A method, system, device and medium for analyzing the content of beta-lactoglobulin allergen in infant formula rice powder
Patent Information
- Application Number
- CN202611075452.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-20
- Publication Date
- 2026-10-27
AI Technical Summary
但单一拉曼光谱检测在复杂基质环境下易受噪声与杂峰干扰，对微量β-乳球蛋白的定量评价精度不足
[0032]本发明提供的一种婴幼儿配方米粉中β-乳球蛋白致敏原含量分析方法，通过将拉曼光谱技术与高斯加权k近邻回归分析相结合，能够突破传统检测方法的局限，实现婴幼儿配方米粉中β-乳球蛋白致敏原的快速、准确定量分析。
Smart Images
Figure CN122983010A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of food testing technology, specifically relating to a method, system, device, and medium for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal. Background Technology
[0002] Milk allergy is one of the most common food allergies in infants and young children, with an incidence rate of approximately 10% to 40%, and can trigger severe allergic reactions. Beta-lactoglobulin is the main allergenic protein in cow's milk, and about 82% of people with milk allergies are sensitive to it. Infant formula rice cereal is a core food in the complementary food stage of infants and young children. Cross-contamination of beta-lactoglobulin is prone to occur during the production process, and since the immune system of infants and young children is not yet fully developed, even ingesting trace amounts of allergens can induce allergies. Currently, there is no effective cure for food allergies; the only solution is to strictly avoid allergens. Therefore, accurate detection of the beta-lactoglobulin content in infant formula rice cereal is of great significance.
[0003] Traditional β-lactoglobulin detection techniques, such as liquid chromatography-mass spectrometry (LC-MS) and enzyme-linked immunosorbent assay (ELISA), while offering high sensitivity and strong qualitative and quantitative capabilities, suffer from drawbacks including cumbersome sample pretreatment, lengthy detection times, reliance on large and expensive equipment, and long antibody preparation cycles with poor stability. Furthermore, the complex matrix composition of infant formula rice cereal, with components such as starch, vitamins, and minerals easily causing matrix interference, further reduces detection efficiency and result reliability, making it difficult to meet the needs of rapid screening of batch samples and on-site testing.
[0004] Raman spectroscopy, as a rapid and non-destructive detection technique, requires no complex pretreatment and has a short detection cycle, making it a significant advantage in the field of food allergen screening. However, single Raman spectroscopy detection is susceptible to noise and extraneous peak interference in complex matrix environments, resulting in insufficient accuracy for the quantitative evaluation of trace amounts of β-lactoglobulin. Summary of the Invention
[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, system, device and medium for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal, so as to achieve rapid and accurate quantitative analysis of β-lactoglobulin allergen in infant formula rice cereal.
[0006] This invention provides the following technical solution:
[0007] In a first aspect, a method for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal is provided, including: acquiring Raman spectral data of the infant formula rice cereal to be tested;
[0008] The Raman spectral data of the infant formula rice cereal to be tested were preprocessed to obtain the spectral data matrix of the infant formula rice cereal to be tested;
[0009] The spectral data matrix of the infant formula rice cereal to be tested is input into a pre-trained β-lactoglobulin content prediction model to obtain the predicted value of β-lactoglobulin content of the infant formula rice cereal to be tested.
[0010] The training process of the β-lactoglobulin content detection model includes:
[0011] The pre-acquired infant formula rice cereal sample set and its corresponding spectral data matrix were divided into a training set and a validation set; the infant formula rice cereal sample set included infant formula rice cereals with different β-lactoglobulin contents;
[0012] The β-lactoglobulin content detection model was fitted and trained using the training set, with the optimization objective being to maximize the coefficient of determination on the validation set, thus obtaining a well-trained β-lactoglobulin content detection model.
[0013] As an optional technical solution of the present invention, the Raman spectral data of the infant formula rice cereal to be tested includes Raman shift and spectral intensity.
[0014] As an optional technical solution of the present invention, the preprocessing of the Raman spectral data of the infant formula rice cereal to be tested to obtain the spectral data matrix of the infant formula rice cereal to be tested includes:
[0015] Using the Raman shift of the infant formula rice cereal to be tested as the row vector and the spectral intensity corresponding to the Raman shift as the column variable, the original Raman spectral data matrix is constructed.
