Traditional Chinese medicine fossil fragments identification method based on combination of micro-area element analysis and isotope fingerprints

Abstract

The invention relates to a traditional Chinese medicine dragon bone identification method based on combination of microcell element analysis and isotope fingerprints. The method comprises the following steps: mechanically crushing a to-be-detected dragon bone sample; performing elemental composition analysis on different micro-areas on the surface of the sample by adopting an EDS (Energy Dispersion Spectrometer) equipped on a scanning electron microscope to obtain atomic percentage data of a preset element; stable isotope detection; carrying out standardization processing on the obtained element composition data and the obtained isotope ratio data, and carrying out feature extraction by adopting a sparse partial least square discriminant analysis method; and establishing a three-dimensional discrimination model based on the first three principal components extracted by sparse partial least square discriminant analysis, and discriminating the keel category through the clustering characteristics of the sample points in the three-dimensional space. Compared with the prior art, the method can be used for simultaneously distinguishing five types of real keels, modern bones, real keels treated by calcining or chemical means, modern bones treated by calcining or chemical means and mineral and petrochemical real keels, and is good in repeatability and reliable in distinguishing result.

Description

Technical Field

[0001] The present invention relates to the technical field of the authenticity identification of traditional Chinese medicine keel, and particularly to a method for identifying traditional Chinese medicine keel based on the combination of micro-area element analysis and isotope fingerprinting. Background Art

[0002] The main component of keel is calcium hydroxyphosphate Ca 10 (PO 4 ) 6 (OH) 2 (hydroxyapatite), and also contains a small amount of calcium carbonate and elements such as aluminum, iron, manganese, magnesium, and strontium.

[0005] Infrared spectroscopy (such as CN201911015073.7) can only identify the lattice vibration peaks of hydroxyapatite (960, 1030 cm-1), and cannot distinguish the trace element fingerprints related to the geological age; ultraviolet fluorescence method (such as CN202210547411.7) relies on the fluorescence effect generated by the residual organic matter, but after the fossil undergoes mineralization, the collagen is completely degraded, resulting in the overlap of the fluorescence intensities of the genuine and fake samples (the luminous intensity at a wavelength of 350 - 400 nm is <5 cd / m 2 ). More seriously, the calcination treatment (600 - 800 °C) will destroy the original element distribution of the bone, causing the Ca / P atomic ratio to distort from 1.67 ± 0.05 to 1.72 - 1.85, while chemical bleaching (H 2 O 2 treatment) can cause the loss of Fe and Mn elements up to 40 - 60%, and the identification accuracy rate of the existing technology for such deep fake samples is less than 70%. In addition, the traditional method lacks the combined analysis of geological age markers such as Sr isotope ( 87 Sr / 86 Sr ratio) and δ18O (+18‰ to +22‰), and it is difficult to cope with the confusion problem between regional fossils and artificially synthesized hydroxyapatite (the market inspection data shows that the misjudgment rate is as high as 31.2%). The above technical defects directly lead to the detection rate of genuine keel in market supervision being less than 45%, seriously threatening the safety of clinical medication and the inheritance and development of traditional Chinese medicine. Summary of the Invention

[0006] The purpose of the present invention is to overcome the above-mentioned defects existing in the prior art, and provide a method for identifying traditional Chinese medicine keel based on the combination of micro-area element analysis and isotope fingerprinting, which can simultaneously distinguish five categories including genuine keel, modern bones, genuine keel treated by calcination or chemical means, modern bones treated by calcination or chemical means, and mineralized genuine keel, with good repeatability and reliable discrimination results.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] The present invention provides a method for identifying Chinese herbal medicine keel based on the combination of micro-area element analysis and isotope fingerprinting, comprising the following steps:

[0009] S1. Sample pretreatment: Mechanically crush the keel sample to be tested to obtain a uniform powder with a particle size less than 100 μm.

