CN104345045A - Chemical pattern recognition and near infrared spectrum-based similar medicinal material identification method - Google Patents
Chemical pattern recognition and near infrared spectrum-based similar medicinal material identification method Download PDFInfo
- Publication number
- CN104345045A CN104345045A CN201410616144.XA CN201410616144A CN104345045A CN 104345045 A CN104345045 A CN 104345045A CN 201410616144 A CN201410616144 A CN 201410616144A CN 104345045 A CN104345045 A CN 104345045A
- Authority
- CN
- China
- Prior art keywords
- near infrared
- infrared spectrum
- medicinal material
- similar
- pattern recognition
- 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.)
- Pending
Links
Abstract
The invention relates to a method for classifying similar medicinal materials through a near infrared spectrum. The method comprises the following specific steps of collecting the similar medicinal materials with certain number as modeling samples, and after crushing the traditional Chinese medicinal materials, testing the near infrared spectrum; testing the classifying ability of the samples by different chemical pattern recognition methods, including hierarchical cluster analysis, partial least squares discriminant analysis and a support vector machine method. Results show that 100 percent of prediction accuracy can be achieved by the support vector machine method and partial least squares discriminant analysis. Medicines are not damaged by the method, pollution to the environment is also not caused, and the method is an environment friendly method. Therefore, the similar traditional Chinese medicinal materials can be quickly and accurately identified without being damaged by the method that the near infrared spectrum is combined with a proper chemical pattern recognition method.
Description
Technical field
The invention belongs to the nondestructive analysis technology in analytical chemistry field, relate to analytical approach and the Chemical Pattern Recognition method of near infrared spectrum.
Background technology
Chinese medicine has long history in China, and it contains China's culture of several thousand.In recent years, the development of Chinese medicine obtains the attention of people gradually, utilizes the conditioning of Chinese medicine, and we can solve the indeterminable problem of some Western medicine, and in addition, Chinese medicine is also smaller to the actual bodily harm of people.But some illegal businessman, in order to obtain interests, utilize the similarity of medicinal material to process some rare medicinal herbs, the drug effect of medicinal material is declined, and adverse influence may be produced to the health of people, finally consumer is caused damage.Therefore, the discrimination method of similar Chinese medicine is China's traditional Chinese medicine industry urgent problem.People gather medicinal material frequently by the method for eye recognition, and this is suitable for really to most of medicinal material.But Chinese medicinal material is of a great variety, and the outward appearance of some medicinal material is extremely similar, with the naked eye cannot differentiate at all.In recent years, various modern instrument analytical method is used to Chinese traditional medicine identification, such as, and the technology such as near infrared spectrum, Raman spectrum and imaging, Magnetic resonance imaging and ultrasonic imaging.
The near-infrared spectrum technique that development in recent years is got up, it has fast, accurately, the feature such as harmless and informative, instead of the Analysis of Chinese Traditional Medicine method that some are traditional gradually, in the research of Chinese medicine, achieve very important effect.But because near infrared spectrum district signal is more weak, overlapping serious, making us when using near infrared spectrum, just must can carry out qualitative and quantitative analysis to complex systems such as Chinese medicines by Chemical Measurement.
Chemical Pattern Recognition utilizes the instruments such as mathematics, statistics, signal transacting, and chemically metric data infers the essential attribute of material type, and then a special kind of skill identifying material and sort out.The method is as an important research content of Chemical Measurement, the fields such as agricultural, medicine, food analysis, oil are widely applied to, in the places of origin of raw materials, true and false discriminating, production quality control and analysis, the aspects such as Quality Identification have played important effect.Chemical Pattern Recognition method includes the mode identification method of supervision and unsupervised mode identification method, and the former mainly contains principal component analysis (PCA), hierarchial-cluster analysis, and the latter mainly contains the methods such as offset minimum binary-differentiation, support vector machine, artificial neural network.Chemical Pattern Recognition method is incorporated in the classification of similar Chinese crude drug, will the quick and precisely classification realizing similar medicinal material be expected to.
