A genetic probability assessment system for children's neurological diseases

Abstract

The present invention belongs to the field of neuroscience, and specifically discloses a genetic probability assessment system for children's neurological diseases, which includes: a clinical feature recording module, a family history entry module, a neurological disease data collection module, and a genetic probability comprehensive assessment module. The present invention uses a GW (Guided Whale Optimization Algorithm) optimization method to extract features from the medical data of historical neurological patients, imitates the behavior pattern of whales, has certain intelligent characteristics, can be flexibly adjusted and optimized according to the characteristics of the data, and conducts efficient searches in complex data spaces; uses a hybrid method based on random forest classifiers and naive Bayes to predict genetic probabilities based on children's clinical features and family history, combines the respective advantages of random forests and naive Bayes, reduces the bias and overfitting risks that may exist in a single model, and can effectively improve the accuracy of predictions.

Description

Technical Field

[0001] The present invention relates to the field of neuroscience, and in particular to a genetic probability assessment system for children's neurological diseases. Background Art

[0002] Disease genetic probability assessment can help doctors and parents better understand the genetic risk of children suffering from specific neurological diseases, and provide a certain reference for early diagnosis, prevention and treatment decisions. When making medical diagnoses, it is necessary to classify a large amount of medical data to achieve automation and accelerate disease diagnosis. When processing medical data, there are inaccuracies and unexpected deviations; at the same time, when performing automated classification, it is easy to be limited by the data dimension, resulting in local minima. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a genetic probability assessment system for pediatric neurological diseases. When performing medical diagnosis, it is necessary to classify a large amount of medical data to achieve automation and accelerate disease diagnosis. In view of the technical problem that there are inaccuracies and unexpected deviations when processing medical data, the present invention uses a GW (Guided Whale Optimization Algorithm) optimization method to extract features from the medical data of historical neurological patients, imitate the behavior pattern of whales, have certain intelligent characteristics, can be flexibly adjusted and optimized according to the characteristics of the data, and perform efficient search and exploration in complex data space to find more representative and relevant features, thereby improving the accuracy of genetic disease probability prediction; in view of the technical problem that when performing automatic classification, it is easy to be limited by the data dimension, resulting in local minimum problems, the present invention uses a hybrid method based on random forest classifier and naive Bayes to predict the genetic probability according to the clinical characteristics and family medical history of children, combining the respective advantages of random forest and naive Bayes, reducing the possible deviation and overfitting risks of a single model, and effectively improving the accuracy of prediction.

[0004] The technical solution adopted by the present invention is as follows: The present invention provides a genetic probability assessment system for children's neurological diseases, the genetic probability assessment system for children's neurological diseases comprises a clinical feature recording module, a family medical history entry module, a neurological disease data collection module and a genetic probability comprehensive assessment module;

[0005] The clinical characteristic recording module records the clinical characteristic data of the child, wherein the clinical characteristic data includes gene test results, cognitive impairment, behavioral abnormalities, language and movement disorders, consciousness level, physical development, and neurological development;

[0006] The family medical history entry module records whether there are patients with neurological diseases in the child's family, including the illness conditions of close and distant relatives;

[0007] The neurological disease data collection module records the medical data of historical neurological patients and performs feature extraction on the medical data of historical neurological patients, wherein the medical data of historical neurological patients include clinical manifestations, family medical history, inheritance pattern and corresponding inheritance probability, and marks the clinical manifestations, family medical history and inheritance pattern as features, and marks the inheritance probability as no genetic risk, slight genetic risk, moderate genetic risk and high genetic risk;

[0008] The genetic probability comprehensive evaluation module comprehensively analyzes the gene test results, clinical feature data and family medical history to evaluate the genetic probability;

[0009] The neurology disease data collection module uses a GW optimization method to extract features from the medical data of historical neurology patients, and the GW optimization method specifically includes the following steps:

[0010] Step A1: Check whether there are missing values ​​in the medical data of historical neurology patients. If missing values ​​are found, replace them with the average value of the same feature. If there are true values ​​and false values, replace the true value with 1 and the false value with 0. Calculate the correlation value between every two different features in the medical data. If there is a correlation between two different features, delete any of them. The formula used is as follows:

