Eriocheir sinensis producing area tracing method based on flavor characteristics
Through a flavour-based method, the origin of Chinese mitten crabs is traced by using top air chromatography-ion migration spectroscopy and statistical model, which solves the problem of high cost and low accuracy of existing aquatic product traceability methods, and achieves efficient and accurate traceability of origin.
Patent Information
- Application Number
- CN202411841861.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-13
- Publication Date
- 2025-04-25
AI Technical Summary
The existing aquatic product traceability methods are costly and the accuracy cannot be guaranteed, making it difficult to effectively solve the problems of irregular land of origin identification and counterfeit and shoddy products.
The origin traceability method of Chinese mitten crab based on flavor characteristics was used to extract the target organ tissue of the sample, and volatile organic compounds were detected using top air chromatography-ion migration spectroscopy, data preprocessing and feature extraction were performed, and the component analysis model and support vector machine model were used for identification to determine the origin.
It improves the accuracy and reliability of the origin traceability of Chinese mitten crabs, reduces analysis costs and time, and achieves efficient origin traceability.
Smart Images
Figure CN119881124A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of aquatic product traceability, and more specifically, to a method for tracing the origin of Chinese mitten crab based on flavor characteristics. Background Art
[0002] Chinese mitten crab (Eriocheir sinensis), also known as hairy crab, is favored by consumers for its delicious meat and rich nutrition. In the market, the origin and quality of Chinese mitten crab have become important factors affecting its price and consumer choice. However, due to the non-standard origin labeling and the existence of counterfeit and inferior products in the market, it is difficult for consumers to distinguish the authenticity when purchasing. In addition, there are significant differences in flavor and taste between Chinese mitten crabs from different origins, which are mainly due to factors such as their growth environment, breeding methods and water quality conditions.
[0003] With the increasing awareness of food safety and quality control, the need for tracing and tracing the origin of aquatic products has become more urgent. At present, the origin of aquatic products is mainly traced based on stable isotope technology. Although stable isotopes have a long half-life and stability, their ratios will still change due to environmental factors. Changes in environmental factors may hide or confuse information about the geographical origin, making the traceability results inaccurate. For example, environmental factors such as climate, soil, and water quality in different regions will affect the composition of stable isotopes in organisms. In addition, the determination of stable isotopes requires high-precision instruments and equipment such as stable isotope mass spectrometers. The experimental costs required for stable isotope determination are also relatively high. The existing methods for tracing the origin of aquatic products are costly and their accuracy cannot be guaranteed. Summary of the invention
[0004] In view of this, the present invention provides a method for tracing the origin of Chinese mitten crab based on flavor characteristics, which is used to solve the problems of existing aquatic product traceability methods, such as high cost and uncertainty in accuracy.
[0005] To achieve the above objectives, the proposed solution is as follows:
[0006] A method for tracing the origin of Chinese mitten crab based on flavor characteristics, comprising:
[0007] Extract target organ tissues from Chinese mitten crab samples;
[0008] The volatile organic compounds in the target organ tissues were detected by headspace gas chromatography-ion mobility spectrometry, and the volatile substance spectrum data of Chinese mitten crab samples were obtained;
[0009] Data preprocessing and feature extraction were performed on the volatile substance spectrum data to obtain the component feature matrix of the Chinese mitten crab sample;
[0010] Input the component feature matrix into the component analysis model for identification to obtain a target feature vector set;
[0011] The target feature vector set is input into the SVM model for identification to obtain the origin of the Chinese mitten crab sample. The SVM model is trained using the component feature matrix of the training sample as the training sample and the origin of the training sample as the sample label.
[0012] Preferably, the target organ tissue is at least one of hepatopancreas, gonad and muscle.
[0013] Preferably, the process of detecting volatile organic compounds in target organ tissues by headspace gas chromatography-ion mobility spectrometry specifically includes:
[0014] Place the target organ tissue in a headspace bottle and allow it to evaporate;
[0015] Extract a certain amount of gas sample from the headspace bottle;
[0016] The extracted gas sample is injected into a gas chromatography system for separation of volatile organic compounds;
[0017] Volatile organic compounds were ionized and mobility analysis was performed in positive ion mode;
[0018] The spectral data of volatile substances were obtained based on ion migration time and ion intensity.
[0019] Preferably, the process of performing data preprocessing and feature extraction on the volatile substance spectrum data specifically includes:
[0020] The spectral data of volatile substances were normalized;
[0021] A smoothing algorithm is used to remove noise from the spectral data of volatile substances;
[0022] The baseline correction technology is used to remove the baseline drift of the spectral data of volatile substances to obtain preprocessed spectral data;
[0023] The main features are extracted from the preprocessed spectral data to obtain a component feature matrix, wherein the main feature is at least one of peak height, peak area and retention time.