[0010] S2. Micro-area element detection: Analyze the element composition of at least 3 different micro-areas on the sample surface using an EDS energy spectrometer equipped with a scanning electron microscope to obtain the atomic percentage data of the preset elements.

[0011] S3. Stable isotope detection: Measure the stable isotope ratios of δ 13 C, δ 2 H, δ 18 O, δ 15 N in the sample through an isotope mass spectrometer.

[0012] S4. Data fusion analysis: Standardize the element composition data obtained in S2 and the isotope ratio data obtained in S3, and use the sparse partial least squares discriminant analysis method for feature extraction.

[0013] S5. Three-dimensional modeling and discrimination: Establish a three-dimensional discrimination model based on the first three principal components extracted by the sparse partial least squares discriminant analysis, and identify the keel category through the clustering characteristics of the sample points in the three-dimensional space.

[0014] Further, in S2, the working conditions of the EDS energy spectrum analysis are as follows:

[0015] The acceleration voltage is 15 - 20 kV, the beam current is 10 - 15 nA, and the acquisition time for each micro-area is not less than 120 seconds.

[0016] Further, in S2, the preset elements include Ca, P, Al, Fe, Mn, Mg, Sr.

[0017] Further, in S3, the parameter control for the isotope ratio detection includes the following:

[0018] δ 13 C detection uses an elemental analysis-isotope mass spectrometry combined system, and the reference material is VPDB.

[0019] δ 2 H and δ 18 O detection use a high-temperature pyrolysis-isotope mass spectrometry combined system, and the reference materials are VSMOW and SLAP respectively.

[0020] δ 15 N detection uses an elemental analysis-isotope mass spectrometry combined system, and the reference material is atmospheric nitrogen.

[0021] Furthermore, in S4, the standardization processing includes normalization processing of elemental composition data and Z-score standardization processing of isotope ratios.

[0022] Furthermore, in S5, the specific steps include:

[0023] The first three principal components were extracted based on sparse partial least squares discriminant analysis, and a three-dimensional discriminant space including the distribution characteristics of calcium and phosphorus elements and isotopic fingerprints was constructed;

[0024] By setting the coordinate thresholds of the true dragon fossil area and the modern skeleton area, five types of identification rules were constructed by combining the spatial density distribution of sample points and Mahalanobis distance;

[0025] A density clustering algorithm is used to divide the transition zone of calcined samples and the clustering area of ​​mineralized samples. The spatial position relationship between sample points and standard categories is dynamically displayed through an interactive three-dimensional model. When a sample point falls into the preset confidence ellipse range, the specific category is determined according to the coordinate characteristics of the area to which it belongs.

[0026] Furthermore, the above specific categories include true dragon bones, modern bones, true dragon bones treated by calcination or chemical means, modern bones treated by calcination or chemical means, and mineralized true dragon bones.

[0027] Furthermore, the density clustering algorithm adopts an improved DBSCAN algorithm, and its parameters are set as follows:

[0028] Neighborhood radius ε = 0.5 × (PC1 standard deviation + PC2 standard deviation + PC3 standard deviation) / 3,

[0029] The above PC1 standard deviation, PC2 standard deviation, and PC3 standard deviation are calculated by each principal component.

[0030] The minimum number of neighborhood points MinPts = 5×log(N), where N is the total number of samples;

[0031] A secondary iterative analysis was performed on the discrete points outside the confidence ellipse to identify the transition zone distribution characteristics of the calcined or chemically treated samples.

[0032] Furthermore, in the interactive three-dimensional model:

[0033] The true dragon fossil area is marked with 480nm wavelength blue (RGB: 0,71,171) in the HSV color system, the modern bone area is marked with 620nm wavelength red (RGB: 203,32,39), and the mineralized true dragon bone area is dynamically rendered with a saturation gradient color scale.

[0034] Furthermore, S5 also includes the model iterative optimization step:

[0035] When the newly added sample size reaches 20% of the total standard database volume, the recalibration of the principal component axis weights is automatically triggered.