[0016] The original data matrix is downsampled using a max-pooling window function to obtain a downsampled spectral matrix;
[0017] Baseline correction was performed on the downsampled spectral matrix to obtain the spectral data matrix of the infant formula rice cereal to be tested.
[0018] As an optional technical solution of the present invention, the β-lactoglobulin content detection model adopts a Gaussian weighted k-nearest neighbor regression model.
[0019] As an optional technical solution of the present invention, the step of dividing the pre-acquired infant formula rice cereal sample set and its corresponding spectral data matrix into a training set and a validation set includes:
[0020] After stratifying the infant formula rice cereal sample set according to different β-lactoglobulin contents, representative samples were selected based on the Euclidean distance in spectral space to obtain training and validation sets. In the training and validation sets, the sample proportions of each β-lactoglobulin content were consistent with those in the infant formula rice cereal sample set.
[0021] As an optional technical solution of the present invention, the step of fitting and training the β-lactoglobulin content detection model through a training set, with the optimization objective of maximizing the determination coefficient of the validation set, to obtain a trained β-lactoglobulin content detection model includes:
[0022] The β-lactoglobulin content detection model was fitted and trained using the training set. The optimization objective was to maximize the coefficient of determination on the validation set. The hyperparameters of the Gaussian weighted k-nearest neighbor regression model were optimized by traversing the grid search method to determine the optimal number of nearest neighbors and the optimal Gaussian kernel width of the Gaussian weighted k-nearest neighbor regression model.
[0023] Secondly, a system for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal is provided, including: a data acquisition module for acquiring Raman spectral data of the infant formula rice cereal to be tested;
[0024] The preprocessing module is used to preprocess the Raman spectral data of the infant formula rice cereal to be tested, and obtain the spectral data matrix of the infant formula rice cereal to be tested.
[0025] The prediction module is used to input the spectral data matrix of the infant formula rice cereal to be tested into a pre-trained β-lactoglobulin content prediction model to obtain the predicted value of β-lactoglobulin content of the infant formula rice cereal to be tested.
[0026] The training process of the β-lactoglobulin content detection model includes:
[0027] The pre-acquired infant formula rice cereal sample set and its corresponding spectral data matrix were divided into a training set and a validation set; the infant formula rice cereal sample set included infant formula rice cereals with different β-lactoglobulin contents;
[0028] The β-lactoglobulin content detection model was fitted and trained using the training set, with the optimization objective being to maximize the coefficient of determination on the validation set, thus obtaining a well-trained β-lactoglobulin content detection model.
[0029] Thirdly, an apparatus for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal is provided, comprising a processor and a medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal described in the first aspect.
[0030] Fourthly, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal as described in the first aspect.
[0031] Compared with the prior art, the beneficial effects of the present invention are:
[0032] This invention provides a method for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal. By combining Raman spectroscopy with Gaussian weighted k-nearest neighbor regression analysis, it can overcome the limitations of traditional detection methods and achieve rapid and accurate quantitative analysis of β-lactoglobulin allergen in infant formula rice cereal. Attached Figure Description
[0033] Figure 1 This is a flowchart of the method for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal according to an embodiment of the present invention;
[0034] Figure 2 The above are the average Raman spectra of infant formula rice cereal samples with different β-lactoglobulin contents in the embodiments of the present invention.
[0035] Figure 3 This is a comparison chart of the predicted and actual values of β-lactoglobulin content in the training set in this embodiment of the invention.
[0036] Figure 4 This is a comparison chart of the predicted and actual values of β-lactoglobulin content in the test set of this invention. Detailed Implementation
[0037] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.
[0038] Example 1
[0039] This embodiment provides a method for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal, such as... Figure 1 As shown, the specific steps include the following:
[0040] Step 1: Obtain the Raman spectral data of the infant formula rice cereal to be tested.
[0041] The Raman spectral data of the infant formula rice cereal to be tested include Raman shift and spectral intensity.