[0024] Preferably, the process of the component analysis model identifying the principal components based on the component feature matrix specifically includes:
[0025] Standardize the component feature matrix;
[0026] Calculate the covariance matrix of the standardized component feature matrix;
[0027] Compute the eigenvalues and eigenvectors of the covariance matrix;
[0028] According to the size of each eigenvalue in the covariance matrix, the eigenvectors corresponding to the first k largest eigenvalues are used as principal components to obtain the principal component eigenvector matrix;
[0029] Project the component feature matrix onto the principal component eigenvector matrix to obtain the target eigenvector set.
[0030] Preferably, the kernel function of the SVM model is a linear kernel, a polynomial kernel or a radial basis kernel.
[0031] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:
[0032] The method for tracing the origin of Chinese mitten crab based on flavor characteristics provided by the present invention first extracts the target organ tissue of the Chinese mitten crab sample, then detects the volatile organic compounds in the target organ tissue respectively by headspace gas chromatography-ion mobility spectrometry, and performs data preprocessing and feature extraction on the obtained volatile substance spectrum data. Finally, the component feature matrix is input into the component analysis model for identification, and the obtained target feature vector set is input into the SVM model for identification to obtain the origin of the Chinese mitten crab sample. The present invention combines the volatile organic compound characteristics obtained by gas chromatography analysis with statistical models and learning models to achieve the origin tracing of the Chinese mitten crab, and improves the accuracy and reliability of origin tracing.
[0033] The invention does not require complicated pre-processing and a long analysis process when tracing the origin of the Chinese mitten crab, thereby improving the efficiency of tracing the origin of the Chinese mitten crab. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.
[0035] Figure 1 A flow chart of a method for tracing the origin of Chinese mitten crab based on flavor characteristics provided by an embodiment of the present invention;
[0036] Figure 2a-2f This is a scatter plot of principal component analysis of different tissues of Chinese mitten crab provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0038] First, combine Figure 1 A method for tracing the origin of Chinese mitten crab based on flavor characteristics provided by an embodiment of the present invention is introduced. Figure 1 As shown, the method includes:
[0039] Step S01, extracting target organ tissues from a Chinese mitten crab sample.
[0040] Specifically, the collected Chinese mitten crab samples are processed to extract at least one target organ tissue from the hepatopancreas, gonads and muscles of the Chinese mitten crab samples. Among them, the hepatopancreas is the main organ for metabolism and flavor synthesis of the Chinese mitten crab, and the volatile organic compounds in the hepatopancreas have a significant impact on the flavor of the crab meat.
[0041] Step S02: Detecting volatile organic compounds in target organ tissues respectively by headspace gas chromatography-ion mobility spectrometry.
[0042] Specifically, headspace gas chromatography-ion mobility spectrometry performs well in analyzing volatile organic compounds, which play a key role in the flavor characteristics of food. Therefore, the extracted target organ tissues were analyzed using headspace gas chromatography-ion mobility spectrometry (HS-GC-IMS) technology to obtain the characteristic information of its volatile flavor substances. By optimizing the analysis conditions to ensure good separation effect, the main volatile compounds in the sample can be accurately identified.
[0043] After processing the Chinese mitten crab samples, the extracted target organ tissues were placed in a 20 ml headspace bottle. To maintain the integrity of the volatile organic compounds, the headspace bottle should be kept free of contamination and allowed to stand at an appropriate temperature (e.g., 60°C) for 15 minutes to promote the release of volatile organic compounds.
[0044] In order to ensure the uniformity of the sample and reduce human error, an automatic sampler was used to extract 500 μL of gas sample from the headspace bottle.
[0045] The extracted gas samples were injected into a gas chromatography (GC) system for separation of volatile organic compounds (VOCs), wherein high-purity nitrogen (99.99%) was used as a carrier gas, injected at a programmed flow rate of 2 ml / min for 2 minutes, then increased to 100 ml / min and maintained for 18 minutes.
[0046] The separated volatile organic compounds enter the IMS chamber and are ionized in a tritium ionization source (300 MBq) and subjected to mobility analysis in positive ion mode, with nitrogen as the drift gas at a flow rate of 150 ml / min and a temperature set at 45°C. Through this process, the volatile organic compounds are converted into ions and migrate in the drift tube.
[0047] After the migration analysis, the ion migration time and ion intensity are transmitted to a computer, and a spectrum is formed based on the ion migration time and ion intensity to obtain the volatile substance spectrum data of the Chinese mitten crab sample.