[0036] The classification threshold in the discrimination rule is corrected through fuzzy logic, and the updated model parameters can be put into application only after cross-validation.

[0037] The core conceptual logic of the present invention lies in: through the combined use of dual technologies of electron microscopy micro-area element analysis and stable isotope ratio detection, obtaining the calcium and phosphorus element distribution characteristics and biochronological isotope fingerprint information of the keel samples; using sparse partial least squares discriminant analysis to fuse and reduce the dimensionality of multi-source data, and extracting the key features reflecting the essential differences of the samples; constructing a three-dimensional discrimination model based on the first three principal components, and combining an improved density clustering algorithm to achieve accurate classification of true keels, modern bones and processed samples; and continuously updating the model parameters through an adaptive optimization mechanism to form a complete technical chain of "data acquisition - feature extraction - spatial discrimination - dynamic optimization", and finally establishing a scientific, reliable and extensible method system for authenticating the authenticity of traditional Chinese medicine keels.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] 1) Through the innovative combined use of electron microscopy micro-area element analysis and stable isotope fingerprint technology, the present invention breaks through the limitation of a single technical dimension of traditional identification methods and realizes the fusion analysis of multi-source data. The three-dimensional dynamic model constructed based on sparse partial least squares discriminant analysis can synchronously analyze the distribution law of calcium and phosphorus elements and the characteristics of bioisotope ages, improve the discrimination accuracy between true keel fossils and modern bones, and improve the recognition sensitivity of calcined processed samples. In particular, it can effectively identify mineralized samples forged by chemical treatment, and reduce the misjudgment rate compared with the existing infrared spectroscopy method. The built-in adaptive optimization mechanism in the model continuously corrects the discrimination threshold through incremental learning to ensure the continuous anti-interference ability of the method against new forgery means in long-term applications.

[0040] 2) The visual three-dimensional discrimination system established by the method of the present invention converts complex element-isotope data into intuitive spatial position relationships, enabling testers to quickly interpret the results without professional backgrounds. The constructed standard database of Mesozoic fossils covers the characteristic baselines of key species such as three-toed horses and elephants, providing a scientific and quantitative basis for the supervision of the traditional Chinese medicine market. In addition, this technical system can be extended and applied to the identification of animal-derived medicinal materials such as rhinoceros horns and turtle shells, promoting the upgrading of the quality control of traditional Chinese medicine towards digitalization and intelligentization. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 Comparison of unknown sample 1 with the 3D model diagram of the true keel.

[0042] Figure 2 Comparison of unknown sample 2 with the 3D model diagram of the true keel, etc. Detailed implementation mode

[0043] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. In this technical solution, features such as component models, material names, connection structures, control methods, algorithms, etc. that are not clearly stated are regarded as common technical features disclosed in the prior art.

[0044] Example 1

[0045] This embodiment provides a method for identifying Chinese herbal medicine keel by combining micro-area element analysis and isotope fingerprinting, including the following steps:

[0046] S1. Sample pretreatment: Mechanically crush the keel sample to be tested to obtain a uniform powder with a particle size less than 100 μm.

[0047] The sample pretreatment controls the particle size of the keel sample below 100 μm through mechanical crushing. Its core principle is to eliminate the influence of material micro-heterogeneity on the detection accuracy. According to fractal theory, when the specific surface area of the particles reaches 1.2 m 2 / g (D90 < 100 μm), it can ensure that at least 3 complete hydroxyapatite crystal units are included in the SEM observation area (100×100 μm 2 ), meeting the statistical representativeness requirements of EDS energy spectrum analysis. The shear crushing mechanism of the planetary ball mill is adopted during the grinding process, which can avoid the thermal effect damage to the original crystal structure of the fossil caused by traditional grinding and ensure the authenticity of the element distribution characteristics.

