Method, device, medium and product for identifying serum of species based on combination of middle and far infrared spectrum and neural network algorithm

By combining medium- and far-infrared spectroscopy and neural network algorithms, the problem of destructive and cost of samples in the blood distinction of species is solved, and rapid and efficient serum identification is achieved, ensuring high-accurate detection results.

CN118980664AInactive Publication Date: 2024-11-19SHANGHAI INT TRAVEL HEALTH CARE CENT (PORT CLINIC OF SHANGHAI ENTRY-EXIT INSPECTION & QUARANTINE BUREAU) +1
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
CN202411055260.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-02
Publication Date
2024-11-19
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The prior art has the problem of destructive, time-consuming and expensive samples in the blood distinction of species, making it difficult to achieve fast, efficient and non-destructive detection methods.

Method used

Using a method based on a combination of far-infrared spectroscopy and neural network algorithm, the far-infrared spectral data of serum samples are obtained, pre-processed and correlation analysis are performed, the data set is constructed and the neural network model is trained to achieve accurate identification of serums of different species.

Benefits of technology

Fast and efficient species serum identification is achieved, high accuracy is ensured, and detection costs are reduced, providing a reliable and efficient method.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118980664A_ABST
    Figure CN118980664A_ABST
Patent Text Reader

Abstract

The invention discloses a species serum recognition method, device, medium and product based on combination of mid-far infrared spectrum and neural network algorithm, and relates to the field of species serum recognition, the method comprises the following steps: acquiring mid-far infrared spectrum data of a serum sample of each species; preprocessing the middle and far infrared spectrum data; performing correlation analysis on the preprocessed spectral data, and extracting spectral features; constructing a data set according to the spectral characteristics of the serum sample of each species; dividing the data set into a training set and a verification set; training a neural network model by using the training set to obtain a serum recognition model; the neural network model comprises a plurality of segmented progressive neural network nested sub-models; and verifying the serum recognition model by using the verification set. According to the invention, accurate identification and analysis of serum of different species can be rapidly and efficiently realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of species serum identification, and in particular to a species serum identification method, equipment, medium and product based on the combination of mid-infrared and far-infrared spectra and neural network algorithms. Background Art

[0002] Blood is a common and information-rich form of biological evidence, playing a key role in analytical chemistry, forensic analysis, customs animal product quarantine and wildlife protection. Serum, as the product obtained after whole blood centrifugation, contains rich biological molecules (lipids, proteins, carbohydrates, nucleic acids) and important biological information of the donor, and can be used for various applications by detecting changes in biomarkers in serum.

[0003] The conventional methods for distinguishing species blood at present include high performance liquid chromatography (HPLC) and mass spectrometry (MS). Due to its high resolution and high sensitivity, HPLC has been used in blood analysis and has been proven to be a reliable method. At present, species identification can be performed by analyzing the characteristic chromatogram and heme peak of blood samples. MS method can currently identify species blood through the difference in molecular mass of hemoglobin α- and β-chains between different species. However, these two methods must add organic solvents such as acetonitrile before detection, which has certain sample destructiveness, and the measurement method is time-consuming and expensive.

[0004] Therefore, how to provide a fast, efficient and non-destructive detection method to accurately distinguish blood between different species remains a problem that needs to be solved urgently. Summary of the invention

[0005] The purpose of this application is to provide a species serum identification method, equipment, medium and product based on the combination of mid-infrared and far-infrared spectroscopy and neural network algorithm, which can quickly and efficiently realize accurate identification and analysis of sera of different species.

[0006] To achieve the above objectives, this application provides the following solutions:

[0007] In a first aspect, the present application provides a species serum identification method based on a combination of mid-infrared spectroscopy and a neural network algorithm, the species serum identification method based on a combination of mid-infrared spectroscopy and a neural network algorithm comprising:

[0008] Obtain mid- and far-infrared spectral data of serum samples from each species;

[0009] Preprocess mid-infrared and far-infrared spectral data;

[0010] Perform correlation analysis on the preprocessed spectral data and extract spectral features;

[0011] A dataset was constructed based on the spectral characteristics of serum samples of each species; and the dataset was divided into a training set and a validation set;

[0012] The neural network model is trained using the training set to obtain a serum recognition model; the neural network model includes: a plurality of segmented progressive neural network nested sub-models;

[0013] The serum recognition model was validated using the validation set.

