S5-S7.
Preferably, the ROC curve is a subject working characteristic curve and can reflect the sensitivity and specificity of the spectral classification model; the ROC curve calculates a series of sensitivity and specificity by continuously changing the classification threshold value, and then is drawn into the ROC curve by taking the sensitivity as a vertical coordinate and the 1-specificity as a horizontal coordinate, and the larger the area under the curve is, the higher the prediction accuracy of the classification model is.
The invention has the beneficial effects that:
1. the Raman spectrum data set preprocessing method has a perfect Raman spectrum data set preprocessing function, can select a spectrum characteristic range of a single acquired Raman spectrum or spectrum data set, remove cosmic rays, process background fluorescence signals based on a polynomial fitting method, smooth spectrum processing based on Savitzky-Golay convolution, and select a normalization (spectrum intensity normalization, spectrum peak area normalization and peak intensity normalization) and mean value centralization processing function according to requirements;
2. the invention integrates and optimizes a plurality of Raman spectrum multivariate data analysis methods commonly used for various organic materials and inorganic materials: a principal component analysis method (PCA), a partial least squares-discriminant analysis method (PLS-DA), a linear discriminant analysis method (LDA), a Support Vector Machine (SVM), a principal component analysis combined support vector machine (PCA-SVM);
3. according to the PCA-SVM classification algorithm model, principal component analysis and a support vector machine are combined, and classification performance of the model is improved on the basis of an SVM;
4. the invention can effectively identify and distinguish the characteristics of samples including various organic and inorganic materials represented by biological tissues and cells, but not limited to the samples;
5. in the feature extraction part, PCA, PLS-DA and single-factor analysis of variance and cross validation are combined to select features with significant meaning in a data set.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a flow chart of data analysis according to the present invention;
FIG. 2 is a schematic view of a Raman spectrum data preprocessing interface according to the present invention;
FIG. 3 is a diagram illustrating the results of smoothing process after removing cosmic rays, removing background noise, and performing cosmic ray removal in the embodiment of the invention;
FIG. 4 is a diagram illustrating a result of mean centering according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a PCA-LDA model cross validation and classification summary interface in an embodiment of the present invention;
FIG. 6 is a schematic diagram of a PLS-DA model cross validation and sorting summary interface according to an embodiment of the present invention;
FIG. 7 is a diagram of an SVM model training interface according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
FIG. 1 shows a flow chart of data analysis according to the present invention;
utilize various commercial and independently developed raman spectrometer, include: the Raman spectrum and the spectrum data set of various organic and inorganic materials are obtained by measurement of a large scientific research-grade Raman spectrum detection instrument and a small portable Raman spectrum detection instrument;
preprocessing the acquired Raman spectrum data through a spectrum preprocessing interface shown in FIG. 2, wherein the spectrum preprocessing comprises spectrum characteristic range selection, cosmic ray removal, background fluorescence signal processing based on a polynomial fitting method and spectrum smoothing processing based on a Savitzky-Golay convolution method (the result and the interface are shown in FIGS. 2 and 3); on the basis of preprocessing, normalization and mean centering processing can be selected according to requirements (the result is shown in fig. 4);
the normalization processing comprises the following steps: in order to eliminate the influence of power disturbance and sample nonuniformity, the spectral intensity normalization can be selected; for the purpose of discussing quantitative information for a substance, spectral peak area normalization may be selected; peak intensity normalization may be selected in order to further highlight certain species content variations by eliminating effects due to sample and instrument variations.
The invention provides two methods for extracting the characteristics of a preprocessed spectral data set: principal Component Analysis (PCA), Partial least squares discriminant analysis (PLS-DA); and selecting any one of the methods to analyze the spectral data set, and then selecting the most significant spectral characteristic components by utilizing one-factor analysis of variance and cross validation respectively.
The main component analysis comprises the following specific steps:
converting a group of linear correlation variables into linear independent variables through orthogonal transformation, thereby reducing the dimensionality of the spectral data set and simultaneously extracting remarkable characteristics in the data set; the sample data set is X (I × J), I is the number of observation samples, and J is the number of spectral features.