[0042] Step 2: Preprocess the Raman spectral data of the infant formula rice cereal to be tested to obtain the spectral data matrix of the infant formula rice cereal to be tested.
[0043] Using the Raman shift of the infant formula rice cereal to be tested as the row vector and the spectral intensity corresponding to the Raman shift as the column variable, the original Raman spectral data matrix is constructed.
[0044] The original data matrix is downsampled using a max-pooling window function, which compresses the spectral dimension while reducing high-frequency random noise, resulting in a downsampled spectral matrix.
[0045] The downsampled spectral matrix was subjected to baseline correction to eliminate baseline drift caused by fluorescence background, thus obtaining the spectral data matrix of the infant formula rice cereal to be tested.
[0046] Step 3: Input the spectral data matrix of the infant formula rice cereal to be tested into the pre-trained β-lactoglobulin content prediction model to obtain the predicted value of β-lactoglobulin content of the infant formula rice cereal to be tested.
[0047] The training process of the β-lactoglobulin content detection model includes:
[0048] S1. Divide the pre-acquired infant formula rice cereal sample set and its corresponding spectral data matrix into a training set and a validation set.
[0049] The infant formula rice cereal sample set includes infant formula rice cereals with different β-lactoglobulin contents.
[0050] In this embodiment, samples of commercially available infant formula rice cereal from different origins, brands, and batches were collected. Enzyme-linked immunosorbent assay (ELISA) was used to detect the β-lactoglobulin content in all samples, and negative baseline samples where β-lactoglobulin was not detected were selected. β-lactoglobulin standards were added to the negative baseline samples to prepare spiked samples with concentration gradients of 0.01 ppm, 0.1 ppm, 1 ppm, 10 ppm, 100 ppm, and 1000 ppm. The content labels were then log-tested. 10 Mathematical transformation is used to prepare a sufficient number of parallel samples for each concentration gradient, ensuring that the dataset covers a wide concentration range and that the sample distribution is spectrally representative.
[0051] Dissolve 0.1 g of rice flour sample into 1 mL of rice flour solution and add the solution to a 96-well plate. Use a confocal Raman spectrometer with a matching microscope stage to acquire the spectra of all spiked samples. Fix the 96-well plate on the microscope stage to acquire Raman signals. Instrument acquisition parameters were set as follows: excitation wavelength 785 nm, laser power 20.0 mW, single spectrum exposure time 0.1 s, and 100 scans per well. The average Raman spectra of samples with different concentrations of spikes are shown below. Figure 2 As shown, the overall spectral shape and characteristic peak positions of all gradient concentration samples are highly consistent. The main characteristic peaks appear at approximately 500 cm⁻¹ and 2500 cm⁻¹. The signal intensity differences caused by different concentrations are very subtle, making it difficult to directly and accurately distinguish the allergen content gradient of the samples with the naked eye. It is necessary to rely on the subsequently constructed Gaussian weighted k-nearest neighbor regression algorithm to deeply mine the subtle feature differences of the full-band spectrum and achieve accurate quantitative discrimination of different β-lactoglobulin contents.
[0052] In this embodiment, the data dimension of the infant formula rice cereal sample set is 3000×1453. The infant formula rice cereal sample set and its corresponding spectral data matrix are obtained by preprocessing the infant formula rice cereal sample set. Specifically, the window function sampling step size is 0.2, corresponding to a window width of 5 original data points. After downsampling the infant formula rice cereal sample set, a downsampled spectral matrix with a dimension of 3000×290 is obtained. Baseline correction processing is performed on the downsampled spectral matrix, with a regularization parameter of 10.0 and an asymmetric residual penalty parameter of 0.01.
[0053] After stratifying the infant formula rice cereal sample set according to different β-lactoglobulin contents, representative samples were selected based on the Euclidean distance in spectral space to obtain training and validation sets. In the training and validation sets, the sample proportions of each β-lactoglobulin content were consistent with those in the infant formula rice cereal sample set.
[0054] S2. The β-lactoglobulin content detection model is fitted and trained using the training set, with the maximum determination coefficient of the validation set as the optimization objective, to obtain the trained β-lactoglobulin content detection model.