To sum up, differentiate similar Chinese crude drug, in case adulterate, even adulterated is current problem demanding prompt solution.This experiment by the analytical approach of near infrared spectrum in conjunction with Chemical Pattern Recognition, thus realizes the discriminating to similar medicine.
Summary of the invention
The object of the invention is for above-mentioned Problems existing, provide fast a kind of, accurately, the method that the harmless Identification chinese herbs medicine material waited is similar.
Comprise the following steps for realizing technical scheme provided by the present invention:
1) sample collection: with the very similar Chinese crude drug of several outward appearance for research object, collect the multiple samples from different manufacturers, by solid pulverizing medicinal materials, sieve, numbering.
2) near infrared spectrum data collection: determine the parameter that near infrared spectrum is tested, chooses the medicinal powder sample collection near-infrared diffuse reflection spectrum signal of suitable order number.
3) Chemical Pattern Recognition Method Modeling: use Chemical Pattern Recognition method to carry out taxonomic history to similar medicinal material.Attempt unsupervised systemic clustering respectively, have the offset minimum binary-discriminatory analysis of linear classification method and the Nonlinear Classification method support vector regression of supervision.System Cluster Analysis adopts system clustering tree to show cluster result.First offset minimum binary-discriminatory analysis and support vector machine need sample to be divided into training set and forecast set, by forecast set, classifying quality predicts that accuracy (suc-pred) judges.Offset minimum binary-discriminatory analysis needs certainty factor number parameter.
4) according to the classifying quality of often kind of method, select best sorting technique and modeling is carried out to this similar medicinal material, prediction.
Adopt near infrared method to test sample in the present invention, sample, without the need to the loaded down with trivial details pre-treatment step such as separation and Extraction, not only can not destroy sample, and test speed quickly.Preserved on computers by the spectroscopic data recorded, utilize the program finished to carry out computing, not only computing velocity is very fast, and achieves the robotization of cluster, does not need to carry out manpower comparing pair to experiment again.Therefore, the present invention is to provide a kind of fast, accurately, the method for harmless etc. the similar Chinese medicine of qualification.
Accompanying drawing explanation
Fig. 1: the near infrared light spectrogram of the root of herbaceous peony, the root of Dahurain angelica, the root of kudzu vine, Poria cocos, the root bark of tree peony five kinds of Chinese medicine 50 samples
The Hierarchical Clustering result figure of Fig. 2: 50 traditional Chinese medicine samples
Specific embodiment
For better understanding the present invention, below in conjunction with embodiment the present invention done and describe in detail further, but the scope of protection of present invention being not limited to the scope that embodiment represents.
Embodiment:
1) sample collection: the sample collecting the root of herbaceous peony, the root of Dahurain angelica, the root of kudzu vine, Poria cocos, the root bark of tree peony five kinds of similar medicinal materials, often kind of medicinal material is bought (the raw large pharmacy of Tianjin good fortune unit, the large pharmacy of Tianjin, Tianjin Yongxin, Tianjin are with one hundred pharmacy, the large pharmacy of up time moral, Tianjin, pharmacy, Tianjin century three pool) from five different pharmacies respectively, the same medicinal material of each pharmacy is divided into two parts, altogether collects five kinds of similar medicinal materials totally 50 increment product.By 50 parts of sample comminution, cross 60 mesh sieves, 120 mesh sieves successively, retain between 60-120 order, i.e. the sample of particle size range 0.118 ~ 0.1154mm, successively in the numbered plastic bottle of tape loaded.
2) near infrared spectrum data collection: sample is put into quartzy bottle to bottle height degree half about, upper and lower oscillator quartz bottle 25 times, uses the near infrared spectrum of integrating sphere diffuse reflectance accessory collected specimens.Gather spectral range 4000 ~ 10000cm
-1, scanning times 64 times, resolution 8cm
-1, each sample collection 3 spectrum, get its averaged spectrum for data processing.Fig. 1 is the near infrared light spectrogram of 50 traditional Chinese medicine samples.As can be seen from the figure, the spectrum atlas peak of Chinese medicine near infrared sample is not obvious, and background is large, and spectra overlapping, directly cannot be distinguished five kinds of Chinese medicines by spectrum itself together.