Firstly, carrying out spectrum peak area normalization and mean value centralization treatment, and then obtaining a covariance matrix XTX; performing singular value decomposition on the covariance matrix to obtain X ═ P delta QTWhere P is the left singular vector, Q is the right singular vector, and Δ is the diagonal matrix of singular values.
F＝PΔ，F＝PΔ＝PΔQTQ is XQ, the matrix Q gives the coefficients for calculating the linear combination of the factor scores and is therefore also called the projection matrix (or loading matrix), and multiplying X by Q yields the projection F of the observed values on the principal component (F is also called the score matrix).
The linear discriminant analysis LDA comprises the following steps:
(1) the convention data set comprises two types of samples, and an interspecies divergence matrix S is calculatedbAnd mu1；
Sb＝(μ0-μ1)(μ0-μ1)Tu0
Projecting the data onto a straight line omega, the centers of the two types of samples are on the straight lineThe projections are respectively omegaTμ0And ωTμ1；
(2) Calculating the similar internal divergence matrix S of the samplew
(3) Calculating an inter-class divergence matrix SbIn-class divergence matrix S of the same kind as the samplewGeneralized Rayleigh entropy of
Solving a projection direction omega;
(4) projection line, i.e. y ═ ωTx；
(5) And projecting the new unknown sample to the straight line, and classifying the class of the point according to the distance from the projected point to the centers of the two types of samples.
FIG. 5 is a cross-validation and sort summary interface of the PCA-LDA model according to an embodiment of the present invention.
The least square discrimination method includes:
(1) carrying out mean value centralization processing on the data;
(2) calculating the predicted response value of each sample according to the least square regression;
(3) and calculating the posterior probability of the sample belonging to each category according to a probability density function and a Bayes formula, such as time A and event B:
P(A|B)＝P(B|A)*P(A)/P(B)
(4) the class with the highest probability is selected as the predictive label.
FIG. 6 shows a PLS-DA model cross validation and sort summary interface;
the support vector machine comprises the following steps:
(1) appointment hyper-plane omegaTx + b ═ y; where ω is the normal vector and b is the displacement.
(2) Calculating the distance d from the point to the hyperplane y;
(3) maximizing the classification interval;
s.t.yi(wT·Φ(xi)+b)≥1,i＝1,2,…,n
wherein, phi (x)i) Is a feature space transformation function, i.e. a mapping function, and s.t. is a constraint.
(4) Introducing slack variables allows some data to be misclassified, preventing overfitting;
the constraint s.t. is:
yi(w·xi+b)≥1-ξ,i＝1,2,…,n
ξi≥0,i＝1,2,…,n
fig. 7 shows an SVM training model interface.
The invention uses Linear Discriminant Analysis (LDA), partial least square-discriminant analysis (PLS-DA), Support Vector Machine (SVM), Principal component analysis (Principal component analysis combined Support vector machine (PCA-SVM)) algorithm to establish four classification models, and extracts the features through the four models respectively.
And (3) cross validation and evaluation of the reliability of each classification model by using an unbiased leave-one method, so that an overfitting phenomenon is prevented.
Taking the total amount of the samples as N in the above steps, and selecting NtThe data are used as training set, then Nts＝N-NtThe number of the taken test samples is the accuracy, the sensitivity and the specificity of the obtained sample classification and the characteristics of the tested workers of the modelThe curve is used for evaluating the performance of Raman spectrum multivariate data analysis methods on Raman spectrum identification of samples (particularly biological tissue samples).
The above-described embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solutions of the present invention can be made by those skilled in the art without departing from the spirit of the present invention, and the technical solutions of the present invention are within the scope of the present invention defined by the claims.