[0055] The β-lactoglobulin content detection model uses a Gaussian weighted k-nearest neighbor regression model.
[0056] The Gaussian weighted k-nearest neighbor regression model uses the spectral data of the training set and the corresponding logarithmic data. 10 The concentration label is fitted. For each sample to be predicted, the k nearest training samples are first retrieved. The distance value is then normalized by dividing by the maximum nearest neighbor distance. For samples with a distance of 0, a minimum perturbation value is set to avoid calculation anomalies. Finally, the weights of each nearest neighbor sample are calculated using a Gaussian kernel function. The weight calculation formula is as follows:
[0057] ;
[0058] in, Indicates the first The normalized Euclidean distance between the nearest neighbor samples and the sample to be predicted. This represents the Gaussian kernel width parameter. Indicates the first The weights of the nearest neighbor samples.
[0059] After normalizing all weights, the log of β-lactoglobulin in the sample to be predicted is obtained by weighted summation. 10 Predicted concentration values.
[0060] The β-lactoglobulin content detection model was fitted and trained using the training set, with the optimization objective being to maximize the coefficient of determination on the validation set. The hyperparameters of the Gaussian weighted k-nearest neighbor regression model were optimized through a grid search method to determine the optimal number of nearest neighbors and the optimal Gaussian kernel width. In this embodiment, the search range for the number of nearest neighbors k was 3~40; the Gaussian kernel width was set... The search range is 0.1 to 1000. After iterating through all hyperparameter combinations, the validation set R is selected. 2 The highest combination is taken as the optimal hyperparameter. After optimization, the optimal nearest neighbor number k=4 and the optimal Gaussian kernel width are obtained. =0.1.
[0061] The performance of the trained β-lactoglobulin content detection model was validated using an independent test set. Figure 3 As shown, the model's logarithmic performance on the training set... 10 Concentration prediction coefficient of determination R 2 If the value is 1.0000, the root mean square error of the prediction is 0. For example... Figure 4 As shown, the model's logarithmic performance on the independent test set... 10 Concentration prediction coefficient of determination R 2 It can reach 0.9996, with a root mean square error as low as 0.0347, demonstrating extremely high quantitative prediction accuracy and stability.
[0062] The β-lactoglobulin content in the sample can be obtained by inverse power transformation of the prediction results by 10.
[0063] Example 2
[0064] This embodiment provides a system for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal, including:
[0065] The data acquisition module is used to acquire the Raman spectral data of the infant formula rice cereal to be tested;
[0066] The preprocessing module is used to preprocess the Raman spectral data of the infant formula rice cereal to be tested, and obtain the spectral data matrix of the infant formula rice cereal to be tested.
[0067] The prediction module is used to input the spectral data matrix of the infant formula rice cereal to be tested into a pre-trained β-lactoglobulin content prediction model to obtain the predicted value of β-lactoglobulin content of the infant formula rice cereal to be tested.
[0068] The training process of the β-lactoglobulin content detection model includes:
[0069] The pre-acquired infant formula rice cereal sample set and its corresponding spectral data matrix were divided into a training set and a validation set; the infant formula rice cereal sample set included infant formula rice cereals with different β-lactoglobulin contents;
[0070] The β-lactoglobulin content detection model was fitted and trained using the training set, with the optimization objective being to maximize the coefficient of determination on the validation set, thus obtaining a well-trained β-lactoglobulin content detection model.
[0071] Example 3
[0072] This embodiment provides an apparatus for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal, including a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal described in Embodiment 1.
[0073] Example 4
[0074] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal as described in Embodiment 1.
