Subtype equine encephalomyelitis virus diagnosis and recognition system based on artificial intelligence

Abstract

The invention discloses a subtype equine encephalomyelitis virus diagnosis and recognition system based on artificial intelligence, and relates to the technical field of virus recognition systems.Mutation mode feature data are extracted from genome data through a feature extraction and aggregation module by using K-mer analysis, and clinical feature data are extracted from case data; the suddenly-changed mode feature data and the clinical feature data are aggregated as sample data, the classification module distributes a plurality of sub-set sample data to be classified to each computing node for parallel processing, and the evaluation module performs clustering processing according to the running state of each computing node, evaluates the classification effect of each cluster, and outputs the result to the database. And selecting an optimal machine training model for application. The recognition system enables each subset to be processed in parallel on an independent computing node, the data processing efficiency is effectively improved, and after clustering processing is carried out by considering the performance of the computing node, the cluster classification effect is evaluated to select the optimal machine training model application.

Description

Technical Field

[0001] The present invention relates to the technical field of virus identification systems, and in particular to a subtype equine encephalomyelitis virus diagnosis and identification system based on artificial intelligence. Background Art

[0002] Equine encephalitis virus (EEV) is a disease caused by equine encephalitis virus, a member of the Flaviviridae family and a type of encephalitis virus. The virus is mainly transmitted by mosquitoes and is a pathogenic factor in horses and other animals. It can also pose a certain threat to humans. The recognition system is used to detect and identify different subtypes of equine encephalitis virus, which is of great practical significance, especially in the rapid response to epidemic outbreaks and high-risk areas.

[0003] The prior art has the following deficiencies:

[0004] 1. In the diagnosis and identification of virus subtypes, genomic data and case data are often large and complex. Traditional data processing methods usually rely on single-machine computing and cannot efficiently process large-scale genomic data and case data, resulting in low data processing efficiency. Especially when faced with massive data, computing bottlenecks and memory limitations will seriously affect the speed and accuracy of diagnosis and identification;

[0005] 2. Existing recognition systems usually select recognition models randomly or based on simple indicator evaluation. Random selection cannot guarantee the recognition system to have the best recognition efficiency and recognition effect due to its strong randomness. Simple indicator evaluation may lead to a lack of evaluation fairness due to the different operating performance of different computing nodes. Secondly, simple indicator evaluation is not comprehensive enough, which makes it impossible for the recognition system to select the best recognition model application.

[0006] Based on this, the present invention proposes an artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system, which enables each subset to be processed in parallel on an independent computing node, effectively improving data processing efficiency, and after clustering processing considering the performance of the computing nodes, evaluates the cluster classification effect to select the optimal machine training model application. Summary of the invention

[0007] The purpose of the present invention is to provide an artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system to address the shortcomings of the background technology.

[0008] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: a diagnosis and identification system for subtype equine encephalomyelitis virus based on artificial intelligence, comprising a subset division module, a feature extraction and aggregation module, a classification module, and an evaluation module;

[0009] Subset division module: collects genome data and case data to build a data set, and after preprocessing the data set, divides the data set into multiple subsets;

[0010] Feature extraction and aggregation module: Use K-mer analysis to extract mutation pattern feature data from genomic data, and extract clinical feature data from case data, and aggregate mutation pattern feature data and clinical feature data as sample data;

[0011] Classification module: distributes multiple subset sample data to be classified to each computing node for parallel processing. Each computing node uses the trained machine learning model for classification to obtain the classification result of each sample data;

[0012] Evaluation module: Clustering is performed based on the operating status of each computing node, and after evaluating the classification effect of each cluster, the optimal machine training model is selected for application.

[0013] In a preferred embodiment, the feature extraction and aggregation module uses a K-mer algorithm to cut each genome sequence in the genome data to generate a K-mer subsequence, and analyzes the mutation features in the genome data based on the K-mer subsequence;

[0014] Collect case data information, standardize or normalize numerical data, and convert categorical data into numerical data through One-Hot encoding or Label encoding;

[0015] Code clinical symptoms, use text analysis techniques to extract standardized symptom categories or severity from symptom descriptions, combine clinical diagnosis results to extract disease progression or complication characteristics, and extract biomarkers or clinical trial results from laboratory test data;

[0016] The genomic data and case data are matched and aggregated based on patient identifiers or viral subtypes, and the genomic data, clinical data, and symptom information of each patient are integrated to obtain a sample data set.

[0017] In a preferred embodiment, the feature extraction and aggregation module obtains the number of K-mer subsequence divisions in the genome sequence, and obtains the genotype difference value, Huffman coding length and Cosine similarity of each K-mer subsequence;

[0018] The mutation factor of the genome sequence is obtained based on the genotype difference value, Huffman coding length and Cosine similarity calculation;

[0019] The obtained mutation factor is compared with the preset mutation threshold. The mutation threshold is used to determine whether there is a gene mutation in the genome sequence. If the mutation factor is less than or equal to the mutation threshold, it is determined that there is no gene mutation in the genome sequence. If the mutation factor is greater than the mutation threshold, it is determined that there is a gene mutation in the genome sequence.