Claims (7)
1. A Raman spectrum multivariate data analysis method is characterized by comprising the following steps:
s1, measuring by using a Raman spectrum detection instrument to obtain original Raman spectrums and spectrum data sets of various organic and inorganic materials;
s2, preprocessing the obtained Raman spectrum data set by using a Raman spectrum multivariate data analysis software system;
s3, preprocessing the obtained Raman spectrum data set, and then performing normalization and mean value centralization processing on the Raman spectrum data;
s4, extracting Raman spectrum characteristic data by adopting a Principal Component Analysis (PCA) method or a partial least squares-discriminant analysis (PLS-DA) method, and extracting significant characteristic components in the Raman spectrum data by respectively utilizing one-way analysis of variance and cross validation;
s5, respectively establishing the features extracted in the step S4 by combining classification models, and performing classification and identification on spectral information by using the four classification models;
s6, evaluating the reliability of the classification model by using unbiased leave-one method for cross validation;
and S7, selecting the residual data for testing to obtain the accuracy, sensitivity and specificity of sample classification and the characteristic curve of the tested worker of the classification model, and evaluating the performance of the classification model.
2. The raman spectroscopy multivariate data analysis method according to claim 1, wherein the preprocessing mainly comprises, in step S2: spectral feature range selection, cosmic ray removal, background fluorescence signal processing based on a polynomial fitting method, and spectral smoothing processing based on a Savitzky-Golay convolution method.
3. The raman spectroscopy multivariate data analysis method according to claim 1, wherein in step S3, spectral intensity normalization, spectral peak area normalization, peak intensity normalization and mean centering are selected as required on the basis of preprocessing.
4. The raman spectroscopy multivariate data analysis method according to claim 1, wherein the Principal Component Analysis (PCA) in step S4 is performed by:
converting a group of linear correlation variables into linear independent variables through orthogonal transformation, reducing the dimensionality of a spectral data set, and simultaneously extracting a significant feature J in the data set; constructing a sample data set X (I multiplied by J) according to the observed sample number I and the spectral feature number J, carrying out spectral peak area normalization and mean value centralization on the sample data set, and then obtaining a covariance matrix XTX; performing singular value decomposition on the covariance matrix to obtain X ═ P delta QTWhere P is the left singular vector, Q is the right singular vector, and Δ is the diagonal matrix of singular values;
F＝PΔ，F＝PΔ＝PΔQTq is XQ, and the matrix Q gives the coefficients for calculating the linear combination of the factor scores, and is therefore also referred to as a projection matrix, and multiplying X by Q yields the projection F of the observed values on the principal component.
5. The raman spectroscopy multivariate data analysis method according to claim 1, wherein the four classification models in step S5 comprise: the method is based on a classification model established by a linear discriminant analysis method LDA, a partial least square-discriminant analysis method PLS-DA, a support vector machine SVM and a principal component analysis combined support vector machine PCA-SVM algorithm.
6. The method of claim 1, wherein the residual data is selected and tested in step S7 to obtain a characteristic curve ROC of the tester for each classification model performance index, and the raman spectrum data and biochemical differences are analyzed in combination with steps S5-S7.