3) Chemical Pattern Recognition Method Modeling:
1. hierarchial-cluster analysis modeling: first adopt principal component analysis (PCA) to carry out dimensionality reduction to data, by parameter optimization, inter-object distance adopts European cluster, and between class distance adopts mean distance, carry out cluster analysis to the near infrared spectrum of 50 samples, cluster result as shown in Figure 2.As can be seen from the figure, the first kind (root of herbaceous peony) and the 4th class (Poria cocos) medicinal material can better must distinguish with other medicinal material, but second (root of Dahurain angelica), three (roots of kudzu vine), five classes (root bark of tree peony) medicinal material do not distinguish by hierarchial-cluster analysis.
2. offset minimum binary method of discrimination modeling: adopt KS group technology, takes out 27 samples and carries out modeling, remain 13 samples and predict from 50 samples.Determine to predict the outcome accuracy with the change because of subnumber, determine offset minimum binary-discriminatory analysis because of subnumber be 9.The accuracy that predicts the outcome that result display offset minimum binary-method of discrimination carries out analyzing is 100%.
3. model construction of SVM: adopt KS group technology, takes out 27 samples and carries out modeling, remain 13 samples and predict from 50 samples.Adopt linear kernel function to set up supporting vector machine model, during result display employing 10 support vectors, the prediction accuracy of 100% can be reached to forecast set.
4) by the comparison of above-mentioned modeling method, adopt unsupervised systemic clustering cannot entirely true differentiation five kinds of Chinese crude drugs, and adopt when having offset minimum binary-discriminatory analysis or the support vector machine of supervision, the prediction accuracy of 100% can be reached.Therefore, near infrared spectrum can be classified to similar Chinese medicine in conjunction with offset minimum binary-discriminatory analysis or support vector machine.
Claims (3)
1. fast, accurately, the discriminating of harmless, free of contamination similar medicinal material, it is characterized in that: it is the infrared spectroscopy signals adopting near-infrared spectral measurement sample, carries out similar discriminating afterwards by Chemical Pattern Recognition method.
2. the discriminating of quick, accurate, harmless similar medicinal material according to claim 1, it is characterized in that: first collect the sample of some for modeling, adopt hierarchial-cluster analysis more afterwards, support vector machine, least square method-diagnostic method is analyzed, and result determines a kind of best modeling method being applicable to this system by experiment.
3. similar medicinal material according to claim 1 is quick, and accurately, harmless discrimination method, is characterized in that: do not limit the species number of medicine, no matter be two kinds or two or morely all can differentiate.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410616144.XA CN104345045A (en) | 2014-11-04 | 2014-11-04 | Chemical pattern recognition and near infrared spectrum-based similar medicinal material identification method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410616144.XA CN104345045A (en) | 2014-11-04 | 2014-11-04 | Chemical pattern recognition and near infrared spectrum-based similar medicinal material identification method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN104345045A true CN104345045A (en) | 2015-02-11 |
Family
ID=52501110
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410616144.XA Pending CN104345045A (en) | 2014-11-04 | 2014-11-04 | Chemical pattern recognition and near infrared spectrum-based similar medicinal material identification method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104345045A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105223161A (en) * | 2015-09-29 | 2016-01-06 | 中国林业科学研究院林产化学工业研究所 | A kind of method differentiating slurrying material based near infrared spectrum |
CN105954222A (en) * | 2016-07-18 | 2016-09-21 | 成都中医药大学 | Identification method for Brag-zhun and faecal pellet substitutes |
CN106990214A (en) * | 2017-05-08 | 2017-07-28 | 云南民族大学 | A kind of method for evaluating Chinese medicine quality |
CN109507123A (en) * | 2018-11-22 | 2019-03-22 | 天津工业大学 | A kind of discrimination method of Radix Angelicae Sinensis based on UV-vis DRS spectrum and Chemical Pattern Recognition and its similar product |