[0048] Step S03, performing data preprocessing and feature extraction on the volatile substance spectrum data.
[0049] Specifically, the volatile substance spectrum data includes ion intensities and migration times of a plurality of volatile organic compounds. In order to better analyze the volatile substance spectrum data, data preprocessing and feature extraction may be performed on the volatile substance spectrum data.
[0050] 1. Data preprocessing:
[0051] First, in order to eliminate the deviation caused by the change of experimental conditions between different samples, the volatile substance spectral data is normalized. Then, a smoothing algorithm such as Savitzky-Golay smoothing is used to remove the noise in the volatile substance spectral data to improve the data quality of the volatile substance spectral data. Finally, the baseline correction technology is used to remove the baseline drift of the volatile substance spectral data to ensure the clarity of the characteristic peaks.
[0052] 2. Feature extraction:
[0053] The main features are extracted from the pre-processed volatile substance spectral data to obtain the component feature matrix of the Chinese mitten crab sample. Among them, the main features include at least one of peak height, peak area and retention time. The component feature matrix is convenient for subsequent analysis and modeling.
[0054] Step S04, inputting the component feature matrix into the component analysis model for identification.
[0055] Specifically, the extracted feature data is input into the component analysis model, and the component analysis model uses dimensionality reduction processing to identify the main flavor feature variables. The core idea of the component analysis model is to find the directions with the largest variance in the data and project the data into these directions. The component analysis model converts high-dimensional data into low-dimensional data through linear transformation while retaining the main features of the data. The process of principal component identification based on the component feature matrix by the component analysis model specifically includes:
[0056] First, the component feature matrix is standardized to ensure that the mean of each feature is 0 and the standard deviation is 1. Then, the covariance matrix of the standardized component feature matrix is calculated, where the covariance matrix is Among them, X is the standardized component feature matrix, X T is the transpose of the standardized component feature matrix, and n is the number of samples. Then, the eigenvalues and eigenvectors of the covariance matrix are calculated, where the eigenvalues reflect the variance in each direction, and the eigenvectors represent the direction of data projection. According to the size of each eigenvalue in the covariance matrix, the eigenvectors corresponding to the first k largest eigenvalues are taken as the principal components to obtain the principal component eigenvector matrix. Finally, the original data of the component feature matrix is projected onto the selected principal component eigenvector matrix to form a new feature space and obtain the target eigenvector set. The target eigenvector set is Z=XW, where W is the principal component eigenvector matrix and Z is the target eigenvector set.
[0057] Step S05, inputting the target feature vector set into the SVM model for identification to obtain the origin of the Chinese mitten crab sample.
[0058] Specifically, support vector machine (SVM) is a supervised learning model for classification and regression analysis. The basic idea is to find an optimal hyperplane to maximize the interval between classes, so as to effectively distinguish data points of different categories. The established SVM model is used to predict the samples to be tested and determine their origin. Among them, the SVM model is trained with the component feature matrix of the training samples as training samples and the origin of the training samples as sample labels. The performance of SVM depends to a large extent on the selection of a suitable kernel function. The choice of kernel function mainly depends on the distribution characteristics of the data. Commonly used kernel functions include linear kernel, polynomial kernel and radial basis kernel.
[0059] The method for tracing the origin of Chinese mitten crab based on flavor characteristics provided in an embodiment of the present invention first extracts the target organ tissue of the Chinese mitten crab sample, then detects the volatile organic compounds in the target organ tissue respectively by headspace gas chromatography-ion mobility spectrometry, and performs data preprocessing and feature extraction on the obtained volatile substance spectrum data. Finally, the component feature matrix is input into the component analysis model for identification, and the obtained target feature vector set is input into the SVM model for identification to obtain the origin of the Chinese mitten crab sample. The embodiment of the present invention combines the volatile organic compound characteristics obtained by gas chromatography analysis with statistical models and learning models to achieve the origin tracing of the Chinese mitten crab, thereby improving the accuracy and reliability of origin tracing.
[0060] The embodiment of the present invention does not require complicated pre-processing and a long analysis process when tracing the origin of the Chinese mitten crab, thereby improving the efficiency of tracing the origin of the Chinese mitten crab.
[0061] Next, the embodiment of the present invention introduces the training process of the SVM model (SVM), and the process is as follows:
[0062] 1. Training Sample Collection
[0063] Collect Chinese mitten crab samples from major breeding areas across the country to ensure that sampling covers different waters, climatic conditions and breeding methods in order to obtain diverse samples. Each sample must record its collection location, time and other relevant information. Sampling time is an important part of ensuring sample quality. The best sampling time is in the morning or evening, which can reduce sample loss caused by temperature and light changes. In addition, collecting samples during the breeding season (such as September to November) can obtain rich volatile organic compounds in hepatopancreatic tissue. Use sterile tools, such as gloves, scissors and sampling bottles to ensure that sample contamination is avoided during the sampling process.