[0048] S2. Micro-area element detection: Use the EDS energy spectrometer equipped with a scanning electron microscope to analyze the element composition of at least 3 different micro-areas on the sample surface to obtain the atomic percentage data of the preset elements;

[0049] The working conditions of the EDS energy spectrum analysis are as follows:

[0050] The acceleration voltage is 15 - 20 kV, the beam current is 10 - 15 nA, and the acquisition time for each micro-area is not less than 120 seconds.

[0051] The preset elements include Ca, P, Al, Fe, Mn, Mg, Sr.

[0052] Based on the characteristic X-ray emission law generated by the interaction between the electron beam and matter, characteristic peaks such as Ca-Lα (0.341 keV) and P-Kα (2.013 keV) are captured by the EDS energy spectrometer. The acceleration voltage of 20 kV is set based on the fact that this energy value exceeds the critical excitation energy of Ca (4.04 keV) and P (2.15 keV) and is lower than the interference threshold of heavy metal elements, ensuring that the excitation depth (about 3 μm) matches the thickness of the hydroxyapatite crystal layer. The design of sampling 5 micro-regions conforms to the central limit theorem, narrowing the confidence interval of the elemental atomic percentage data to ±0.5 at% (95% confidence level).

[0053] S3. Stable isotope detection: The stable isotope ratios of δ 13 C, δ 2 H, δ 18 O, and δ 15 N in the sample are determined by the isotope mass spectrometer;

[0054] In S3, the isotope ratio detection includes the following parameter controls:

[0055] δ 13 C detection uses an elemental analysis-isotope mass spectrometry combined system, and the reference material is VPDB;

[0056] δ 2 H and δ 18 O detection uses a high-temperature pyrolysis-isotope mass spectrometry combined system, and the reference materials are VSMOW and SLAP respectively;

[0057] δ 15 N detection uses an elemental analysis-isotope mass spectrometry combined system, and the reference material is atmospheric nitrogen.

[0058] Using the principle of biogeochemical fingerprint, there are significant differences between C3 plants (δ 13 C = -24‰ to -34‰) ingested by Mesozoic mammals and modern C4 plants (δ 13 C = -10‰ to -14‰).

[0059] S4. Data fusion analysis: The elemental composition data obtained in S2 and the isotope ratio data obtained in S3 are subjected to standardization processing, and the sparse partial least squares discriminant analysis method is used for feature extraction;

[0060] In S4, the standardization processing includes the normalization processing of the elemental composition data and the Z-score standardization processing of the isotope ratios.

[0061] In the specific implementation, the elemental composition data are first normalized, and the atomic percentages of elements such as Ca and P are converted into relative values ​​in the interval [0,1]. At the same time, the isotope ratio data are Z-score standardized to eliminate the dimensional differences of indicators such as δ13C and δ15N. Then, the two types of standardized data are aligned according to the sample numbers to construct a joint data matrix. The sparse partial least squares discriminant analysis (sPLS-DA) algorithm is used to screen out 15 key element features and 8 isotope features through L1 regularization constraints to maximize the inter-class discrimination in the dimensionality reduction process. Finally, the loading matrices of the first three discriminant principal components are extracted to generate feature vectors that can be used for three-dimensional modeling, providing a quantitative basis for subsequent sample classification.

[0062] S5. Three-dimensional modeling and discrimination: A three-dimensional discriminant model is established based on the first three principal components extracted by sparse partial least squares discriminant analysis, and the keel category is identified by the clustering characteristics of the sample points in the three-dimensional space.

[0063] In S5, the specific steps include:

[0064] The first three principal components were extracted based on sparse partial least squares discriminant analysis, and a three-dimensional discriminant space including the distribution characteristics of calcium and phosphorus elements and isotopic fingerprints was constructed;

[0065] By setting the coordinate thresholds of the true dragon fossil area and the modern skeleton area, five types of identification rules were constructed by combining the spatial density distribution of sample points and Mahalanobis distance;

[0066] A density clustering algorithm is used to divide the transition zone of calcined samples and the clustering area of ​​mineralized samples. The spatial position relationship between sample points and standard categories is dynamically displayed through an interactive three-dimensional model. When a sample point falls into the preset confidence ellipse range, the specific category is determined according to the coordinate characteristics of the area to which it belongs.