CN110160987A (en) * | 2019-07-03 | 2019-08-23 | 温州大学 | A kind of method for quickly detecting quality of Radix Bupleuri |
CN113917003A (en) * | 2021-08-30 | 2022-01-11 | 浙江工业大学 | Method for detecting content of imperatorin in angelica dahurica decoction pieces |
-
2014
- 2014-11-04 CN CN201410616144.XA patent/CN104345045A/en active Pending
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105223161A (en) * | 2015-09-29 | 2016-01-06 | 中国林业科学研究院林产化学工业研究所 | A kind of method differentiating slurrying material based near infrared spectrum |
CN105954222A (en) * | 2016-07-18 | 2016-09-21 | 成都中医药大学 | Identification method for Brag-zhun and faecal pellet substitutes |
CN106990214A (en) * | 2017-05-08 | 2017-07-28 | 云南民族大学 | A kind of method for evaluating Chinese medicine quality |
CN106990214B (en) * | 2017-05-08 | 2019-11-15 | 云南民族大学 | A method of evaluation Chinese medicine quality |
CN109507123A (en) * | 2018-11-22 | 2019-03-22 | 天津工业大学 | A kind of discrimination method of Radix Angelicae Sinensis based on UV-vis DRS spectrum and Chemical Pattern Recognition and its similar product |
CN110160987A (en) * | 2019-07-03 | 2019-08-23 | 温州大学 | A kind of method for quickly detecting quality of Radix Bupleuri |
CN113917003A (en) * | 2021-08-30 | 2022-01-11 | 浙江工业大学 | Method for detecting content of imperatorin in angelica dahurica decoction pieces |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104345045A (en) | Chemical pattern recognition and near infrared spectrum-based similar medicinal material identification method | |
CN104215623B (en) | Laser Raman spectroscopy intelligence discrimination method and system towards conglomerate detection | |
CN107677647A (en) | Chinese medicine place of production discrimination method based on principal component analysis and BP neural network | |
CN102179375B (en) | Nondestructive detecting and screening method based on near-infrared for crop single-grain components | |
CN102778442B (en) | Method for rapidly identifying type of balsam material liquid for cigarette | |
CN103033486B (en) | Method for near infrared spectrum monitoring of quality of pericarpium citri reticulatae and citrus chachiensis hortorum medicinal materials | |
CN103411906B (en) | The near infrared spectrum qualitative identification method of pearl powder and oyster shell whiting | |
CN103439266A (en) | Identification and analysis system for jewelry jade and rock minerals | |
CN102288551A (en) | Method for distinguishing traditional Chinese medicinal material by utilizing hyperspectral data | |
CN106124445A (en) | A kind of quick, Undamaged determination genetically engineered soybean method | |
CN104376325A (en) | Method for building near-infrared qualitative analysis model | |
CN104062262A (en) | Crop seed variety authenticity identification method based on near infrared spectrum | |
CN103499552A (en) | Fast and intelligent waste plastic sorting method | |
CN102243170A (en) | Method for identifying production place of dwarf lilyturf root by using near infrared spectrum technology | |
CN107132194A (en) | A kind of pseudo-ginseng and its adulterant discrimination method based on uv-vis spectra and Chemical Pattern Recognition | |
CN108593592A (en) | A kind of tuber of pinellia based on near-infrared spectrum technique mixes pseudo- discrimination method | |
CN108760677A (en) | A kind of rhizoma pinellinae praeparata based on near-infrared spectrum technique mixes pseudo- discrimination method | |
CN101620178B (en) | Method for quickly detecting additive chemical component in Chinese patent medicine, health-care food or food based on near-infrared spectrum technique | |
CN104020128A (en) | Method for rapidly identifying propolis source | |
CN104297206A (en) | Quick liquid milk brand identifying method | |
CN104132968A (en) | Identification method of rice geographical indications and application of identification method of rice geographical indications | |
Cui et al. | Identification of maize seed varieties based on near infrared reflectance spectroscopy and chemometrics | |
CN105223140A (en) | The method for quickly identifying of homology material | |
CN103076300B (en) | Method for judging and analyzing traditional Chinese medicine resource fingerprint information by specific mode identification model | |
CN111426657B (en) | Identification comparison method of three-dimensional fluorescence spectrogram of soluble organic matter |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150211 |
|
WD01 | Invention patent application deemed withdrawn after publication |