[0064] Collect Chinese mitten crab samples one by one, paying special attention to the consistency of individual size and gender to reduce the impact of biological variation. Each Chinese mitten crab sample should record the specific location, time, environmental factors (such as water temperature, salinity, oxygen content) and other information collected.
[0065] The collected Chinese mitten crab samples were placed in an incubator, quickly transferred to the laboratory, and refrigerated at 4°C to avoid the loss of volatile organic compounds.
[0066] 2. Sample Pretreatment
[0067] The collected Chinese mitten crab samples were processed and the hepatopancreas tissue, gonads and muscles were extracted.
[0068] 3. Headspace Gas Chromatography-Ion Mobility Spectrometry Analysis
[0069] The volatile organic compounds of the organs and tissues of each Chinese mitten crab sample were detected by headspace gas chromatography-ion mobility spectrometry to obtain the volatile substance spectrum data in different organs and tissues of Chinese mitten crab samples from different regions. The specific analysis process is shown in step S02 of the above embodiment and will not be repeated here.
[0070] 4. Data Preprocessing and Feature Extraction
[0071] Use appropriate data processing software to pre-process the volatile substance spectrum data obtained by headspace gas chromatography-ion mobility spectrometry to remove noise and interference signals and improve data quality. At the same time, extract effective volatile flavor substance characteristics. The specific analysis process is shown in step S03 of the above embodiment and will not be repeated here.
[0072] 5. Principal Component Analysis (PCA)
[0073] The extracted characteristic data is input into the principal component analysis model, and the dimension reduction process is performed to identify the main flavor characteristic variables. This step is to find out the differences between samples from different origins, which is convenient for subsequent analysis. The specific analysis process is shown in step S04 of the above embodiment, and will not be repeated here.
[0074] The results showed that 17 signals were detected in the gonads of Chinese mitten crab samples from different regions, including 10 identifiable VOCs, among which the concentrations of 2-butanone, 2-propanol, 2-acetone, 3-pentanol, 1-butanol and 3-methyl were relatively high. Among them, the contents of 2-butanone, 3-pentanol, 1-butanol, 3-methyl and 2-propanol were significantly different.
[0075] Thirty signals were detected in the hepatopancreas of Chinese mitten crab samples from different regions, including 19 identifiable VOCs, among which 2-butanone, (Z)-2-pentenol, 2,3-pentanedione, 1,4-dioxane, 2-methyl-2-propanol and 1-propanol had significant differences in content.
[0076] Twenty-nine signals, including 18 identifiable VOCs, were detected in the muscles of Chinese mitten crab samples from different regions. Among them, the relative abundance of VOCs such as 2-methyl-2-propanol, 4-methyl-2-pentanone, 1-propanol, 2-methyl, 1-butanol, 3-pentanol and 2,3-pentanedione showed significant differences.
[0077] The criteria P<0.05 and VIP≥1 were selected to identify different VOCs in the three tissues, and then PCA was performed to reduce the dimension of the VOCs data, where P is significance and VIP is the variable importance projection value of the data. The P value and VIP value are used to illustrate the differences in the data. Principal component analysis generated a scatter plot of the first two principal components (PCs) of the three Chinese mitten crab sample tissues from different sources, where Figure 2a-2b This is the distribution diagram of the top three volatile organic compounds in the gonad of Chinese mitten crab; Figure 2c-2d This is the composition distribution diagram of the top three volatile organic compounds in the hepatopancreas of Chinese mitten crab; Figure 2e-2f The composition distribution diagram of the top three volatile organic compounds in the muscle of Chinese mitten crab. CH (Chenghu), GC (Guchenghu), XH (Xinghua), YR (Yangtze River) and YZ (Yangzhong) are different sampling origins. The cumulative variance contribution rate of the first two principal components in all three PCAs exceeded 50%, indicating that it is reliable to select PC1 and PC2 for sample analysis. This method selected VOCs suitable for geographical traceability.
[0078] 6. Support Vector Machine (SVM) Model Establishment
[0079] Data preprocessing: The scale of feature values has a great impact on the classification effect, so it is usually necessary to standardize the data. You can use the Z-score standardization method to transform the features into a distribution with a mean of 0 and a standard deviation of 1. Divide the data set into a training set and a test set. A common split ratio is 70% for training and 30% for testing.