7. The method for multivariate data analysis by Raman spectroscopy according to claim 6, wherein the ROC curve is a subject operating characteristic curve capable of reflecting spectral classification model sensitivity and specificity; the ROC curve calculates a series of sensitivity and specificity by continuously changing the classification threshold value, and then is drawn into the ROC curve by taking the sensitivity as a vertical coordinate and the 1-specificity as a horizontal coordinate, and the larger the area under the curve is, the higher the prediction accuracy of the classification model is.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011482853.5A CN112712108B (en) | 2020-12-16 | 2020-12-16 | Raman spectrum multivariate data analysis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011482853.5A CN112712108B (en) | 2020-12-16 | 2020-12-16 | Raman spectrum multivariate data analysis method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112712108A true CN112712108A (en) | 2021-04-27 |
| CN112712108B CN112712108B (en) | 2023-08-18 |
Family
ID=75542146
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011482853.5A Active CN112712108B (en) | 2020-12-16 | 2020-12-16 | Raman spectrum multivariate data analysis method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112712108B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113408616A (en) * | 2021-06-18 | 2021-09-17 | 北京航空航天大学 | Spectrum classification method based on PCA-UVE-ELM |
| CN113947117A (en) * | 2021-10-08 | 2022-01-18 | 南京航空航天大学 | A CNN transfer learning method based on polynomial reconstruction algorithm for Raman spectroscopy |
| CN113989578A (en) * | 2021-12-27 | 2022-01-28 | 季华实验室 | Method, system, terminal device and medium for analyzing peak position of Raman spectrum |
| CN114295600A (en) * | 2021-12-30 | 2022-04-08 | 西北大学 | An Improved Raman Spectroscopy Multivariate Data Analysis and Imaging Method |
| CN115184336A (en) * | 2022-07-15 | 2022-10-14 | 新疆维吾尔自治区人民医院 | A method for identification of Sjögren's syndrome and interstitial lung disease based on serum Raman spectroscopy |
| CN116559143A (en) * | 2023-05-15 | 2023-08-08 | 西北大学 | Method and system for analyzing composite Raman spectrum data of glucose component in blood |
| CN116933056A (en) * | 2023-07-24 | 2023-10-24 | 哈尔滨工业大学 | Method and system for determining the peak area of characteristic peaks of Raman spectrum without subtracting Raman background |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011174906A (en) * | 2010-02-25 | 2011-09-08 | Olympus Corp | Vibration spectrum analysis method |
| WO2017019988A1 (en) * | 2015-07-30 | 2017-02-02 | The Research Foundation For The State University Of New York | Gender and race identification from body fluid traces using spectroscopic analysis |
| CN106680241A (en) * | 2017-01-13 | 2017-05-17 | 北京化工大学 | Novel spectrum multi-analysis classification and identification method and application thereof |
| CN108802002A (en) * | 2018-05-08 | 2018-11-13 | 华南农业大学 | A kind of quick nondestructive differentiates the silkworm seed Raman spectrum model building method of termination of diapause |
| CN108802000A (en) * | 2018-03-16 | 2018-11-13 | 上海交通大学 | A kind of lossless quick cholecalciferol-cholesterol content quantitative method based on the full spectrum analysis of Raman |
-
2020
- 2020-12-16 CN CN202011482853.5A patent/CN112712108B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011174906A (en) * | 2010-02-25 | 2011-09-08 | Olympus Corp | Vibration spectrum analysis method |
| WO2017019988A1 (en) * | 2015-07-30 | 2017-02-02 | The Research Foundation For The State University Of New York | Gender and race identification from body fluid traces using spectroscopic analysis |
| CN106680241A (en) * | 2017-01-13 | 2017-05-17 | 北京化工大学 | Novel spectrum multi-analysis classification and identification method and application thereof |
| CN108802000A (en) * | 2018-03-16 | 2018-11-13 | 上海交通大学 | A kind of lossless quick cholecalciferol-cholesterol content quantitative method based on the full spectrum analysis of Raman |
| CN108802002A (en) * | 2018-05-08 | 2018-11-13 | 华南农业大学 | A kind of quick nondestructive differentiates the silkworm seed Raman spectrum model building method of termination of diapause |
Non-Patent Citations (2)
| Title |
|---|
| 王宁;王驰;卞海溢;王钧;王鹏;白鹏利;尹焕才;田玉冰;高静;: "应用Hilbert变换提取拉曼光谱相位信息进行血液识别分类方法的研究", 光谱学与光谱分析, no. 08 * |
| 袁玉峰;陶站华;刘军贤;: "化学计量学结合拉曼光谱在生物材料检测中的应用", 光谱实验室, no. 06 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113408616A (en) * | 2021-06-18 | 2021-09-17 | 北京航空航天大学 | Spectrum classification method based on PCA-UVE-ELM |