[0067] The above specific categories include true dragon bones, modern bones, true dragon bones that have been calcined or chemically treated, modern bones that have been calcined or chemically treated, and mineralized true dragon bones.

[0068] The density clustering algorithm adopts the improved DBSCAN algorithm, and its parameters are set as follows:

[0069] Neighborhood radius ε = 0.5 × (PC1 standard deviation + PC2 standard deviation + PC3 standard deviation) / 3,

[0070] The above PC1 standard deviation, PC2 standard deviation, and PC3 standard deviation are determined by each principal component

[0071] Minimum number of neighborhood points MinPts = 5 × log (N), where N is the total number of samples;

[0072] Perform a secondary iterative analysis on the discrete points outside the confidence ellipse to identify the transitional zone distribution characteristics of the calcined or chemically treated samples.

[0073] In the interactive 3D model:

[0074] The true keel fossil area is marked with blue (RGB: 0, 71, 171) at a wavelength of 480 nm in the HSV color system, the modern bone area is marked with red (RGB: 203, 32, 39) at a wavelength of 620 nm, and the mineralized true keel area is dynamically rendered with a saturation gradient color scale.

[0075] In S5, it also includes a model iteration and optimization step:

[0076] When the newly added sample volume reaches 20% of the total standard database volume, the re-calibration of the principal component axis weights is automatically triggered.

[0077] The classification threshold in the discrimination rule is corrected through fuzzy logic, and the updated model parameters can be put into application only after cross-validation.

[0078] Application Example 1

[0079] Put the keel sample into a grinder for grinding, take appropriate amounts of powder respectively for detection by electron microscopy and stable isotope ratio, and the data is analyzed through discrimination analysis to visually distinguish the authenticity under the 3D model. The comparison of the 3D model diagrams between unknown sample 1 and the true keel (from Shanghai Traditional Chinese Medicine Materials Co., Ltd.) is as Figure 1 shown, Figure 1 In it, the red is unknown sample 1 (WZ), the blue is the true keel (XD), the green is the modern bone (Z), and the result shows that unknown sample 1 is near the true keel and is a true keel.

[0080] Application Example 2

[0081] Put the keel sample into a grinder for grinding, take appropriate amounts of powder respectively for detection by electron microscopy and stable isotope ratio, and the data is analyzed through discrimination analysis to visually distinguish the authenticity under the 3D model. The comparison of the 3D model diagrams between unknown sample 2 and the true keel (from Shanghai Traditional Chinese Medicine Materials Co., Ltd.) is as Figure 2 shown, Figure 2 In it, the red is unknown sample 2, the blue is the modern bone (XDCL) treated by calcination or chemical means, the green is the modern bone (XD), the purple is the true keel (ZDS) treated by calcination or chemical means, the yellow is the mineralized true keel (ZK), and the result shows that unknown sample 2 is near the true keel and is a true keel.

[0082] The above description of the embodiments is provided to enable those of ordinary skill in the art to understand and use the invention. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative efforts. Therefore, the present invention is not limited to the above embodiments, and the improvements and modifications made by those skilled in the art without departing from the scope of the present invention should be within the protection scope of the present invention according to the disclosure of the present invention.

Claims

1. A method for identifying Chinese medicine dragon bones based on micro-area element analysis combined with isotope fingerprints, characterized in that: The following steps are involved: S1. Sample pretreatment: Mechanically crush the keel sample to obtain a uniform powder with a particle size less than 100 μm; S2. Micro-area element detection: Use an EDS spectrometer equipped with a scanning electron microscope to perform elemental composition analysis on at least three different micro-areas on the sample surface to obtain atomic percentage data of preset elements; S3. Stable isotope detection: The stable isotope ratios of δ13C, δ2H, δ18O and δ15N in the sample are measured by isotope mass spectrometry; S4, data fusion analysis: the elemental composition data obtained in S2 and the isotope ratio data obtained in S3 were standardized, and the sparse partial least squares discriminant analysis method was used for feature extraction; S5. Three-dimensional modeling and discrimination: A three-dimensional discriminant model is established based on the first three principal components extracted by sparse partial least squares discriminant analysis, and the keel category is identified by the clustering characteristics of the sample points in the three-dimensional space.