[0020] In a preferred embodiment, the evaluation module obtains the real-time temperature, temperature rise rate and voltage standard deviation of each computing node, normalizes the real-time temperature, temperature rise rate and voltage standard deviation, maps the value range of the real-time temperature, temperature rise rate and voltage standard deviation to [0, 1], and adds the normalized real-time temperature to the temperature rise rate and the voltage standard deviation to obtain the operation index of the computing node;

[0021] Assume that all computing nodes need to be divided into K clusters. First, randomly select the operation index of K computing nodes as the cluster center value of K clusters, and calculate the absolute value of the difference between the remaining computing nodes and the cluster center value of K clusters. Then assign the computing nodes to the cluster with the smallest absolute value of the difference. After all computing nodes are assigned, recalculate the mean of the operation index of each cluster, and use the mean of the operation index as the cluster center value of the cluster. Repeat the clustering steps. When the difference between the cluster center value obtained in any iteration and the cluster center value obtained in the previous iteration is less than 0.1, it is judged that the convergence condition is met and K clusters are output.

[0022] Get the classification effect coefficient of each computing node in the cluster. The classification effect coefficient is obtained by adding the normalized classification speed and classification accuracy. Calculate the mean of the classification effect coefficients of all clusters, select the cluster with the largest classification effect coefficient, and get the machine learning model used by each computing node in the cluster. Apply the machine learning model with the largest number of uses as the optimal machine training model.

[0023] In a preferred embodiment, the classification module assigns a subset to each computing node, and the computing node uses a pre-trained machine learning model to perform a classification task, and the machine learning model includes a decision tree model, a support vector machine model, a random forest model, and a neural network model;

[0024] After processing the subset assigned to it, each computing node outputs the classification result of the subset. The classification result of each subset is a label. After all computing nodes complete the classification task, the classification results are summarized.

[0025] In a preferred embodiment, the calculation expression of the Huffman code length is:

[0026] , where is the Huffman code length, is the number of K-mer subsequences in the genome sequence, represents the frequency of occurrence of the i-th K-mer subsequence in the genome data, Indicates the bit length of the Huffman code of the i-th K-mer.

[0027] In a preferred embodiment, the calculation expression of the Cosine similarity is: , where is the Cosine similarity, is the frequency vector of a K-mer subsequence in the genome sequence, is the frequency vector of another K-mer subsequence in the genome sequence, is the frequency vector and the frequency vector The dot product of The frequency vectors are The norm and frequency vector of The norm of .

[0028] In a preferred embodiment, the calculation expression of the genotype difference value is: , where is the genotype difference value, is the number of K-mer subsequences in the genome sequence, is the frequency of occurrence of sample data in the i-th K-mer subsequence, is the frequency of occurrence of the i-th K-mer subsequence in the reference genome.

[0029] In a preferred embodiment, the subset division module collects genome data and case data, collects genome data from different virus samples, each virus sample contains a gene sequence of the virus, obtains genome data through genome sequencing results of the virus or a public gene database, and the data format is FASTQ, FASTA genome data format;

[0030] Collect patient-related case data, including patient information, clinical symptoms, diagnostic information, and laboratory test results. Obtain case data through patient electronic health records, medical databases, or laboratory reports. The data format is CSV, JSON, XML structured or semi-structured format;

[0031] Genomic data uses a standard format to save sequence information, case data is unified into a tabular form, numerical features are standardized or normalized, and text data is encoded and converted into numerical data.

[0032] In a preferred embodiment, the subset partitioning module partitions the data set by viral subtype, patient group, balanced data distribution or parallel processing.

[0033] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0034] The present invention uses K-mer analysis to extract mutation pattern feature data from genome data and clinical feature data from case data through feature extraction and aggregation modules, aggregates mutation pattern feature data and clinical feature data as sample data, distributes multiple subset sample data to be classified to each computing node for parallel processing by a classification module, and each computing node uses a trained machine learning model for classification, and an evaluation module performs clustering processing according to the operating status of each computing node, and after evaluating the classification effect of each cluster, selects the optimal machine training model for application. The recognition system enables each subset to be processed in parallel on an independent computing node, effectively improving data processing efficiency, and after clustering processing considering the performance of the computing node, evaluates the cluster classification effect to select the optimal machine training model for application. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] 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 recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0036] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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] Example 1: Please refer to Figure 1 As shown, the artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system described in this embodiment includes a subset division module, a feature extraction and aggregation module, a classification module, and an evaluation module;

[0039] Subset partitioning module: collects genome data and case data to build a data set. After preprocessing the data set (including data formatting and data cleaning), the data set is divided into multiple subsets for parallel processing. The data set can be divided according to virus subtypes, patient groups, etc. The subset partitioning results are sent to the feature extraction and aggregation module and the classification module;

[0040] Feature extraction and aggregation module: Use K-mer analysis to extract mutation pattern feature data from genomic data and clinical feature data from case data. Aggregate mutation pattern feature data and clinical feature data into sample data. For example, genomic data and case data may be stored separately. Aggregate them according to virus subtype or patient identifier to ensure that all relevant information (genomic, clinical, symptom, etc.) of each patient can be combined. The sample data of each subset is sent to the classification module;

[0041] Classification module: distributes multiple subset sample data to be classified to each computing node for parallel processing. Each computing node uses the trained machine learning model to perform classification, obtains the classification result of each sample data, and sends the classification result to the evaluation module;

[0042] Evaluation module: Clustering is performed based on the operating status of each computing node, and after evaluating the classification effect of each cluster, the optimal machine training model is selected for application.