[0014] Optionally, the step of obtaining mid-infrared and far-infrared spectral data of serum samples of each species specifically includes:

[0015] Serum samples from each species were thawed and vortexed;

[0016] The sample is evenly dropped onto the surface of the substrate material, and dried to obtain a sample film formed on the surface of the substrate;

[0017] The substrate on which the sample film is currently formed is placed in a mid-infrared and far-infrared spectrometer for spectral detection to obtain mid-infrared and far-infrared spectral data.

[0018] Optionally, the step of obtaining mid-infrared and far-infrared spectral data of serum samples of each species specifically includes:

[0019] Serum samples from each species were thawed and vortexed;

[0020] The sample is evenly dropped onto the surface of the substrate material, and dried to obtain a sample film formed on the surface of the substrate;

[0021] The substrate on which the sample film is currently formed is placed in a mid-infrared and far-infrared spectrometer for spectral detection to obtain mid-infrared and far-infrared spectral data; the amount of serum sample used and the drying time are adjusted according to the actual test sample requirements.

[0022] Optionally, the substrate material includes: high-resistance silicon wafer, quartz and Teflon.

[0023] Optionally, the preprocessing of the mid-infrared and far-infrared spectral data specifically includes: baseline removal operation and normalization operation; the baseline removal operation includes: direct baseline subtraction, polynomial baseline fitting and wavelet transformation; the normalization operation includes: extreme value normalization, logarithmic normalization and standard deviation normalization.

[0024] Optionally, the mid-infrared and far-infrared spectrum data specifically include: amplitude information of each frequency point, relative amplitude information of different peaks, area under the peak, relative information of areas under different peaks and half-height width of each peak.

[0025] Optionally, the neural network model is a convolutional neural network model, a support vector machine, a recurrent neural network, and a long short-term memory model.

[0026] In a second aspect, the present application provides a serum identification device for different species, the serum identification device for different species comprising:

[0027] A spectral data acquisition module, used to acquire mid-infrared and far-infrared spectral data of serum samples of each species;

[0028] A preprocessing module is used to preprocess mid-infrared and far-infrared spectral data;

[0029] The spectral feature extraction module is used to perform correlation analysis on the preprocessed spectral data and extract spectral features;

[0030] A dataset construction module, used to construct a dataset according to the spectral characteristics of serum samples of each species; and divide the dataset into a training set and a validation set;

[0031] The serum recognition model determination module is used to train the neural network model using the training set to obtain the serum recognition model; the neural network model includes: a plurality of segmented progressive neural network nested sub-models;

[0032] The serum recognition model verification module is used to verify the serum recognition model using the verification set.

[0033] In a third aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for identifying sera of different species.

[0034] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for identifying sera of different species.

[0035] In a fifth aspect, the present application provides a computer program product, including a computer program, which implements the method for identifying sera of different species when executed by a processor.

[0036] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0037] The present application provides a method, device, medium and product for species serum identification based on the combination of mid- and far-infrared spectroscopy and neural network algorithms. The neural network model is trained using a training set to obtain a serum identification model; and then when identification is performed, the spectral characteristics of the serum directly obtained are directly input into the serum identification model for automatic identification; the present application reliably and efficiently performs qualitative identification and analysis on serum samples of different species, and ensures high accuracy. The neural network model in the present application includes: multiple segmented progressive neural network nested sub-models; that is, multiple sub-models with different parameters are used to conduct targeted analysis of different characteristic areas in the spectral data, which effectively accelerates network convergence, improves model training efficiency, and provides guarantees for subsequent model detection accuracy. The present application combines a species serum identification method combining mid- and far-infrared spectral data with a neural network algorithm to achieve efficient and accurate species serum identification based on the combination of mid- and far-infrared spectroscopy and a neural network algorithm. The present application provides a reliable and efficient method for quickly distinguishing species serum, and at the same time has important reference significance for studying and identifying body fluids of other species. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 This is an application environment diagram of a species serum identification method based on the combination of mid-infrared and far-infrared spectra and neural network algorithms in an embodiment of the present application;