[0080] Select kernel function: The performance of SVM depends largely on the selection of appropriate kernel function. Kernel function can map data to high-dimensional feature space, so that data that is linearly inseparable in low-dimensional space becomes linearly separable in high-dimensional space. The choice of kernel function mainly depends on the distribution characteristics of the data. Commonly used kernel functions include: Linear kernel (LinearKernel): Among them, the linear kernel is suitable for linearly separable data. Polynomial Kernel: Where α is the adjustment factor, c is the constant term, and d is the degree of the polynomial. The polynomial kernel is suitable for data in the polynomial feature space. Radial Basis Function (RBF Kernel): K(x i ,x j )=exp(-γ||x i -x j || 2 ), where γ is the parameter of the kernel function, which is used to control the distance between data points in high-dimensional space. The radial basis kernel is suitable for nonlinearly separable data.
[0081] Model training: The SVM model is trained using the component feature matrix of the training samples (such as the relative abundance of volatile organic compounds) as the training samples and the origin of the training samples as the sample labels. The SVM model is established using the selected kernel function and training data. During the training process, the algorithm optimizes the objective function to determine the optimal hyperplane and support vector.
[0082] Model evaluation: The trained SVM model is evaluated through the test set. Common evaluation indicators include accuracy, precision, recall and F1 value.
[0083] Based on the test results, it can be seen that the SVM model of the FM group of Chinese mitten crab muscle and the FHe group of Chinese mitten crab hepatopancreas achieved a 100% discrimination rate. This shows that the use of several special VOCs in muscle and hepatopancreas can trace the origin of river crabs with an accuracy rate of 100%.
[0084] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.
[0085] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.
[0086] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for tracing the origin of Chinese mitten crab based on flavor characteristics, characterized in that: include: Extract target organ tissues from Chinese mitten crab samples; The volatile organic compounds in the target organ tissues were detected by headspace gas chromatography-ion mobility spectrometry, and the volatile substance spectrum data of Chinese mitten crab samples were obtained; Data preprocessing and feature extraction were performed on the volatile substance spectrum data to obtain the component feature matrix of the Chinese mitten crab sample; Input the component feature matrix into the component analysis model for identification to obtain a target feature vector set; The target feature vector set is input into the SVM model for identification to obtain the origin of the Chinese mitten crab sample. The SVM model is trained using the component feature matrix of the training sample as the training sample and the origin of the training sample as the sample label.
2. The method for tracing the origin of Chinese mitten crab based on flavor characteristics according to claim 1, characterized in that: The target organ tissue is at least one of hepatopancreas, gonad and muscle.
3. The method for tracing the origin of Chinese mitten crab based on flavor characteristics according to claim 1, characterized in that: The process of detecting volatile organic compounds in target organ tissues by headspace gas chromatography-ion mobility spectrometry specifically includes: Place the target organ tissue in a headspace bottle and allow it to evaporate; Extract a certain amount of gas sample from the headspace bottle; The extracted gas sample is injected into a gas chromatography system for separation of volatile organic compounds; Volatile organic compounds were ionized and mobility analysis was performed in positive ion mode; The spectral data of volatile substances were obtained based on ion migration time and ion intensity.
4. The method for tracing the origin of Chinese mitten crab based on flavor characteristics according to claim 1, characterized in that: The process of performing data preprocessing and feature extraction on the volatile substance spectrum data specifically includes: The spectral data of volatile substances were normalized; A smoothing algorithm is used to remove noise from the spectral data of volatile substances; The baseline correction technology is used to remove the baseline drift of the spectral data of volatile substances to obtain preprocessed spectral data; The main features are extracted from the preprocessed spectral data to obtain a component feature matrix, wherein the main feature is at least one of peak height, peak area and retention time.
5. The method for tracing the origin of Chinese mitten crab based on flavor characteristics according to claim 1, characterized in that: The component analysis model is based on the process of principal component identification based on the component feature matrix, which specifically includes: Standardize the component feature matrix; Calculate the covariance matrix of the standardized component feature matrix; Compute the eigenvalues and eigenvectors of the covariance matrix; According to the size of each eigenvalue in the covariance matrix, the eigenvectors corresponding to the first k largest eigenvalues are used as principal components to obtain the principal component eigenvector matrix; Project the component feature matrix onto the principal component eigenvector matrix to obtain the target eigenvector set.
6. The method for tracing the origin of Chinese mitten crab based on flavor characteristics according to claim 1, characterized in that: The kernel function of the SVM model is a linear kernel, a polynomial kernel or a radial basis kernel.