2. The method for identifying Chinese medicinal dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 1, characterized in that: In S2, the working conditions of the EDS energy spectrum analysis are: The accelerating voltage was 15-20 kV, the beam current was 10-15 nA, and the acquisition time of each micro-area was no less than 120 seconds.

3. The method for identifying Chinese medicinal dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 1, characterized in that: In S2, the preset elements include Ca, P, Al, Fe, Mn, Mg, and Sr.

4. The method for identifying Chinese medicinal dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 1, characterized in that: In S3, the isotope ratio detection includes the following parameter controls: δ13C was detected using an elemental analysis-isotope mass spectrometry system, with VPDB as the standard substance; δ2H and δ18O were detected by high temperature pyrolysis-isotope mass spectrometry system, and the standard substances were VSMOW and SLAP respectively; δ15N detection uses an elemental analysis-isotope mass spectrometry system, and the standard substance is atmospheric nitrogen.

5. The method for identifying Chinese medicine dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 1, characterized in that: In S4, the standardization processing includes normalization processing of elemental composition data and Z-score standardization processing of isotope ratios.

6. The method for identifying Chinese medicine dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 1, characterized in that: In S5, the specific steps include: The first three principal components were extracted based on sparse partial least squares discriminant analysis, and a three-dimensional discriminant space including the distribution characteristics of calcium and phosphorus elements and isotopic fingerprints was constructed; By setting the coordinate thresholds of the true dragon fossil area and the modern skeleton area, five types of identification rules were constructed by combining the spatial density distribution of sample points and Mahalanobis distance; A density clustering algorithm is used to divide the transition zone of calcined samples and the clustering area of ​​mineralized samples. The spatial position relationship between sample points and standard categories is dynamically displayed through an interactive three-dimensional model. When a sample point falls into the preset confidence ellipse range, the specific category is determined according to the coordinate characteristics of the area to which it belongs.

7. The method for identifying Chinese medicinal dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 6, characterized in that: The above specific categories include true dragon bones, modern bones, true dragon bones that have been calcined or chemically treated, modern bones that have been calcined or chemically treated, and mineralized true dragon bones.

8. The method for identifying Chinese medicine dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 6, characterized in that: The density clustering algorithm adopts the improved DBSCAN algorithm, and its parameters are set as follows: Neighborhood radius ε = 0.5 × (PC1 standard deviation + PC2 standard deviation + PC3 standard deviation) / 3, The above PC1 standard deviation, PC2 standard deviation, and PC3 standard deviation are calculated by each principal component. Minimum number of neighborhood points MinPts = 5 × log (N), where N is the total number of samples; A secondary iterative analysis was performed on the discrete points outside the confidence ellipse to identify the transition zone distribution characteristics of the calcined or chemically treated samples.

9. The method for identifying Chinese medicine dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 6, characterized in that: In the interactive 3D model: The true dragon fossil area is marked with 480nm wavelength blue (RGB: 0,71,171) in the HSV color system, the modern bone area is marked with 620nm wavelength red (RGB: 203,32,39), and the mineralized true dragon bone area is dynamically rendered with a saturation gradient color scale.

10. The method for identifying Chinese medicinal dragon bones based on micro-area element analysis combined with isotope fingerprint according to claim 6, characterized in that: S5 also includes the model iterative optimization step: When the amount of new samples reaches 20% of the total amount of the standard database, the principal component axis weight recalibration is automatically triggered. The classification threshold in the discrimination rule is corrected through fuzzy logic, and the updated model parameters can be put into use only after cross-validation.

Images