[0040] Figure 2 A schematic diagram of a process for species serum identification based on a combination of mid-infrared and far-infrared spectra and a neural network algorithm provided in an embodiment of the present application;

[0041] Figure 3 The spectra of serum samples produced from 8 different species are shown below;

[0042] Figure 4 It is a schematic diagram of the CNN network model structure;

[0043] Figure 5 Schematic diagram of the confusion matrix of the nested algorithm model and classification results for multiple neural network sub-models in a piecewise progressive manner. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0045] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0046] The species serum identification method based on the combination of mid-infrared and far-infrared spectra and neural network algorithm provided in the embodiments of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the mid-infrared spectrum data of the serum to the server 104. After the server 104 receives the mid-infrared spectrum data of the serum, the server 104 pre-processes the mid-infrared spectrum data of the serum; performs correlation analysis on the pre-processed spectrum data to extract spectrum features; and uses the serum recognition model to identify the spectrum features. The server 104 can feedback the obtained recognition results to the terminal 102. In addition, in some embodiments, the species serum recognition method based on the combination of mid-infrared spectrum and neural network algorithm can also be implemented by the server 104 or the terminal 102 alone, such as the terminal 102 can directly process the mid-infrared spectrum data of the serum, or the server 104 can obtain the mid-infrared spectrum data of the serum from the data storage system and process the mid-infrared spectrum data of the serum.

[0047] In an exemplary embodiment, Figure 2 As shown, a species serum identification method based on the combination of mid-infrared spectroscopy and neural network algorithm is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following S201 to S206. Among them:

[0048] S201, obtaining mid-infrared and far-infrared spectral data of serum samples of each species; the mid-infrared and far-infrared spectral data specifically include but are not limited to: amplitude information of each frequency point, relative amplitude information of different peaks, area under the peak, relative information of area under different peaks and half-height width of each peak. The mid-infrared and far-infrared spectral data provide information on vibrational rotational frequency resonance absorption of chemical bonds, groups, etc. in serum;

[0049] S202, preprocessing mid-infrared and far-infrared spectrum data;

[0050] S203, performing correlation analysis on the preprocessed spectral data to extract spectral features; the spectral features include but are not limited to: frequency, amplitude, relative peak height, peak area and half-peak full width;

[0051] S204, constructing a data set according to the spectral characteristics of the serum samples of each species; and dividing the data set into a training set and a validation set;

[0052] S205, training a neural network model using the training set to obtain a serum recognition model; the neural network model includes: a plurality of segmented progressive neural network nested sub-models; the neural network model is a convolutional neural network model, a support vector machine, a recurrent neural network, and a long short-term memory model;

[0053] S206, using the validation set to validate the serum recognition model.

[0054] By implementing the above S201-S206, data preprocessing, feature extraction and network learning are performed from the training data, which realizes the automatic and efficient identification of species sera, and reduces the amount of calculation by an order of magnitude while ensuring an accuracy rate of 98.95%, thereby providing a reliable and efficient method for quickly distinguishing species sera, and at the same time has important reference significance for the research and identification of body fluids of other species.

[0055] In another exemplary embodiment of the present application, S201 specifically includes:

[0056] S21, thawing the serum samples of each species and performing a vortex operation; wherein, the sample thawing conditions, vortex time and specific parameter settings of the vortex machine can be adjusted according to the actual process.

[0057] S22, dripping the sample evenly on the surface of the substrate material, and drying it to obtain a sample film formed on the surface of the substrate; the substrate material includes but is not limited to high-resistance silicon wafers, and can be any material that is highly transparent to terahertz waves, such as quartz and Teflon; the sample amount and drying time can be adjusted according to the test sample requirements.

[0058] S23, placing the substrate on which the sample film is currently formed into a mid-infrared and far-infrared spectrometer for spectrum detection to obtain mid-infrared and far-infrared spectrum data. Ensure that the test sample chamber environment is in a vacuum state.