[0043] This application uses K-mer analysis to extract mutation pattern feature data from genomic data and clinical feature data from case data through feature extraction and aggregation modules, aggregates mutation pattern feature data and clinical feature data as sample data, and distributes multiple subset sample data to be classified to each computing node for parallel processing. Each computing node uses a trained machine learning model for classification. The evaluation module performs clustering processing based on the operating status of each computing node, and after evaluating the classification effect of each cluster, selects the optimal machine training model application. The recognition system enables each subset to be processed in parallel on an independent computing node, effectively improving data processing efficiency, and after clustering processing considering the performance of the computing node, evaluates the cluster classification effect to select the optimal machine training model application.

[0044] The workflow of the recognition system is:

[0045] The recognition system collects genomic data and case data to construct a data set. After preprocessing the data set (including data formatting and data cleaning), the data set is divided into multiple subsets for parallel processing. The data set can be segmented by virus subtype, patient group, etc. K-mer analysis is used to extract mutation pattern feature data from the genomic data, and clinical feature data from the case data. The mutation pattern feature data and clinical feature data are aggregated into sample data. For example, genomic data and case data may be stored separately. They are aggregated according to virus subtype or patient identifier to ensure that all relevant information (genomic, clinical, symptom, etc.) of each patient can be combined, and multiple subset sample data to be classified are distributed to each computing node for parallel processing. Each computing node uses a trained machine learning model for classification to obtain the classification result of each sample data. Clustering is performed according to the running status of each computing node. After evaluating the classification effect of each cluster, the optimal machine training model is selected for application.

[0046] Embodiment 2: The subset partitioning module collects genome data and case data to construct a data set, and after preprocessing the data set (including data formatting and data cleaning), the data set is divided into multiple subsets for parallel processing. The data set can be divided according to virus subtype, patient group, etc.;

[0047] The subset partitioning module collects genomic data and case data;

[0048] Genome data collection: Collect genome data from different virus samples. Each sample contains the gene sequence of the virus, which may involve large-scale genome information (such as thousands of gene markers, gene data of multiple samples). The data is obtained through the genome sequencing results of the virus or the public gene database. The data format is FASTQ, FASTA and other genome data formats.

[0049] Case data collection: Collect case data related to patients, including basic information, clinical symptoms, diagnosis information, laboratory test results, etc. Data is obtained through patient electronic health records (EHR), medical databases or laboratory reports. The data format is structured or semi-structured formats such as CSV, JSON, and XML.

[0050] Genomic data and case data are often stored separately and need to be integrated together through patient identifiers or virus subtype identifiers to form a complete sample data set.

[0051] Unify the data from different sources (genomic data and case data) into a unified format for subsequent processing. For genomic data, you can use a standard format (such as FASTA) to save sequence information; for case data, you can unify it into a table format to ensure that the field format of each record is consistent. Detect and process missing values ​​in the data. For example, some fields in the case data may be missing, and some sequence information in the genomic data may be incomplete. Interpolate or delete missing values ​​to ensure data integrity. Delete duplicate data entries to avoid the same data from being repeatedly involved in model training. Identify and correct outliers in the data. For example, there are extreme values ​​in the patient's age field, or abnormal mutation markers that do not conform to biological laws appear in the genomic data.

[0052] Standardize or normalize numerical features (such as sequencing intensity values ​​in genomic data, laboratory index values ​​in case data, etc.) so that data of different scales can be compared within the same range. Encode text data (such as symptom descriptions, diagnosis results, etc.) and convert them into numerical data for processing by machine learning models. Generate different variants of data as needed, such as simulating mutations in genomic data or adding noise to case data, to increase data diversity and enhance the generalization ability of the model.

[0053] The data set partitioning includes the following steps:

[0054] Divide by virus subtype: If the goal is to identify different virus subtypes, the dataset can be divided according to the data of different subtypes. For example, the data can be divided into subtype A, subtype B, subtype C, etc. Each subset contains only the genome data and case data of a specific subtype.

[0055] Split by patient group: The dataset can be divided according to different groups of patients, such as age, gender, disease severity (mild, moderate, severe), etc. The purpose of this is to allow samples from different groups to be processed independently and enable the model to learn the differences between groups.

[0056] Balanced data distribution: When dividing the data set, ensure that the data in each subset has a relatively balanced distribution. For example, in each subset, ensure that the number of samples of different virus subtypes, different patient groups, etc. is roughly the same to avoid too many or too few samples of a certain subtype or group, which will affect the subsequent training effect.

[0057] Parallel processing partitioning: For each subset, they can be further divided into smaller task blocks based on computing resources, which can be processed in parallel on multiple computing nodes. For example, a genomic data subset may be very large, and the task blocks can be further divided by chromosome or gene segment.

[0058] Feature extraction and aggregation module: Use K-mer analysis to extract mutation pattern feature data from genomic data and clinical feature data from case data. Aggregate mutation pattern feature data and clinical feature data as sample data. For example, genomic data and case data may be stored separately. Aggregate them according to virus subtype or patient identifier to ensure that all relevant information (genomic, clinical, symptom, etc.) of each patient can be combined;

[0059] K-mer refers to a substring of length K in a gene sequence (for example, when K=3, all subsequences of length 3). K-mer analysis can help identify important mutation information or variation patterns in genomic sequences, especially the repeated and variable parts of short sequences.

[0060] The feature extraction and aggregation module uses the K-mer algorithm to cut each genome sequence in the genome data, generate all possible K-mer subsequences, and analyze the mutation features in the genome data based on the K-mer subsequences.