[0075] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0076] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0077] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0078] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0079] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A method for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal, characterized in that, include: Obtain Raman spectral data of the infant formula rice cereal to be tested; The Raman spectral data of the infant formula rice cereal to be tested were preprocessed to obtain the spectral data matrix of the infant formula rice cereal to be tested; The spectral data matrix of the infant formula rice cereal to be tested is input into a pre-trained β-lactoglobulin content prediction model to obtain the predicted value of β-lactoglobulin content of the infant formula rice cereal to be tested. The training process of the β-lactoglobulin content detection model includes: The pre-acquired infant formula rice cereal sample set and its corresponding spectral data matrix were divided into a training set and a validation set; the infant formula rice cereal sample set included infant formula rice cereals with different β-lactoglobulin contents; The β-lactoglobulin content detection model was fitted and trained using the training set, with the optimization objective being to maximize the coefficient of determination on the validation set, thus obtaining a well-trained β-lactoglobulin content detection model.
2. The method for analyzing the β-lactoglobulin allergen content in infant formula rice cereal according to claim 1, characterized in that, The Raman spectral data of the infant formula rice cereal to be tested include Raman shift and spectral intensity.
3. The method for analyzing the β-lactoglobulin allergen content in infant formula rice cereal according to claim 2, characterized in that, The Raman spectral data of the infant formula rice cereal to be tested are preprocessed to obtain a spectral data matrix of the infant formula rice cereal to be tested, including: Using the Raman shift of the infant formula rice cereal to be tested as the row vector and the spectral intensity corresponding to the Raman shift as the column variable, the original Raman spectral data matrix is constructed. The original data matrix is downsampled using a max-pooling window function to obtain a downsampled spectral matrix; Baseline correction was performed on the downsampled spectral matrix to obtain the spectral data matrix of the infant formula rice cereal to be tested.
4. The method for analyzing the β-lactoglobulin allergen content in infant formula rice cereal according to claim 1, characterized in that, The β-lactoglobulin content detection model uses a Gaussian weighted k-nearest neighbor regression model.
5. The method for analyzing the β-lactoglobulin allergen content in infant formula rice cereal according to claim 1, characterized in that, The step of dividing the pre-acquired infant formula rice cereal sample set and its corresponding spectral data matrix into a training set and a validation set includes: After stratifying the infant formula rice cereal sample set according to different β-lactoglobulin contents, representative samples were selected based on the Euclidean distance in spectral space to obtain training and validation sets. In the training and validation sets, the sample proportions of each β-lactoglobulin content were consistent with those in the infant formula rice cereal sample set.
6. The method for analyzing the β-lactoglobulin allergen content in infant formula rice cereal according to claim 4, characterized in that, The process of fitting and training the β-lactoglobulin content detection model using a training set, with the optimization objective of maximizing the determination coefficient on the validation set, yields a well-trained β-lactoglobulin content detection model, including: The β-lactoglobulin content detection model was fitted and trained using the training set. The optimization objective was to maximize the coefficient of determination on the validation set. The hyperparameters of the Gaussian weighted k-nearest neighbor regression model were optimized by traversing the grid search method to determine the optimal number of nearest neighbors and the optimal Gaussian kernel width of the Gaussian weighted k-nearest neighbor regression model.
7. A system for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal, characterized in that, include: The data acquisition module is used to acquire the Raman spectral data of the infant formula rice cereal to be tested; The preprocessing module is used to preprocess the Raman spectral data of the infant formula rice cereal to be tested, and obtain the spectral data matrix of the infant formula rice cereal to be tested. The prediction module is used to input the spectral data matrix of the infant formula rice cereal to be tested into a pre-trained β-lactoglobulin content prediction model to obtain the predicted value of β-lactoglobulin content of the infant formula rice cereal to be tested. The training process of the β-lactoglobulin content detection model includes: The pre-acquired infant formula rice cereal sample set and its corresponding spectral data matrix were divided into a training set and a validation set; the infant formula rice cereal sample set included infant formula rice cereals with different β-lactoglobulin contents; The β-lactoglobulin content detection model was fitted and trained using the training set, with the optimization objective being to maximize the coefficient of determination on the validation set, thus obtaining a well-trained β-lactoglobulin content detection model.
8. A device for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal, characterized in that, Including processor and media; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method for analyzing the content of β-lactoglobulin allergen in infant formula rice cereal according to any one of claims 1 to 6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method for analyzing the content of β-lactoglobulin allergens in infant formula rice cereal as described in any one of claims 1 to 6.