[0059] In another exemplary embodiment of the present application, S202 specifically includes: baseline removal operation and normalization operation. The baseline removal operation may be direct baseline subtraction, polynomial baseline fitting, wavelet transform, etc.; the normalization operation may be extreme value normalization, logarithmic normalization, standard deviation normalization, etc.

[0060] In another exemplary embodiment of the present application, S203 calculates the covariance matrix of the original matrix data column (the first n-1 columns) and the label column (the nth column), and adjusts the feature weight of each data column according to the variance corresponding to each data column of the original matrix to obtain a new data set. 20% of the weighted data are randomly selected as a validation set, and the rest are used as a training set.

[0061] In another exemplary embodiment of the present application, S205 uses multiple sub-models with different parameters to conduct targeted analysis on different characteristic regions in the spectral data, and continuously adjusts the network hyperparameters to find the neural network model with the best effect.

[0062] The accuracy of the serum recognition model verified in S206 is:

[0063]

[0064] In the formula, Accuracy represents the accuracy rate, TP represents the number of positive samples predicted as positive samples; TN represents the number of negative samples predicted as negative samples; FP represents the number of negative samples predicted as positive samples; FN represents the number of positive samples predicted as negative samples.

[0065] The following is described by specific examples:

[0066] S1: Vortex the sera of different species for 5 minutes in a vortex machine to homogenize the samples. In the experiment, the serum samples used included 24 different human serum samples and 123 animal serum samples, and the animal species involved horses, pigs, sheep, cattle, chickens, cats and dogs. Among them, there were 12, 36, 4, 17, 18, 18 and 18 serum samples from different individuals of the seven species of horses, pigs, sheep, cattle, chickens, cats and dogs, respectively;

[0067] S2: Use a pipette to take 10 μL of the sample and evenly drop it on the surface of the high-resistance silicon wafer, and then use a dryer to dry it for 15 minutes;

[0068] S3: The silicon wafer with the sample film obtained in S2 is placed in a Fourier transform infrared spectrometer (FTIR) for detection, and the corresponding spectrum is obtained by collecting signals in the mid- and far-infrared bands. The spectrum is as follows: Figure 3 As shown;

[0069] S4: Perform data preprocessing on the mid-infrared and far-infrared spectra of serum of different species collected in S3. First, remove the baseline of the obtained far-infrared spectrum signal, determine the spectral baseline by measuring the ambient noise, and then subtract the baseline from the original data to obtain the baseline-removed data. Normalize the baseline-removed data to accelerate network convergence.

[0070] S5: Use feature correlation to analyze spectral information to improve network performance. By calculating the covariance matrix of the original matrix data column (the first n-1 columns) and the label column (the nth column), the feature weight of each data column is adjusted according to the variance corresponding to each data column of the original matrix to obtain a new data set. 20% of the adjusted weighted data is randomly selected as a validation set, and the rest is used as a training set.

[0071] S6: Figure 4 As shown, a segmented and progressive nesting method of multiple neural network sub-models is constructed, and multiple sub-models with different parameters are used to analyze different characteristic regions in the spectral data in a targeted manner, and the convolutional neural network model containing multiple sub-models is input using the training set divided in S5 to train the model. In the process of training the network, the influence of each spectral information on the neural network is analyzed, and the network hyperparameters, such as learning rate, number of iterations, number of hidden layers, number of unit nodes, activation function, etc., are continuously adjusted to find the best neural network model.

[0072] In another exemplary embodiment of the present application, taking the convolutional neural network (CNN) model as an example, CNN has good performance in different fields such as data classification, image recognition and natural language processing, making it an indispensable tool for data analysis. Its outstanding feature is that CNN can learn and extract features from input data. For a large amount of spectral data, CNN can keenly discover the characteristic differences between different types, which is suitable for blood spectrum classification tasks. Since the spectral data of the sample is a set of one-dimensional data containing amplitude, frequency, and other information, considering the amount of calculation, this application uses a 1-dimensional CNN architecture to classify blood types.

[0073] The convolutional neural network (CNN) model structure consists of two convolutional layers, two pooling layers and a fully connected layer, such as Figure 5The structure is shown in part (a) of the paper. The convolutional layer uses convolution kernels to perform element-wise multiplication and summation on the input data, with the goal of extracting local features from the input data and capturing spatial relationships and specific patterns. Given that the input data consists of a one-dimensional spectrum, both convolutional layers are designed based on the 1D CNN architecture, with a convolution kernel size of 3 and a stride of 1. In order to reduce the data dimensionality and the number of parameters while retaining important features, a maximum pooling layer is selected. The parameters of the two maximum pooling layers are standardized, with a kernel size and stride of 2. Finally, the neural network understands the relationship and weights between features through a fully connected layer, and at the output layer, a softmax activation function is used to map the final features to their corresponding output categories. The Adaptive MomentEstimation optimizer (hereinafter referred to as the Adam optimizer) is selected as the optimization method for the model. By adopting strategies such as adaptive learning rate and dynamic momentum, the Adam optimizer effectively speeds up the model convergence and produces better results, thereby reducing the burden of hyperparameter tuning.

[0074] The above model was trained under the conditions of 200 epochs, batch-size of 64 per epoch, and learning-rate of 0.001. A total of 95 groups of sample spectra were used for external validation. The final classification results are shown in Figure 5 Part (b) and Figure 5 As shown in part (c) of the figure, the model has a direct classification accuracy of 95%, with a small number of cat samples being misclassified as humans. The reason for this may be that there is too much redundant information in the spectrum, which causes the subtle differences between human blood and cat blood to be submerged. Therefore, a correlation analysis was performed on the sample spectra to reduce the weight of the redundant information.

[0075] S7: Input the validation set spectral data into the serum recognition model to verify the network performance and obtain the accuracy of the serum recognition model.

[0076] For the identification of human and non-human serum, the accuracy of the test network performance was 99%, and the corresponding confusion matrix is ​​as follows Figure 5 As shown in part (b) of the figure, for all eight species, the accuracy of the test network performance is 99%, and the corresponding confusion matrix is ​​as follows Figure 5 As shown in part (c), only a few human and cat serum samples were confused, and the serum samples of other species could be accurately classified.

[0077] The serum samples are not limited to the above eight types, and can be in the form of any animal species and age. The far infrared spectra of serum of different species can be in the form of any band and intensity. When searching for the best neural network model, the network hyperparameters can be adjusted arbitrarily.

[0078] Based on the same inventive concept, the embodiment of the present application also provides a species serum identification device based on a combination of mid- and far-infrared spectra and a neural network algorithm for implementing the species serum identification method based on a combination of mid- and far-infrared spectra and a neural network algorithm involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more embodiments of the species serum identification device based on a combination of mid- and far-infrared spectra and a neural network algorithm provided below can refer to the above limitations on the species serum identification method based on a combination of mid- and far-infrared spectra and a neural network algorithm, which will not be repeated here.

[0079] In an exemplary embodiment, a species serum identification device based on a combination of mid-infrared and far-infrared spectra and a neural network algorithm is provided, comprising:

[0080] A spectral data acquisition module, used to acquire mid-infrared and far-infrared spectral data of serum samples of each species;

[0081] A preprocessing module is used to preprocess mid-infrared and far-infrared spectral data;

[0082] The spectral feature extraction module is used to perform correlation analysis on the preprocessed spectral data and extract spectral features;

[0083] A dataset construction module, used to construct a dataset according to the spectral characteristics of serum samples of each species; and divide the dataset into a training set and a validation set;

[0084] The serum recognition model determination module is used to train the neural network model using the training set to obtain the serum recognition model; the neural network model includes: a plurality of segmented progressive neural network nested sub-models;

[0085] The serum recognition model verification module is used to verify the serum recognition model using the verification set.

[0086] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a species serum identification method based on a combination of mid- and far-infrared spectroscopy and a neural network algorithm is implemented.

[0087] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0088] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0089] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0090] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0091] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0092] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0093] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0094] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for species serum identification based on the combination of mid-infrared and far-infrared spectroscopy and neural network algorithm, characterized in that: The species serum identification method based on the combination of mid-infrared and far-infrared spectra and neural network algorithms includes: Obtain mid- and far-infrared spectral data of serum samples from each species; Preprocess mid-infrared and far-infrared spectral data; Perform correlation analysis on the preprocessed spectral data and extract spectral features; A dataset was constructed based on the spectral characteristics of serum samples of each species; and the dataset was divided into a training set and a validation set; The neural network model is trained using the training set to obtain a serum recognition model; the neural network model includes: a plurality of segmented progressive neural network nested sub-models; The serum recognition model was validated using the validation set.

2. The method for species serum identification based on the combination of mid-infrared and far-infrared spectra and neural network algorithm according to claim 1, characterized in that: The step of obtaining mid-infrared and far-infrared spectral data of serum samples of each species specifically includes: Serum samples from each species were thawed and vortexed; The sample is evenly dropped onto the surface of the substrate material, and dried to obtain a sample film formed on the surface of the substrate; The substrate on which the sample film is currently formed is placed in a mid-infrared and far-infrared spectrometer for spectral detection to obtain mid-infrared and far-infrared spectral data; the amount of serum sample used and the drying time are adjusted according to the actual test sample requirements.

3. The method for species serum identification based on the combination of mid-infrared and far-infrared spectra and neural network algorithm according to claim 2 is characterized in that: Substrate materials include: high-resistance silicon wafer, quartz and Teflon.

4. The method for species serum identification based on the combination of mid-infrared and far-infrared spectra and neural network algorithm according to claim 1, characterized in that: The preprocessing of mid-infrared and far-infrared spectral data specifically includes: baseline removal operation and normalization operation; the baseline removal operation includes: direct baseline subtraction, polynomial baseline fitting and wavelet transformation; the normalization operation includes: extreme value normalization, logarithmic normalization and standard deviation normalization.

5. The method for species serum identification based on the combination of mid-infrared and far-infrared spectra and neural network algorithm according to claim 1, characterized in that: The mid-infrared and far-infrared spectrum data specifically include: amplitude information of each frequency point, relative amplitude information of different peaks, area under the peak, relative information of areas under different peaks and half-height width of each peak.

6. The method for species serum identification based on the combination of mid-infrared and far-infrared spectra and neural network algorithm according to claim 1, characterized in that: The neural network models include convolutional neural network model, support vector machine, recurrent neural network and long short-term memory model.

7. A species serum identification device based on the combination of mid-infrared and far-infrared spectra and neural network algorithms, characterized in that: The species serum identification device based on the combination of mid-infrared and far-infrared spectra and neural network algorithms includes: A spectral data acquisition module, used to acquire mid-infrared and far-infrared spectral data of serum samples of each species; A preprocessing module is used to preprocess mid-infrared and far-infrared spectral data; The spectral feature extraction module is used to perform correlation analysis on the preprocessed spectral data and extract spectral features; A dataset construction module, used to construct a dataset according to the spectral characteristics of serum samples of each species; and divide the dataset into a training set and a validation set; The serum recognition model determination module is used to train the neural network model using the training set to obtain the serum recognition model; the neural network model includes: a plurality of segmented progressive neural network nested sub-models; The serum recognition model verification module is used to verify the serum recognition model using the verification set.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the species serum identification method based on the combination of mid-infrared and far-infrared spectra and a neural network algorithm as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the species serum identification method based on the combination of mid-infrared and far-infrared spectra and a neural network algorithm described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the species serum identification method based on the combination of mid-infrared and far-infrared spectra and a neural network algorithm described in any one of claims 1 to 6 is implemented.

Citation Information

Patent Citations

  • Identification of immunoglobulin (IG) disorders using fourier transform infrared spectroscopy

    CA2637992A1

  • Method for discriminating microorganism by utilizing Fourier infrared spectrum

    CN101556242A

  • Spectrum-based blood species identification instrument and identification method

    CN107045637A

  • Infrared spectrum-based sheep echinococcosis granulosus and sheep brucellosis classification model and establishment method and detection device thereof

    CN115266594A

  • Plasma spectrum-based disease screening model construction method and disease screening device

    CN115274100A