[0061] The feature extraction and aggregation module obtains the number of K-mer subsequence divisions in the genome sequence, and obtains the genotype difference value, Huffman coding length and Cosine similarity of each K-mer subsequence;

[0062] The mutation factor of the genome sequence is obtained based on the genotype difference value, Huffman coding length and Cosine similarity calculation, and the expression is:

[0063] , where is a mutation factor, is the genotype difference value, is the Huffman code length, is the Cosine similarity, , , is the adjustment coefficient, and the adjustment coefficient , , All are greater than 0;

[0064] The calculation expression of genotype difference value is: , where is the genotype difference value, is the number of K-mer subsequences in the genome sequence, is the frequency of occurrence of sample data in the i-th K-mer subsequence, is the frequency of occurrence of the i-th K-mer subsequence in the reference genome;

[0065] The genotype difference value measures the frequency difference between the sample and the reference genome on the K-mer subsequence. A larger difference value indicates that there is a significant variation between the sample genome and the reference genome at that position. K-mers with large genotype differences usually correspond to mutation regions, where mutations may affect gene function or lead to the appearance of different phenotypes. An increase in this value indicates an increase in the difference between genomes, which usually means the accumulation of mutations.

[0066] The calculation expression of Huffman code length is:

[0067] , where is the Huffman code length, is the number of K-mer subsequences in the genome sequence, represents the frequency of occurrence of the i-th K-mer subsequence in the genome data, Indicates the bit length of the Huffman code of the i-th K-mer;

[0068] The Huffman code length can help detect mutation regions in the genome. If the code length of a K-mer is significantly longer, it may mean that the K-mer appears less frequently and may represent a region of mutation or genomic variation. Conversely, a shorter code length means that the K-mer is more common in the genome and the incidence of mutation is lower.

[0069] The calculation expression of Cosine similarity is: , where is the Cosine similarity, is the frequency vector of a K-mer subsequence in the genome sequence, is the frequency vector of another K-mer subsequence in the genome sequence, is the frequency vector and the frequency vector The dot product of The frequency vectors are The norm and frequency vector of The norm of, for example, , norm Calculation method and norm of is calculated in the same way;

[0070] The higher the cosine similarity, the more consistent the frequency distribution of the K-mer subsequences of the two genomes. Low cosine similarity means that there are large differences between the genomes. K-mers with low cosine similarity indicate that there are large mutation differences between the genomes, which may be reflected in certain mutation regions or variant regions.

[0071] The larger the mutation factor of the genome sequence, the greater the mutation difference between the genomes of the genome sequence. The obtained mutation factor is compared with the preset mutation threshold. The mutation threshold is used to determine whether there is a gene mutation in the genome sequence. If the mutation factor is less than or equal to the mutation threshold, it is determined that there is no gene mutation in the genome sequence. If the mutation factor is greater than the mutation threshold, it is determined that there is a gene mutation in the genome sequence.

[0072] Collect key information in case data, such as the patient's basic information (gender, age, weight, etc.), clinical symptoms (fever, cough, etc.), diagnostic information (virus type, infection status, etc.), laboratory test results (such as blood tests, imaging examinations, etc.) and treatment plans.

[0073] Numerical data (such as age, body temperature, and laboratory test values) are standardized or normalized, and categorical data (such as symptom description and virus type) are converted into numerical data through One-Hot encoding or Label encoding.

[0074] Encode clinical symptoms, use text analysis techniques (such as natural language processing) to extract standardized symptom categories or severity from symptom descriptions, combine clinical diagnosis results, such as viral infection type, patient's immune response, etc., extract disease progression or possible complication characteristics, and extract biomarkers or clinical trial results from laboratory test data, which often reflect the patient's health status or immune response to the virus. For example, detect certain specific gene expressions, antibody levels, etc.

[0075] Match and aggregate genomic data and case data based on patient identifiers or viral subtypes. Each patient's genomic data, clinical data, symptom information, etc. should be integrated together to form a complete sample data set to ensure that all relevant data for each patient can be stored in the same data table to avoid data loss or redundancy. For each sample, a unified data structure is designed so that information such as genomic features, clinical features, and symptoms can be seamlessly combined. For example, each patient can be represented by a table or vector containing fields such as basic patient information, genomic features (such as K-mer frequency), clinical symptoms, and laboratory test results. This step ensures that in subsequent machine learning models, the input data can cover all relevant information of the patient, so that the model can comprehensively consider genomic variations, clinical manifestations, and other clinical information.

[0076] During the aggregation process, some patients may be missing certain feature information (such as genomic data or certain clinical features), and the missing values ​​need to be filled or marked. Imputation, filling, median filling, or KNN filling can be used to handle missing values ​​in the data to ensure the integrity of the sample data. After data aggregation, the merged data needs to be standardized to ensure that the scale of all features is consistent. Numerical features (such as K-mer frequency, age, etc.) can be standardized (such as subtracting the standard deviation) or normalized (such as compressing the value to between 0 and 1). Categorical features (such as symptom type, treatment plan, etc.) are converted into numerical data through One-Hot encoding or Label encoding.

[0077] All features are integrated to form a high-dimensional feature vector or matrix. Each patient's genomic features (frequency obtained through K-mer analysis) and clinical features (such as symptoms, signs, treatment response, etc.) will be input into the model as a whole. For some features, dimensionality reduction processing (such as PCA, principal component analysis) may be required to reduce the dimension and reduce the computational complexity. Output the aggregated and processed sample data set, ready to be sent to the subsequent classification module for training or inference. The sample data set can be stored and transmitted in formats such as CSV, JSON, and Parquet.

[0078] The feature extraction and aggregation module in this application can effectively extract useful features from genomic data and case data, and aggregate them into unified sample data, providing accurate and comprehensive input for subsequent virus subtype identification. This feature extraction method that integrates genomic and clinical information can significantly improve the diagnostic accuracy and generalization ability of the model.

[0079] The feature extraction and aggregation module collects case data information, standardizes or normalizes the numerical data, and converts the categorical data into numerical data through One-Hot encoding or Label encoding, including the following steps:

[0080] Collect relevant information about patients, including clinical data (such as age, gender, symptoms, etc.), laboratory test results (such as blood tests, imaging test results, etc.), and other clinical variables (such as treatment plans, medication history, etc.). Data formats usually include electronic medical records (EMR), medical database records, or other structured / unstructured formats. Integrate case data from different sources to ensure the relevance of data (such as genomic data corresponding to patient ID, clinical symptoms, test results, etc.). Handle missing values, outliers, or missing entries to ensure data integrity and consistency.

[0081] Standardize the numerical data so that the data has the same dimension. The standardized data distribution has zero mean and unit standard deviation, so that features of different scales can be compared at the same scale to avoid the impact of large values ​​on model training. Normalize the numerical data and scale it to a specified interval (usually [0,1]). Normalization ensures that the values ​​of all features are in the same interval, which helps the convergence speed of algorithm training. Select the standardization or normalization method according to the characteristics of the data and the requirements of the subsequent model. Standardization is applicable to most machine learning algorithms, especially distance-based algorithms (such as KNN, SVM, etc.). Normalization is often used in deep learning and neural networks, where input data needs to be scaled consistently.

[0082] For unordered categorical data (such as gender, blood type, drug use history, etc.), One-Hot encoding is used to convert the categorical data into binary vectors. For example, gender (Male, Female) can be represented by One-Hot encoding as: Male→[1,0], Female→[0,1]. One-Hot encoding can be extended to variables with multiple categories, generating multiple binary feature columns, one feature for each category.

[0083] For ordered categorical data (such as disease severity, age group, etc.), use Label encoding to convert the categories into integers. Label encoding maps each category to a unique integer value, such as Mild→0, Moderate→1, Severe→2. This method can retain the order information between categories and is suitable for categorical variables with a clear order.

[0084] For variables with a large number of categories (such as city names, drug names, etc.), using One-Hot encoding may lead to dimensionality disaster (feature dimension is too high). In this case, other methods such as frequency encoding or embedding representation (such as Word2Vec, embedding layer, etc.) can be used to reduce the dimension and avoid performance degradation.

[0085] The feature extraction and aggregation module encodes clinical symptoms, uses text analysis technology to extract standardized symptom categories or severity from symptom descriptions, combines clinical diagnosis results, extracts disease progression or complication features, and extracts biomarkers or clinical trial results from laboratory test data, including the following steps:

[0086] Symptom classification system: Based on the symptom description text, a standardized symptom classification system (such as ICD-10 or other clinical standards) is applied to classify the symptoms. Each symptom can be mapped to a specific category, for example: headache → neurological symptoms, cough → respiratory symptoms, and the specific symptom category can be identified using natural language processing (NLP) technology (such as entity recognition) based on the context of the symptom description.

[0087] Analyze the severity of symptoms using text analysis techniques such as sentiment analysis or sentiment classification. For example, detect the difference between "severe headache" and "mild headache" and assign a severity value (such as a scale of 1 to 5) to each symptom based on the intensity of the emotion described. Determine the severity of symptoms by extracting keywords or phrases (such as "severe", "mild", "sudden") in the symptom description to form a standardized grading system. Generate standardized symptom codes based on the preprocessed symptom text and combined with the severity assessment. For each patient, the symptom data will be converted into numerical features in a unified format for subsequent analysis.

[0088] Extract disease progression information from electronic medical records or doctors' diagnostic records, such as disease stage, course of disease, clinical diagnosis results, etc. Use medical coding systems (such as ICD-10, SNOMED-CT, etc.) to standardize diagnostic information for subsequent analysis. If the diagnostic information has a time sequence, time series analysis methods can be used to extract disease progression characteristics. For example, information such as the patient's disease progression and treatment response can be converted into time series data to reflect the progression of the disease. According to the patient's diagnostic records, extract the trend of clinical indicators (such as body temperature, blood pressure, pulse, etc.) over time as a sign of disease progression. Extract possible complication information from the case, such as respiratory failure, organ failure, etc. Analyze the timing and frequency of complications and their relationship with the disease itself. Code complications using standardized coding systems (such as ICD-10) and input these codes into subsequent analysis models as one of the disease characteristics.

[0089] Collect laboratory test results from patients, such as blood tests, urine tests, imaging tests, etc. The data may include numerical results (such as blood sugar levels, white blood cell counts, etc.) and categorical results (such as virus test results, imaging scores, etc.). Based on known disease-related biomarkers, extract disease-related marker information from laboratory test data. For example, certain blood markers (such as C-reactive protein, white blood cell count) may be related to the severity of the disease or inflammatory response. For each patient, record the levels of all relevant biomarkers for further analysis. If the patient participates in a clinical trial, extract relevant clinical trial results (such as drug response, treatment effect, etc.). These results can be used as important features such as disease progression and treatment response. Convert clinical trial results into numerical or categorical features, and combine them with the patient's medical history, treatment records, etc. to form a complete clinical feature.

[0090] The feature extraction and aggregation module matches and aggregates genomic data and case data based on patient identifiers or virus subtypes. The genomic data, clinical data, and symptom information of each patient are integrated to obtain a sample data set, including the following steps:

[0091] Collect patients' genomic sequence data from genomic databases or laboratory test results. The data includes information such as gene mutations, single nucleotide polymorphisms (SNPs), insertions and deletions (InDels). Collect patients' clinical information, including basic demographic data (such as age, gender), medical history, medication history, laboratory test results, etc. These data may be stored in electronic medical record (EMR) systems or other medical databases. Clean genomic data and case data, including removing missing values, duplicate data, etc. Standardize or normalize numerical data, and perform One-Hot encoding or Label encoding on categorical data to convert them into numerical forms suitable for analysis.

[0092] Use the patient's unique identifier (such as medical record number or ID number) as a keyword to match genomic data and case data. Ensure that each patient's genomic information is associated with their clinical, symptom information, and other data. For example, a patient's genomic data may contain mutation information, while the patient's case data contains clinical symptoms and laboratory test results. Combine these data using patient identifiers. If the subtype information of the virus is available, the data can be grouped according to the virus subtype. This helps analyze the association between genomic features and clinical manifestations of different subtypes. Combine each patient's genomic data, clinical information, and symptom data, and divide them into subsets based on the virus subtype.

[0093] Aggregate genomic data, clinical data, and symptom information. For example, for the same patient, genomic data and case data may be stored in different tables or databases, and the aggregation process combines these data into a complete data set. The result of the aggregation is a multidimensional feature vector for each patient, which contains their genomic information, clinical characteristics, symptom descriptions, etc. If the case data and genomic data are stored separately, they are merged according to specific rules by virus subtype or patient identifier. Merge the processed genomic data, case data, and symptom information into a complete sample data set. Each patient will correspond to a sample, and the sample includes the following features:

[0094] Genomic data: mutation, SNP information, etc.

[0095] Clinical data: diagnosis results, laboratory test results, etc.;

[0096] Symptom information: symptom category, severity level, etc.;

[0097] The sample data of each patient is represented by a feature vector, and each feature corresponds to a column of data. For example, the feature of genomic data can be the presence or absence of gene mutations, the feature of case data can be clinical diagnosis, laboratory indicators, etc., and the feature of symptom data can be symptom category and severity. The final output data set will contain the feature data of all patients, providing input for subsequent machine learning model training, virus subtype identification, clinical prediction and other tasks.

[0098] Classification module: distributes multiple subset sample data to be classified to each computing node for parallel processing. Each computing node uses the trained machine learning model for classification to obtain the classification result of each sample data;

[0099] The classification module assigns subsets to each computing node. Each computing node is an independent processing unit, usually a working unit in a cluster. Each node runs a machine learning model to classify the data subset assigned to it. When performing classification tasks, the computing node uses the pre-trained machine learning model for reasoning to generate classification results for each subset. These models may be based on different machine learning algorithms, such as decision tree models, support vector machine (SVM) models, random forest models, neural network models, etc.

[0100] The classification module assigns a subset to each computing node. After processing the subset assigned to it, each computing node outputs the classification result of the subset. The classification result of each subset can be a label (such as "subtype 1" or "subtype 2") or a probability distribution (such as "90% belongs to subtype 1", "10% belongs to subtype 2"). After all computing nodes complete the classification task, the classification results are summarized. The classification module integrates these results and passes them to subsequent modules, such as the evaluation module, for further analysis and evaluation of the performance of the model.

[0101] The computing node processes the subset through the pre-trained neural network model. In this application, the pre-training of the neural network model includes the following steps:

[0102] Collect data sets for pre-training, usually including genomic data, case data, clinical data, etc. These data sets should be representative and reflect the diversity of virus subtypes and classification needs. Before inputting the data into the neural network, clean the data to ensure that there are no missing values, outliers, or duplicates. Standardize the data (for example, convert numerical data into a standard normal distribution with a mean of 0 and a variance of 1) or normalize (for example, scale all data to the interval [0, 1]) to ensure the consistency of the scale of the data and avoid certain features from having too much impact on model training. If the data set is small, more training samples can be generated through data enhancement techniques (such as random perturbations, data transformations, etc.) to increase the generalization ability of the model.

[0103] The architecture of a fully connected neural network includes an input layer, a hidden layer, and an output layer. Each layer consists of multiple neurons (nodes). The number of neurons in the input layer is usually equal to the number of features, and the number of neurons in the output layer is equal to the number of categories (in classification tasks, usually the number of virus subtypes).

[0104] Input layer: Each node corresponds to a feature (such as genomic mutations, clinical data, symptoms, etc.).

[0105] Hidden layer: Usually one or more hidden layers are selected to increase the complexity of the network and help capture nonlinear relationships in the data. The number of nodes in the hidden layer is usually between 100 and 1000, adjusted according to the size of the data set.

[0106] Output layer: The number of nodes in the output layer is equal to the number of classification categories. For virus subtype classification, the number of nodes in the output layer is the number of different virus subtypes.

[0107] Activation function: Each hidden layer and output layer node usually uses an activation function (such as ReLU (Rectified-Linear-Unit) or Sigmoid) to increase nonlinear features.

[0108] For hidden layers, the commonly used activation function is ReLU because it can effectively avoid the gradient vanishing problem and speed up training.

[0109] For the output layer, if it is a multi-classification problem, the Softmax activation function is often used to output the probability of each category.

[0110] The dataset is divided into training set, validation set and test set, usually in a ratio of 7:2:1. The training set is used to train the model, the validation set is used to tune the model's hyperparameters, and the test set is used to finally evaluate the model's performance. According to the category of the virus subtype, create corresponding labels. For example, if it is a "subtype 1", "subtype 2" task, the label can be an integer such as 0, 1, etc.

[0111] The input data is passed through the network layer by layer, and weighted summation is performed at each layer and processed through the activation function. The final output layer will generate a probability distribution for each category, and the cross-entropy loss function (Cross-Entropy-Loss) is used to calculate the difference between the output and the actual label. The cross-entropy loss function works well for multi-classification tasks, and can measure the difference between the probability distribution predicted by the model and the actual label. According to the result of the loss function calculation, the gradient of each weight and bias is calculated through the back propagation algorithm. Back propagation is a process that uses the chain rule. It passes the error from the output layer back to the input layer, and updates the parameters (weights and biases) of each layer. The optimization algorithm (such as Adam, SGD, etc.) is used to update the parameters in the network and minimize the loss function. The Adam optimizer performs well when dealing with sparse gradients and large-scale data, and it is usually selected for training.

[0112] The learning rate is a hyperparameter that controls the step size of each parameter update. Techniques such as learning rate decay can be used to avoid oscillation or overfitting during training, and to select an appropriate batch size (such as 32, 64, etc.). Too small a batch size may lead to unstable training, while too large a batch size may lead to insufficient memory, and define the number of training rounds. Early-Stopping is usually used, that is, training is stopped in advance when the performance of the validation set no longer improves to prevent overfitting. After training is completed, the optimal model (including network structure and weights) is saved. You can use the framework API (such as TensorFlow, PyTorch) to save and load the model. In the classification module, load the trained neural network model and use it to predict the subset data to be classified to obtain the classification results of each sample.

[0113] It should be noted that the pre-training process for other machine learning models belongs to the prior art and will not be elaborated in this application.

[0114] Subset data: Suppose we have a dataset containing 1000 samples, and the dataset is divided into 5 subsets, each containing 200 samples.

[0115] Distribute data: Distribute each subset to 5 different computing nodes for parallel processing.

[0116] Node calculation:

[0117] Computing node 1 uses the trained model to classify 200 samples and outputs the classification results: 100 samples are "subtype 1" and 100 samples are "subtype 2".

[0118] Computing node 2 performs the same operation and outputs its classification results: 150 samples are "subtype 1" and 50 samples are "subtype 2".

[0119] Similarly, other computing nodes will also be classified and obtain their own results.

[0120] Result summary: The classification module summarizes the results of the five computing nodes to obtain the final classification result.

[0121] The evaluation module performs clustering according to the operating status of each computing node, evaluates the classification effect of each cluster, and selects the optimal machine training model for application;

[0122] The real-time temperature, temperature rise rate and voltage standard deviation of each computing node are obtained, and the real-time temperature, temperature rise rate and voltage standard deviation are normalized so that the value range of the real-time temperature, temperature rise rate and voltage standard deviation is mapped to [0,1], and the normalized real-time temperature is added to the temperature rise rate and the voltage standard deviation to obtain the operation index of the computing node;

[0123] Assume that all computing nodes need to be divided into K (K is set to 3 in this application) clusters. First, randomly select the operation indexes of K computing nodes as the cluster center values ​​of the K clusters, and calculate the absolute value of the difference between the remaining computing nodes and the cluster center values ​​of the K clusters. Then, assign the computing nodes to the cluster with the smallest absolute value of the difference. After all computing nodes are assigned, recalculate the mean of the operation index of each cluster, and use the mean of the operation index as the cluster center value of the cluster. Repeat the clustering steps. When the difference between the cluster center value obtained in any iteration and the cluster center value obtained in the previous iteration is less than 0.1, it is judged that the convergence condition is met and K clusters are output.

[0124] Get the classification effect coefficient of each computing node in the cluster. The classification effect coefficient is obtained by adding the normalized classification speed and classification accuracy. Calculate the mean of the classification effect coefficients of all clusters, select the cluster with the largest classification effect coefficient, and get the machine learning model used by each computing node in the cluster. Apply the machine learning model with the largest number of uses as the optimal machine training model.

[0125] In the present application, by clustering the computing nodes, it is possible to evaluate the classification effect of computing nodes with the same performance or similar performance, and the evaluation result is more accurate.

[0126] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0127] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0128] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0129] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0130] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. The diagnostic and identification system of subtype equine encephalomyelitis virus based on artificial intelligence is characterized by: It includes subset partitioning module, feature extraction and aggregation module, classification module and evaluation module; Subset division module: collects genome data and case data to build a data set, and after preprocessing the data set, divides the data set into multiple subsets; Feature extraction and aggregation module: Use K-mer analysis to extract mutation pattern feature data from genomic data, and extract clinical feature data from case data, and aggregate mutation pattern feature data and clinical feature data as sample data; Classification module: distributes multiple subset sample data to be classified to each computing node for parallel processing. Each computing node uses the trained machine learning model for classification to obtain the classification result of each sample data; Evaluation module: Clustering is performed based on the operating status of each computing node, and after evaluating the classification effect of each cluster, the optimal machine training model is selected for application.

2. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 1 is characterized in that: The feature extraction and aggregation module uses the K-mer algorithm to cut each genome sequence in the genome data, generates K-mer subsequences, and analyzes mutation features in the genome data based on the K-mer subsequences; Collect case data information, standardize or normalize numerical data, and convert categorical data into numerical data through One-Hot encoding or Label encoding; Code clinical symptoms, use text analysis techniques to extract standardized symptom categories or severity from symptom descriptions, combine clinical diagnosis results to extract disease progression or complication characteristics, and extract biomarkers or clinical trial results from laboratory test data; The genomic data and case data are matched and aggregated based on patient identifiers or viral subtypes, and the genomic data, clinical data, and symptom information of each patient are integrated to obtain a sample data set.

3. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 2 is characterized in that: The feature extraction and aggregation module obtains the number of K-mer subsequence divisions in the genome sequence, and obtains the genotype difference value, Huffman coding length and Cosine similarity of each K-mer subsequence; The mutation factor of the genome sequence is obtained based on the genotype difference value, Huffman coding length and Cosine similarity calculation; The obtained mutation factor is compared with the preset mutation threshold. The mutation threshold is used to determine whether there is a gene mutation in the genome sequence. If the mutation factor is less than or equal to the mutation threshold, it is determined that there is no gene mutation in the genome sequence. If the mutation factor is greater than the mutation threshold, it is determined that there is a gene mutation in the genome sequence.

4. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 3 is characterized in that: The evaluation module obtains the real-time temperature, temperature rise rate and voltage standard deviation of each computing node, normalizes the real-time temperature, temperature rise rate and voltage standard deviation, maps the value range of the real-time temperature, temperature rise rate and voltage standard deviation to [0,1], and adds the normalized real-time temperature to the temperature rise rate and the voltage standard deviation to obtain the operation index of the computing node; Assume that all computing nodes need to be divided into K clusters. First, randomly select the operation index of K computing nodes as the cluster center value of K clusters, and calculate the absolute value of the difference between the remaining computing nodes and the cluster center value of K clusters. Then assign the computing nodes to the cluster with the smallest absolute value of the difference. After all computing nodes are assigned, recalculate the mean of the operation index of each cluster, and use the mean of the operation index as the cluster center value of the cluster. Repeat the clustering steps. When the difference between the cluster center value obtained in any iteration and the cluster center value obtained in the previous iteration is less than 0.1, it is judged that the convergence condition is met and K clusters are output. Get the classification effect coefficient of each computing node in the cluster. The classification effect coefficient is obtained by adding the normalized classification speed and classification accuracy. Calculate the mean of the classification effect coefficients of all clusters, select the cluster with the largest classification effect coefficient, and get the machine learning model used by each computing node in the cluster. Apply the machine learning model with the largest number of uses as the optimal machine training model.

5. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 4 is characterized in that: The classification module assigns a subset to each computing node, and the computing node uses a pre-trained machine learning model to perform classification tasks. The machine learning model includes a decision tree model, a support vector machine model, a random forest model, and a neural network model; After processing the subset assigned to it, each computing node outputs the classification result of the subset. The classification result of each subset is a label. After all computing nodes complete the classification task, the classification results are summarized.

6. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 5 is characterized by: The calculation expression of the Huffman code length is: , where is the Huffman code length, is the number of K-mer subsequences in the genome sequence, represents the frequency of occurrence of the i-th K-mer subsequence in the genome data, Indicates the bit length of the Huffman code of the i-th K-mer.

7. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 6 is characterized by: The calculation expression of the Cosine similarity is: , where is the Cosine similarity, is the frequency vector of a K-mer subsequence in the genome sequence, is the frequency vector of another K-mer subsequence in the genome sequence, is the frequency vector and the frequency vector The dot product of The frequency vectors are The norm and frequency vector of The norm of .

8. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 7 is characterized by: The calculation expression of the genotype difference value is: , where is the genotype difference value, is the number of K-mer subsequences in the genome sequence, is the frequency of occurrence of sample data in the i-th K-mer subsequence, is the frequency of occurrence of the i-th K-mer subsequence in the reference genome.

9. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 8, characterized in that: The subset division module collects genome data and case data, collects genome data from different virus samples, each virus sample contains the gene sequence of the virus, obtains genome data through the genome sequencing results of the virus or the public gene database, and the data format is FASTQ, FASTA genome data format; Collect patient-related case data, including patient information, clinical symptoms, diagnostic information, and laboratory test results. Obtain case data through patient electronic health records, medical databases, or laboratory reports. The data format is CSV, JSON, XML structured or semi-structured format; Genomic data uses a standard format to save sequence information, case data is unified into a tabular form, numerical features are standardized or normalized, and text data is encoded and converted into numerical data.

10. The artificial intelligence-based subtype equine encephalitis virus diagnosis and identification system according to claim 9, characterized in that: The subset partitioning module partitions the data set in the following ways: partitioning by virus subtype, partitioning by patient group, balanced data distribution or parallel processing.

Images