[0011] ;

[0012] In the formula, is the correlation value, is the number of eigenvalues, and are the eigenvalues ​​of the two features respectively;

[0013] Step A2: Preset the number of iterations and use the GW optimization method to update the features over time. The formula used is as follows:

[0014] ;

[0015] In the formula, is the number of iterations, is the solution during iteration, is the optimal solution during iteration, and are two calculation vectors, , , is a vector that varies linearly between 0 and 2. and Randomly varies between 0 and 1;

[0016] Step A3: Perform feature selection to change the features output by the GW optimization method into binary solutions, represented by 0 and 1, using the following formula:

[0017] ;

[0018] Step A4: Set the fitness function and calculate the fitness value of the feature. The formula used is as follows:

[0019] ;

[0020] In the formula, is the fitness value, is a number in the range [0, 1], , is the number of features output by the GW optimization method, is the number of features in the dataset, represents the error of the GW optimization method;

[0021] Step A5: preset a fitness threshold, and mark the features whose fitness values ​​are greater than the fitness threshold as training features;

[0022] The genetic probability comprehensive evaluation module uses a hybrid method based on a random forest classifier and naive Bayes to make predictions, and the hybrid method based on a random forest classifier and naive Bayes specifically includes the following steps:

[0023] Step B1: Create and initialize the genetic probability prediction model, use Bayes' theorem, and use the training features to calculate the posterior probability. The formula used is as follows:

[0024] ;

[0025] In the formula, is the posterior probability of the inherited probability, is the prior probability of the inheritance probability, is the likelihood, is the prior probability of the training feature;

[0026] Step B2: Perform random sampling with replacement on the training features to obtain the pre-probability of the training features, and perform random sampling on the training features according to the pre-probability to obtain random samples;

[0027] Step B3: construct a decision tree by mapping random samples of the feature subspace, build a decision tree model for each training feature, use the z-dimensional subspace to randomly extract training features from the training features to obtain extracted features, calculate the probabilities of no genetic risk, slight genetic risk, moderate genetic risk and high-risk genetic risk according to the extracted features, generate data segmentation at the leaf nodes of the decision tree, and iterate until saturation is reached;

[0028] Step B4: Combine decision trees to form a random forest ensemble and use high probability values ​​to make classification decisions;

[0029] Step B5: The genetic probability comprehensive evaluation module calculates and outputs the genetic probability of the child's neurological disease based on the child's clinical characteristic data and family medical history.

[0030] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0031] (1) In order to solve the technical problem of inaccuracy and unexpected deviation in processing medical data, the present invention uses a GW (Guided Whale Optimization Algorithm) optimization method to extract features from the medical data of historical neurology patients, imitating the behavior pattern of whales. It has certain intelligent characteristics and can be flexibly adjusted and optimized according to the characteristics of the data. It can perform efficient search and exploration in complex data space to find more representative and relevant features, thereby improving the accuracy of genetic disease probability prediction;

[0032] (2) In order to address the technical problem that when performing automated classification, local minima may occur due to limitations in data dimensions, the present invention uses a hybrid method based on a random forest classifier and naive Bayes to predict genetic probability based on children's clinical characteristics and family medical history. This method combines the respective advantages of random forest and naive Bayes, reduces the bias and overfitting risks that may exist in a single model, and can effectively improve the accuracy of prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 The present invention is a flowchart of a hybrid method based on random forest classifier and naive Bayes in the genetic probability comprehensive evaluation module.

[0034] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0035] The technical solutions 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 only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0036] Example 1: See Figure 1 , this embodiment provides a genetic probability assessment system for children's neurological diseases, the genetic probability assessment system for children's neurological diseases includes a clinical feature recording module, a family medical history entry module, a neurological disease data collection module and a genetic probability comprehensive assessment module;

[0037] The clinical characteristic recording module records the clinical characteristic data of the child, wherein the clinical characteristic data includes gene test results, cognitive impairment, behavioral abnormalities, language and movement disorders, consciousness level, physical development, and neurological development;

[0038] The family medical history entry module records whether there are patients with neurological diseases in the child's family, including the illness conditions of close and distant relatives;

[0039] The neurological disease data collection module records the medical data of historical neurological patients and performs feature extraction on the medical data of historical neurological patients, wherein the medical data of historical neurological patients include clinical manifestations, family medical history, inheritance pattern and corresponding inheritance probability, and marks the clinical manifestations, family medical history and inheritance pattern as features, and marks the inheritance probability as no genetic risk, slight genetic risk, moderate genetic risk and high genetic risk;

[0040] The genetic probability comprehensive evaluation module comprehensively analyzes the gene test results, clinical characteristic data and family medical history to evaluate the genetic probability.

[0041] Embodiment 2: This embodiment is based on the above embodiment. The neurological disease data collection module uses a GW optimization method to extract features from the medical data of historical neurological patients. The GW optimization method specifically includes the following steps:

[0042] Step A1: Check whether there are missing values ​​in the medical data of historical neurology patients. If missing values ​​are found, replace them with the average value of the same feature. If there are true values ​​and false values, replace the true value with 1 and the false value with 0. Calculate the correlation value between every two different features in the medical data. If there is a correlation between two different features, delete any of them. The formula used is as follows:

[0043] ;

[0044] In the formula, is the correlation value, is the number of eigenvalues, and are the eigenvalues ​​of the two features respectively;

[0045] Step A2: Preset the number of iterations and use the GW optimization method to update the features over time. The formula used is as follows:

[0046] ;

[0047] In the formula, is the number of iterations, is the solution during iteration, is the optimal solution during iteration, and are two calculation vectors, , , is a vector that varies linearly between 0 and 2. and Randomly varies between 0 and 1;

[0048] Step A3: Perform feature selection to change the features output by the GW optimization method into binary solutions, represented by 0 and 1, using the following formula:

[0049] ;

[0050] Step A4: Set the fitness function and calculate the fitness value of the feature. The formula used is as follows:

[0051] ;

[0052] In the formula, is the fitness value, is a number in the range [0, 1], , is the number of features output by the GW optimization method, is the number of features in the dataset, represents the error of the GW optimization method;

[0053] Step A5: preset a fitness threshold, and mark the features whose fitness values ​​are greater than the fitness threshold as training features.

[0054] Through the above operations, in order to solve the technical problem of inaccurate and unexpected deviations in processing medical data, the present invention uses a GW (Guided Whale Optimization Algorithm) optimization method to extract features from the medical data of historical neurology patients, imitate the behavior patterns of whales, have certain intelligent characteristics, can be flexibly adjusted and optimized according to the characteristics of the data, and conduct efficient search and exploration in complex data space to find more representative and relevant features, thereby improving the accuracy of genetic disease probability prediction.

[0055] Example 3: See Figure 1 This embodiment is based on the above embodiment, and the genetic probability comprehensive evaluation module uses a hybrid method based on a random forest classifier and naive Bayes to make predictions. The hybrid method based on a random forest classifier and naive Bayes specifically includes the following steps:

[0056] Step B1: Create and initialize the genetic probability prediction model, use Bayes' theorem, and use the training features to calculate the posterior probability. The formula used is as follows:

[0057] ;

[0058] In the formula, is the posterior probability of the inherited probability, is the prior probability of the inheritance probability, is the likelihood, is the prior probability of the training feature;

[0059] Step B2: Perform random sampling with replacement on the training features to obtain the pre-probability of the training features, and perform random sampling on the training features according to the pre-probability to obtain random samples;

[0060] Step B3: construct a decision tree by mapping random samples of the feature subspace, build a decision tree model for each training feature, use the z-dimensional subspace to randomly extract training features from the training features to obtain extracted features, calculate the probabilities of no genetic risk, slight genetic risk, moderate genetic risk and high-risk genetic risk according to the extracted features, generate data segmentation at the leaf nodes of the decision tree, and iterate until saturation is reached;

[0061] Step B4: Combine decision trees to form a random forest ensemble and use high probability values ​​to make classification decisions;

[0062] Step B5: The genetic probability comprehensive evaluation module calculates and outputs the genetic probability of the child's neurological disease based on the child's clinical characteristic data and family medical history.

[0063] Through the above operations, in order to address the technical problem that when performing automatic classification, it is easy to be limited by the data dimension, resulting in local minima, the present invention uses a hybrid method based on random forest classifier and naive Bayes to predict the genetic probability according to the clinical characteristics of children and family medical history. It combines the respective advantages of random forest and naive Bayes, reduces the possible bias and overfitting risk of a single model, and can effectively improve the accuracy of prediction.

[0064] Example 4: This example is based on the above example. The medical data of the historical neurology patients described in Example 1 are obtained internally by the unit, including medical records, diagnostic records and examination results. It is implemented in compliance with the provisions of the Personal Information Protection Law and other relevant laws, is obtained under legal circumstances, and is a case of legal use.

[0065] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0066] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

[0067] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A genetic probability assessment system for children's neurological diseases, characterized in that: It includes a clinical feature recording module, a family medical history entry module, a neurological disease data collection module, and a genetic probability comprehensive assessment module; The clinical characteristic recording module records the clinical characteristic data of the child; The family medical history entry module records whether there are patients with neurological diseases in the child's family, including the illness conditions of close and distant relatives; The neurological disease data collection module records the medical data of historical neurological patients and performs feature extraction on the medical data of historical neurological patients, wherein the medical data of historical neurological patients include clinical manifestations, family medical history, inheritance pattern and corresponding inheritance probability, and marks the clinical manifestations, family medical history and inheritance pattern as features, and marks the inheritance probability as no genetic risk, slight genetic risk, moderate genetic risk and high genetic risk; The genetic probability comprehensive evaluation module comprehensively analyzes the gene test results, clinical feature data and family medical history to evaluate the genetic probability; The neurology disease data collection module uses a GW optimization method to extract features from the medical data of historical neurology patients, and the GW optimization method specifically includes the following steps: Step A1: Check whether there are missing values ​​in the medical data of historical neurology patients. If missing values ​​are found, replace them with the average value of the same feature. If there are true values ​​and false values, replace the true value with 1 and the false value with 0. Calculate the correlation value between every two different features in the medical data. If there is a correlation between two different features, delete any of them. The formula used is as follows: ； In the formula, is the correlation value, is the number of eigenvalues, and are the eigenvalues ​​of the two features respectively; Step A2: Preset the number of iterations and use the GW optimization method to update the features over time. The formula used is as follows: ； In the formula, is the number of iterations, is the solution during iteration, is the optimal solution during iteration, and are two calculation vectors, , , is a vector that varies linearly between 0 and 2. and Randomly varies between 0 and 1; Step A3: Perform feature selection to change the features output by the GW optimization method into binary solutions, represented by 0 and 1, using the following formula: ； Step A4: Set the fitness function and calculate the fitness value of the feature. The formula used is as follows: ； In the formula, is the fitness value, is a number in the range [0, 1], , is the number of features output by the GW optimization method, is the number of features in the dataset, represents the error of the GW optimization method; Step A5: preset a fitness threshold, and mark the features whose fitness values ​​are greater than the fitness threshold as training features; The genetic probability comprehensive evaluation module uses a hybrid method based on a random forest classifier and naive Bayes to make predictions, and the hybrid method based on a random forest classifier and naive Bayes specifically includes the following steps: Step B1: Create and initialize a genetic probability prediction model, use Bayes’ theorem, and use training features to calculate the posterior probability; Step B2: Perform random sampling with replacement on the training features to obtain the pre-probability of the training features, and perform random sampling on the training features according to the pre-probability to obtain random samples; Step B3: construct a decision tree by mapping random samples of the feature subspace, build a decision tree model for each training feature, use the z-dimensional subspace to randomly extract training features from the training features to obtain extracted features, calculate the probabilities of no genetic risk, slight genetic risk, moderate genetic risk and high-risk genetic risk according to the extracted features, generate data segmentation at the leaf nodes of the decision tree, and iterate until saturation is reached; Step B4: Combine decision trees to form a random forest ensemble and use high probability values ​​to make classification decisions; Step B5: The genetic probability comprehensive evaluation module calculates and outputs the genetic probability of the child's neurological disease based on the child's clinical characteristic data and family medical history.

Images