| CN113408616B (en) * | 2021-06-18 | 2024-03-26 | 北京航空航天大学 | Spectral classification method based on PCA-UVE-ELM |
| CN113947117A (en) * | 2021-10-08 | 2022-01-18 | 南京航空航天大学 | A CNN transfer learning method based on polynomial reconstruction algorithm for Raman spectroscopy |
| CN113989578A (en) * | 2021-12-27 | 2022-01-28 | 季华实验室 | Method, system, terminal device and medium for analyzing peak position of Raman spectrum |
| CN113989578B (en) * | 2021-12-27 | 2022-04-26 | 季华实验室 | Raman spectroscopy peak position analysis method, system, terminal equipment and medium |
| CN114295600A (en) * | 2021-12-30 | 2022-04-08 | 西北大学 | An Improved Raman Spectroscopy Multivariate Data Analysis and Imaging Method |
| CN115184336A (en) * | 2022-07-15 | 2022-10-14 | 新疆维吾尔自治区人民医院 | A method for identification of Sjögren's syndrome and interstitial lung disease based on serum Raman spectroscopy |
| CN115184336B (en) * | 2022-07-15 | 2024-03-15 | 新疆维吾尔自治区人民医院 | Method for identifying Sjogren syndrome and interstitial lung disease based on serum Raman spectrum |
| CN116559143A (en) * | 2023-05-15 | 2023-08-08 | 西北大学 | Method and system for analyzing composite Raman spectrum data of glucose component in blood |
| CN116933056A (en) * | 2023-07-24 | 2023-10-24 | 哈尔滨工业大学 | Method and system for determining the peak area of characteristic peaks of Raman spectrum without subtracting Raman background |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112712108B (en) | 2023-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112712108B (en) | Raman spectrum multivariate data analysis method | |
| Sharma et al. | Trends of chemometrics in bloodstain investigations | |
| CN113177919B (en) | Lithology classification and principal component element content detection method combining LIBS and deep learning | |
| JP2022525427A (en) | Automatic boundary detection in mass spectrometry data | |
| CN114184599B (en) | Single-cell Raman spectrum acquisition number estimation method, data processing method and device | |
| Wang et al. | Mid-level data fusion of Raman spectroscopy and laser-induced breakdown spectroscopy: Improving ores identification accuracy | |
| Li et al. | Raman spectrum matching with contrastive representation learning | |
| US7991223B2 (en) | Method for training of supervised prototype neural gas networks and their use in mass spectrometry | |
| Lotter et al. | Identifying plastics with photoluminescence spectroscopy and machine learning | |
| Huffman et al. | Laser-induced breakdown spectroscopy spectral feature selection to enhance classification capabilities: A t-test filter approach | |
| Coic et al. | Assessment of essential information in the fourier domain to accelerate raman hyperspectral microimaging | |
| Huang et al. | The application of wavelet transform of Raman spectra to facilitate transfer learning for gasoline detection and classification | |
| Cai et al. | Deep metric learning framework combined with Gramian angular difference field image generation for Raman spectra classification based on a handheld Raman spectrometer | |
| Chen et al. | Authentication and inference of seal stamps on Chinese traditional painting by using multivariate classification and near-infrared spectroscopy | |
| Koyun et al. | RamanFormer: A transformer-based quantification approach for Raman mixture components | |
| Burlacu et al. | Convolutional Neural Network detecting synthetic cannabinoids | |
| Gholami et al. | A statistical modeling approach for tumor-type identification in surgical neuropathology using tissue mass spectrometry imaging | |
| US20240011910A1 (en) | Methods and systems for raman spectra-based identification of chemical compounds | |
| CN106383103A (en) | Time-resolved oil fluorescence identification method based on main component analysis | |
| JP6280910B2 (en) | Method for measuring the performance of a spectroscopic system | |
| Vyas et al. | Phenotypic profiling based on body fluid traces discovered at the scene of crime: Raman spectroscopy of urine stains for race differentiation | |
| Campanella et al. | Data Fusion: LIBS+ Raman | |
| Li et al. | Forensic Characterization of Mobile Phone Cases Using Infrared Spectroscopy with Deep Learning Methods | |
| Ratle et al. | Pattern analysis in illicit heroin seizures: a novel application of machine learning algorithms. | |
| Grissa et al. | A hybrid data mining approach for the identification of biomarkers in metabolomic data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |