Auxiliary assessment method and system for acute and critical diseases based on big data analysis

Abstract

The invention discloses an acute and critical disease auxiliary evaluation method and system based on big data analysis, relates to the technical field of data processing, and is used for improving the accuracy and efficiency of acute and critical diseases. The method comprises the following steps: acquiring medical detection data and medical image data of a target user; converting the physical sign data, the medical history data and the medical detection data into a first feature vector, and converting the evoked cause data into a second feature vector; inputting the first feature vector into a first prediction model to obtain a weight value corresponding to each item of data in the first feature vector, which affects the change of each item of data in the second feature vector; inputting the second feature vector into a second prediction model to obtain a weight value corresponding to each item of data in the second feature vector, which affects the change of each item of data in the first feature vector; and determining a disease degree prediction value of the target user according to the first feature vector and the weight value of each data thereof, the second feature vector and the weight value of each data thereof, and the medical image data.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an emergency and critical illness auxiliary assessment method and system based on big data analysis. Background Art

[0002] Critical illness refers to clinical signs of a serious, variable, and life-threatening emergency. Critical illness identification is to assess and judge the patient's condition through various methods to determine whether he or she is in a critical state. Critical illness usually refers to a serious, life-threatening disease or condition, such as a heart attack, stroke, severe infection, etc.

[0003] At present, we can only rely on the expert experience of the attending physician to make judgments on critical and severe diseases, but relying solely on the expert experience of the attending physician may lead to errors or low efficiency. Summary of the invention

[0004] The embodiments of the present invention provide a method and system for assisting emergency and critical illness assessment based on big data analysis, which are used to improve the accuracy and efficiency of emergency and critical illness assessment.

[0005] An embodiment of the present invention provides an emergency and critical illness auxiliary assessment method based on big data analysis, the method comprising: Obtain the target user's medical history data, physical sign data, and induced disease cause data; Obtaining a physical sign data comparison table corresponding to the induced disease cause data and the medical history data, and determining whether the physical sign data of the target user is within a normal data range according to the physical sign data comparison table; If the target user's vital sign data is not within the normal data range, obtaining the target user's medical examination data and medical imaging data; Converting the target user's vital sign data, medical history data and medical test data into a first feature vector, and converting the target user's disease-inducing cause data into a second feature vector; Input the first feature vector into a first prediction model to obtain weight values ​​corresponding to each item of data in the first feature vector that affect changes in each item of data in the second feature vector; input the second feature vector into a second prediction model to obtain weight values ​​corresponding to each item of data in the second feature vector that affect changes in each item of data in the first feature vector; The predicted value of the severity of the disease of the target user is determined by using the weight value of the first feature vector and its various data items, the weight value of the second feature vector and its various data items, and the medical imaging data.

[0006] In an optional embodiment provided by the present invention, the step of obtaining a physical sign data comparison table corresponding to the causal data and the medical history data, and determining whether the physical sign data of the target user is within a normal data range according to the physical sign data comparison table includes: Determine the disease type based on the causative agent data and the medical history data, and determine a physical sign data comparison table corresponding to the disease type, wherein the physical sign data comparison table includes the normal data range of each physical sign data corresponding to the disease type; The target user's vital sign data is compared with the vital sign data in the vital sign data comparison table. If the target user's vital sign data is not within the normal data range of the corresponding vital sign data in the vital sign data comparison table, it is determined that the target user's vital sign data is not within the normal data range.

[0007] In an optional embodiment provided by the present invention, the determining the disease type based on the causative disease data and the medical history data includes: Extracting entity keywords from the causative disease data and the medical history data, wherein the entity keywords at least include disease name, drug name, and examination result; Determine a connection relationship between entity keywords based on the semantic relationship between the entity keywords, wherein the connection relationship at least includes a cause, a symptom, and a complication; The disease type is determined based on the logical relationship between the entity keywords and the connection relationship.

[0008] In an optional embodiment provided by the present invention, the converting of the target user's vital sign data, medical history data and the medical test data into a first feature vector, and the converting of the target user's induced disease cause data into a second feature vector includes: Performing data preprocessing on the target user's vital sign data, medical history data, medical test data, and induced disease cause data, wherein the data preprocessing at least includes: abnormal data filtering, data standardization, and normalization; Input the target user's physical sign data, medical history data, medical test data, and induced disease cause data that have undergone data preprocessing into the feature extraction model to obtain data features corresponding to the physical sign data, medical history data, medical test data, and induced disease cause data; Combining the target user's vital sign data, medical history data, and data features corresponding to the medical test to obtain the first feature vector; The data feature corresponding to the disease-inducing cause data of the target user is determined as a second feature vector.

[0009] In an optional embodiment provided by the present invention, before inputting the first feature vector into a first prediction model to obtain weight values ​​corresponding to each data item in the first feature vector that affect changes in each data item in the second feature vector; and inputting the second feature vector into a second prediction model to obtain weight values ​​corresponding to each data item in the second feature vector that affect changes in each data item in the first feature vector, the method further includes: Determine a first sample feature vector according to the physical sign data, medical history data and medical test data in the sample data, and determine a second sample feature vector according to the induced disease cause data in the corresponding sample data; Using the second sample feature vector as a sample label of the first sample feature vector to perform model training to obtain the first prediction model; The first sample feature vector is used as a sample label of the second sample feature vector to perform model training to obtain the second prediction model.

[0010] In an optional embodiment provided by the present invention, the step of using the second sample feature vector as a sample label of the first sample feature vector to perform model training to obtain the first prediction model includes: Inputting the first sample feature vector into the first prediction model to obtain a first prediction weight value of the influence of each data in the first sample feature vector on the second sample feature vector; Determine, based on the data relationship between the data in the preset knowledge base, the influence value of each data in the first sample feature vector on the change of each data in the second sample feature vector; Determine an average value of all influence values ​​corresponding to the same item of data in the first sample feature vector as a first actual weight value; Calculate a first model loss value according to the first predicted weight value and the first actual weight value; When the loss value of the first model is less than a preset value, the training of the first prediction model is stopped.

[0011] In an optional embodiment provided by the present invention, the step of performing model training using the first sample feature vector as a sample label of the second sample feature vector to obtain the second prediction model includes: Inputting the second sample feature vector into the second prediction model to obtain a second prediction weight value of the influence of each data in the second sample data on the first sample data; the initial model parameters of the first prediction model are the model parameters of the second prediction model; Calculating a second actual weight value according to a numerical relationship between the first sample feature vector and the second sample feature vector; Calculate a second model loss value according to the second predicted weight value and the second actual weight value; When the loss value of the second model is less than a preset value, the training of the second prediction model is stopped.

[0012] In an optional embodiment provided by the present invention, determining the predicted value of the severity of illness of the target user by using the first feature vector and the weight value of each item of data thereof, the second feature vector and the weight value of each item of data thereof, and the medical imaging data includes: Determine a first disease severity prediction value based on the first feature vector and the weight values ​​of its various data items, and the second feature vector and the weight values ​​of its various data items; determine a second disease severity prediction value based on the medical imaging data; The predicted illness severity value of the target user is determined by using the first illness severity predicted value and the second illness severity predicted value.

[0013] In an optional embodiment provided by the present invention, determining the first severity prediction value according to the first feature vector and the weight value of each data thereof, and the second feature vector and the weight value of each data thereof; determining the second severity prediction value according to the medical imaging data includes: Inputting the first feature vector and the weight value of each item of data thereof, and the second feature vector and the weight value of each item of data thereof into a first illness severity prediction model to obtain a first illness severity prediction value; The medical imaging data is input into a second illness severity prediction model to obtain a second illness severity prediction value.

[0014] The embodiment of the present invention provides an emergency and critical illness auxiliary assessment system based on big data analysis, wherein the system comprises: The acquisition module is used to obtain the target user's medical history data, physical sign data, and induced disease cause data; The acquisition module is further used to obtain a comparison table of physical sign data corresponding to the induced disease cause data and the medical history data; A determination module, used to determine whether the target user's vital sign data is within a normal data range according to the vital sign data comparison table; The acquisition module is further configured to acquire the medical examination data and medical imaging data of the target user if the vital sign data of the target user is not within the normal data range; A conversion module, used to convert the target user's vital sign data, medical history data and the medical test data into a first feature vector, and convert the target user's induced disease cause data into a second feature vector; A prediction module, configured to input the first feature vector into a first prediction model to obtain weight values ​​corresponding to each item of data in the first feature vector that affect changes in each item of data in the second feature vector; and input the second feature vector into a second prediction model to obtain weight values ​​corresponding to each item of data in the second feature vector that affect changes in each item of data in the first feature vector; The determination module is further used to determine the predicted value of the severity of the disease of the target user through the weight value of the first feature vector and its various data, the weight value of the second feature vector and its various data, and the medical imaging data.

[0015] The present invention provides an emergency and critical illness auxiliary assessment method and system based on big data analysis. First, the medical history data, physical sign data, and induced cause data of a target user are obtained, and then a physical sign data comparison table corresponding to the induced cause data and the medical history data is obtained, and whether the physical sign data of the target user is within a normal data range is determined according to the physical sign data comparison table; if the physical sign data of the target user is not within the normal data range, the medical test data and medical imaging data of the target user are obtained; then, the physical sign data, medical history data, and medical test data of the target user are converted into a first feature vector, and the induced cause data of the target user are converted into a second feature vector; the first feature vector is input into a first prediction model to obtain weight values ​​corresponding to each data item in the first feature vector that affect changes in each data item in the second feature vector; the second feature vector is input into a second prediction model to obtain weight values ​​corresponding to each data item in the second feature vector that affect changes in each data item in the first feature vector; finally, the predicted value of the severity of the disease of the target user is determined through the weight values ​​of the first feature vector and its each data item, the weight values ​​of the second feature vector and its each data item, and the medical imaging data. Compared with the existing method of relying on the expert experience of the attending physician to judge the emergency and critical illness, the present invention determines the weight value based on the acquisition of various data, and then jointly determines the predicted value of the severity of the target user based on the weight value and data such as the first eigenvector and the second eigenvector. Therefore, the accuracy and efficiency of the emergency and critical illness assessment can be improved through the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of an emergency and critical illness auxiliary assessment method based on big data analysis provided by the present invention; Figure 2 A schematic diagram of the structure of an emergency and critical illness auxiliary assessment system based on big data analysis provided by the present invention. DETAILED DESCRIPTION

[0017] In order to better understand the above technical scheme, the technical scheme of the embodiment of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiment of the present invention and the specific features in the embodiments are detailed descriptions of the technical scheme of the embodiment of the present invention, rather than limitations on the technical scheme of the present invention. In the absence of conflict, the embodiment of the present invention and the technical features in the embodiments can be combined with each other.

[0018] See also Figure 1 This embodiment provides an emergency and critical illness auxiliary assessment method based on big data analysis, and the method execution steps are: S101, obtaining medical history data, physical sign data, and induced disease cause data of the target user; Among them, the medical history data may include: the current symptoms, physical signs, medication status, etc. of the target user; the target user's past medical history, surgical history, drug allergy history, etc.; whether there are people with hereditary diseases or similar symptoms in the target user's family. Physical sign data may include: body temperature, pulse, respiratory rate, blood pressure, etc.; induced disease cause data may include: whether the target user's environment is likely to cause disease factors, such as air pollution, radiation, etc.; the target user's eating habits, exercise, sleep conditions, etc.; whether the target user's mental state is stable, whether there is anxiety, depression, etc. This embodiment limits the specific forms of medical history data, physical sign data, and induced disease cause data.

[0019] S102, obtaining a physical sign data comparison table corresponding to the causal data and the medical history data, and determining whether the physical sign data of the target user is within a normal data range according to the physical sign data comparison table.

[0020] In this embodiment, the type of disease the target user suffers from can be determined based on the induced disease data and the medical history data, and then the corresponding physical sign data comparison table is determined based on the disease type. For example, if the target user is determined to have diabetes through the induced disease data and the medical history data, then a physical sign data comparison table corresponding to diabetes needs to be determined, that is, the table contains the normal data range of each physical sign data focused on by diabetes, such as the normal data range of parameters such as blood sugar monitoring, body mass index (BMI) and blood pressure.

[0021] In an optional embodiment provided by the present invention, the step of obtaining a physical sign data comparison table corresponding to the causal data and the medical history data, and determining whether the physical sign data of the target user is within a normal data range according to the physical sign data comparison table includes: S1021, determining the disease type based on the induced disease cause data and the medical history data, and determining a physical sign data comparison table corresponding to the disease type; Among them, the physical sign data comparison table includes the normal data range of each physical sign data corresponding to the disease type. Specifically, in this implementation, the disease type is determined based on the induced cause data and the medical history data, including: extracting entity keywords from the induced cause data and the medical history data, the entity keywords at least include the disease name, drug name, and test results; determining the connection relationship between the entity keywords based on the semantic relationship between the entity keywords, the connection relationship at least includes the cause, symptoms, and complications; determining the disease type based on the logical relationship between the entity keywords and the connection relationship.

[0022] In this embodiment, the causative agent data and the medical history data can be matched specifically according to the keywords in the preset database, and the matched keywords are determined as entity keywords. Then, the connection relationship between the entity keywords is determined based on the contextual semantic relationship of the entity keywords in the causative agent data and the medical history data. The connection relationship is used to represent the cause, symptoms, complications, etc. Finally, the disease type is determined through the logical relationship between the entity keywords and the connection relationship.

[0023] For example, the name of the drug is A, and the test result is a high blood sugar level. Combining the semantic relationship, it can be determined that the high blood sugar level leads to drug A. Drug A is used to assist in controlling blood sugar. Combining this logical relationship, it can be determined that the disease type is diabetes.

[0024] S1022, comparing the target user's vital sign data with the vital sign data in the vital sign data comparison table; if the target user's vital sign data is not within the normal data range of the corresponding vital sign data in the vital sign data comparison table, determining that the target user's vital sign data is not within the normal data range.

[0025] S103, if the vital sign data of the target user is not within the normal data range, obtaining the medical examination data and medical imaging data of the target user; Among them, medical test data include: blood sugar, blood lipids, liver function, kidney function, white blood cell count, red blood cell count, hemoglobin concentration, platelet count; urine color, transparency, pH value, protein, sugar and other indicators. Medical imaging data include X-ray images (such as chest X-rays, abdominal X-rays, etc.), CT images (such as head CT, chest CT, abdominal CT, etc.), MRI images (such as head MRI, spine MRI, joint MRI, etc.), ultrasound images (such as cardiac ultrasound, liver ultrasound, kidney ultrasound, etc.).

[0026] S104, converting the target user's vital sign data, medical history data and the medical test data into a first feature vector, and converting the target user's disease-causing data into a second feature vector.

[0027] In an optional embodiment provided by the present invention, the converting of the target user's vital sign data, medical history data and the medical test data into a first feature vector, and the converting of the target user's induced disease cause data into a second feature vector includes: S1041, preprocessing the target user's vital sign data, medical history data, medical test data, and induced disease cause data; The data preprocessing includes at least: abnormal data filtering, data standardization, normalization, etc.

[0028] S1042, inputting the target user's physical sign data, medical history data, medical test data, and induced disease cause data after data preprocessing into the feature extraction model to obtain data features corresponding to the physical sign data, medical history data, medical test data, and induced disease cause data; Among them, the feature extraction model can use the existing neural network model. By inputting the processed data into the model, the data features corresponding to each data item can be obtained, so as to determine the feature vector based on the data features in the subsequent steps.

[0029] S1043: Combine the target user's vital sign data, medical history data, and data features corresponding to the medical examination to obtain the first feature vector.

[0030] S1044: Determine the data feature corresponding to the disease-inducing cause data of the target user as a second feature vector.

[0031] In this embodiment, by pre-processing the data of the target user and inputting them into the feature extraction model, data features corresponding to the physical sign data, medical history data, medical test data, and induced disease data can be obtained, thereby obtaining the first feature vector and the second feature vector based on the extracted data features. Since the feature extraction model is a neural network model, the data features of the corresponding data can be well extracted based on the model, so that the first feature vector and the second feature vector determined by this embodiment can better reflect the characteristics of the data itself.

[0032] S105, input the first eigenvector into the first prediction model to obtain the weight values ​​corresponding to the data items in the first eigenvector that affect the changes of the data items in the second eigenvector; input the second eigenvector into the second prediction model to obtain the weight values ​​corresponding to the data items in the second eigenvector that affect the changes of the data items in the first eigenvector.

[0033] In this embodiment, before obtaining the weight values ​​corresponding to each data item, it is necessary to perform model training on the first prediction model and the second prediction model. The training process is as follows: S201, determining a first sample feature vector according to the physical sign data, medical history data and medical test data in the sample data, and determining a second sample feature vector according to the induced disease cause data in the corresponding sample data.

[0034] It should be noted that, in this embodiment, the first sample feature vector and the second sample feature vector are Figure 1 The method for determining the first eigenvector and the second eigenvector in the corresponding steps is the same, and will not be repeated in this embodiment.

[0035] S202: Use the second sample feature vector as a sample label of the first sample feature vector to perform model training to obtain the first prediction model.

[0036] Specifically, the step of using the second sample feature vector as a sample label of the first sample feature vector to perform model training to obtain the first prediction model includes: S2021: Input the first sample feature vector into the first prediction model to obtain a first prediction weight value of the influence of each data in the first sample feature vector on the second sample feature vector.

[0037] S2022, determining the influence of each data item in the first sample feature vector on the change of each data item in the second sample feature vector based on the data relationship between each data item in the preset knowledge base; The data relationship between each data in the preset knowledge base is a specific value written in advance, such as the influence value of data A in the first sample feature vector on data B in the second sample feature vector is 0.6. The response value can be a proportional multiple, an inverse multiple of the data change, or an influence value of a linear relationship, etc., which is not specifically limited in this embodiment.

[0038] S2023, determining an average value of all influence values ​​corresponding to the same item of data in the first sample feature vector as a first actual weight value; As through the formula: Determine a first actual weight value, where n is the amount of data in the first sample feature vector, m is the amount of data in the second sample feature vector, is the impact value of the i-th item of data in the first sample feature vector on the data change of the j-th item of data in the second sample feature vector.

[0039] S2024, calculating a first model loss value according to the first predicted weight value and the first actual weight value; ; Among them, loss is the loss value of the first model, n is the total number of samples, and m is the amount of data in each sample data. is the first prediction weight value of the influence of the j-th item of data in the first sample vector of the i-th sample data on the corresponding second sample feature vector, is the first actual weight value of the j-th item of data in the first sample vector of the i-th sample data affecting the corresponding second sample feature vector.

[0040] S2025: When the loss value of the first model is less than a preset value, stop training the first prediction model.

[0041] S203: Using the first sample feature vector as the sample label of the second sample feature vector to perform model training to obtain the second prediction model.

[0042] Specifically, the step of using the first sample feature vector as a sample label of the second sample feature vector to perform model training to obtain the second prediction model includes: S2031, inputting the second sample feature vector into the second prediction model to obtain a second prediction weight value of the influence of each data in the second sample data on the first sample data; The initial model parameters of the first prediction model are the model parameters of the second prediction model.

[0043] S2032: Calculate a second actual weight value according to a numerical relationship between the first sample feature vector and the second sample feature vector.

[0044] Specifically, this embodiment can calculate the numerical relationship between the first sample feature vector and the second feature vector through similarity, that is, the numerical relationship is the similarity of the two feature vectors, and then preset the correspondence between the similarity and the weight value in the mapping table, and determine the weight value corresponding to the similarity as the second actual weight value.

[0045] S2033: Calculate a second model loss value based on the second predicted weight value and the second actual weight value.

[0046] S2034: When the loss value of the second model is less than a preset value, stop training the second prediction model.

[0047] S106: Determine the predicted value of the severity of the illness of the target user through the weight value of the first feature vector and its various data items, the weight value of the second feature vector and its various data items, and the medical imaging data.

[0048] In an optional embodiment provided by the present invention, determining the predicted value of the severity of illness of the target user by using the first feature vector and the weight value of each item of data thereof, the second feature vector and the weight value of each item of data thereof, and the medical imaging data includes: S1061, determining a first illness severity prediction value based on the weight values ​​of the first eigenvector and its various data items, and the weight values ​​of the second eigenvector and its various data items; determining a second illness severity prediction value based on the medical imaging data.

[0049] Specifically, the determining of a first illness severity prediction value according to the weight values ​​of the first feature vector and its various data items, and the second feature vector and its various data items; determining a second illness severity prediction value according to the medical imaging data includes: inputting the weight values ​​of the first feature vector and its various data items, and the second feature vector and its various data items into a first illness severity prediction model to obtain a first illness severity prediction value; inputting the medical imaging data into a second illness severity prediction model to obtain a second illness severity prediction value; Among them, the first severity prediction model and the second severity prediction value are models trained according to sample data and their corresponding labels. The sample data in the first severity prediction model are the first sample feature vector and the weight values ​​of its various data, the second sample feature vector and the weight values ​​of its various data, and the corresponding labels are the severity of the illness, which can be emergency, severe, critical, etc.; the sample data in the second severity prediction value are medical sample imaging data, and the corresponding label is also the severity of the illness.

[0050] S1062: Determine the target user's predicted illness severity value using the first predicted illness severity value and the second predicted illness severity value.

[0051] Specifically, in this embodiment, the first illness severity prediction value and the second illness severity prediction value may be weighted calculated or averaged to obtain the illness severity prediction value of the target user.

[0052] An embodiment of the present invention provides an emergency and critical illness auxiliary assessment method based on big data analysis. First, the medical history data, vital sign data, and induced cause data of the target user are obtained, and then a vital sign data comparison table corresponding to the induced cause data and the medical history data is obtained, and whether the vital sign data of the target user is within the normal data range is determined according to the vital sign data comparison table; if the vital sign data of the target user is not within the normal data range, the medical test data and medical imaging data of the target user are obtained; then, the vital sign data, medical history data, and medical test data of the target user are converted into a first feature vector, and the induced cause data of the target user are converted into a second feature vector; the first feature vector is input into a first prediction model to obtain the weight value corresponding to each data in the first feature vector that affects the change of each data in the second feature vector; the second feature vector is input into a second prediction model to obtain the weight value corresponding to each data in the second feature vector that affects the change of each data in the first feature vector; finally, the predicted value of the severity of the target user is determined by the weight value of the first feature vector and its each data, the weight value of the second feature vector and its each data, and the medical imaging data. Compared with the existing method of relying on the expert experience of the attending physician to judge the emergency and critical illness, the present invention determines the weight value based on the acquisition of various data, and then jointly determines the predicted value of the severity of the target user based on the weight value and data such as the first eigenvector and the second eigenvector. Therefore, the accuracy and efficiency of the emergency and critical illness assessment can be improved through the present invention.

[0053] It should be understood that the order of execution of the steps in the above embodiment does not necessarily 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 embodiment of the present invention.

[0054] In one embodiment, a system for assisting the evaluation of critical and severe diseases based on big data analysis is provided, and the system for assisting the evaluation of critical and severe diseases based on big data analysis corresponds one-to-one to the method for assisting the evaluation of critical and severe diseases based on big data analysis in the above-mentioned embodiment. Figure 2 As shown in the figure, the functional modules of the emergency and critical care auxiliary assessment system based on big data analysis are described in detail as follows: An acquisition module 21 is used to acquire the medical history data, physical sign data, and induced disease cause data of the target user; The acquisition module 21 is further used to obtain a comparison table of physical sign data corresponding to the induced disease cause data and the medical history data; A determination module 22, configured to determine whether the target user's vital sign data is within a normal data range according to the vital sign data comparison table; The acquisition module 21 is further used to acquire the medical examination data and medical imaging data of the target user if the vital sign data of the target user is not within the normal data range; A conversion module 23, used to convert the target user's vital sign data, medical history data and the medical test data into a first feature vector, and convert the target user's induced disease cause data into a second feature vector; Prediction module 24, used for inputting the first feature vector into a first prediction model to obtain weight values ​​corresponding to each data item in the first feature vector that affect changes in each data item in the second feature vector; inputting the second feature vector into a second prediction model to obtain weight values ​​corresponding to each data item in the second feature vector that affect changes in each data item in the first feature vector; The determination module 22 is further configured to determine a predicted value of the severity of illness of the target user through the weight values ​​of the first feature vector and its various data items, the weight values ​​of the second feature vector and its various data items, and the medical imaging data.

[0055] In an optional embodiment provided by the present invention, the determination module 22 is specifically configured to: Determine the disease type based on the causative agent data and the medical history data, and determine a physical sign data comparison table corresponding to the disease type, wherein the physical sign data comparison table includes the normal data range of each physical sign data corresponding to the disease type; The target user's vital sign data is compared with the vital sign data in the vital sign data comparison table. If the target user's vital sign data is not within the normal data range of the corresponding vital sign data in the vital sign data comparison table, it is determined that the target user's vital sign data is not within the normal data range.

[0056] In an optional embodiment provided by the present invention, the determination module 22 is specifically configured to: Extracting entity keywords from the causative disease data and the medical history data, wherein the entity keywords at least include disease name, drug name, and examination result; Determine a connection relationship between entity keywords based on the semantic relationship between the entity keywords, wherein the connection relationship at least includes a cause, a symptom, and a complication; The disease type is determined based on the logical relationship between the entity keywords and the connection relationship.

[0057] In an optional embodiment provided by the present invention, the conversion module 23 is specifically configured to: Performing data preprocessing on the target user's vital sign data, medical history data, medical test data, and induced disease cause data, wherein the data preprocessing at least includes: abnormal data filtering, data standardization, and normalization; Input the target user's physical sign data, medical history data, medical test data, and induced disease cause data that have undergone data preprocessing into the feature extraction model to obtain data features corresponding to the physical sign data, medical history data, medical test data, and induced disease cause data; Combining the target user's vital sign data, medical history data, and data features corresponding to the medical test to obtain the first feature vector; The data feature corresponding to the disease-inducing cause data of the target user is determined as a second feature vector.

[0058] In an optional embodiment provided by the present invention, the emergency and critical illness auxiliary assessment system based on big data analysis further includes a training module 25 for: Determine a first sample feature vector according to the physical sign data, medical history data and medical test data in the sample data, and determine a second sample feature vector according to the induced disease cause data in the corresponding sample data; Using the second sample feature vector as a sample label of the first sample feature vector to perform model training to obtain the first prediction model; The first sample feature vector is used as a sample label of the second sample feature vector to perform model training to obtain the second prediction model.

[0059] In an optional embodiment provided by the present invention, the training module 25 is specifically used for: Inputting the first sample feature vector into the first prediction model to obtain a first prediction weight value of the influence of each data in the first sample feature vector on the second sample feature vector; Determine, based on the data relationship between the data in the preset knowledge base, the influence value of each data in the first sample feature vector on the change of each data in the second sample feature vector; Determine an average value of all influence values ​​corresponding to the same item of data in the first sample feature vector as a first actual weight value; Calculate a first model loss value according to the first predicted weight value and the first actual weight value; When the loss value of the first model is less than a preset value, the training of the first prediction model is stopped.

[0060] In an optional embodiment provided by the present invention, the training module 25 is specifically used for: Inputting the second sample feature vector into the second prediction model to obtain a second prediction weight value of the influence of each data in the second sample data on the first sample data; the initial model parameters of the first prediction model are the model parameters of the second prediction model; Calculating a second actual weight value according to a numerical relationship between the first sample feature vector and the second sample feature vector; Calculate a second model loss value according to the second predicted weight value and the second actual weight value; When the loss value of the second model is less than a preset value, the training of the second prediction model is stopped.

[0061] In an optional embodiment provided by the present invention, the determination module 22 is specifically configured to: Determine a first disease severity prediction value based on the first feature vector and the weight values ​​of its various data items, and the second feature vector and the weight values ​​of its various data items; determine a second disease severity prediction value based on the medical imaging data; The predicted illness severity value of the target user is determined by using the first illness severity predicted value and the second illness severity predicted value.

[0062] In an optional embodiment provided by the present invention, the determination module 22 is specifically configured to: Inputting the first feature vector and the weight value of each item of data thereof, and the second feature vector and the weight value of each item of data thereof into a first illness severity prediction model to obtain a first illness severity prediction value; The medical imaging data is input into a second illness severity prediction model to obtain a second illness severity prediction value.

[0063] For the specific limitations of the emergency and critical care auxiliary assessment system based on big data analysis, please refer to the limitations of the emergency and critical care auxiliary assessment method based on big data analysis above, which will not be repeated here. Each module in the above-mentioned device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0064] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0065] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. An emergency and critical illness auxiliary assessment method based on big data analysis, characterized in that: The method comprises: Obtain the target user's medical history data, physical sign data, and induced disease cause data; Obtaining a physical sign data comparison table corresponding to the induced disease cause data and the medical history data, and determining whether the physical sign data of the target user is within a normal data range according to the physical sign data comparison table; If the target user's vital sign data is not within the normal data range, obtaining the target user's medical examination data and medical imaging data; Converting the target user's vital sign data, medical history data and medical test data into a first feature vector, and converting the target user's disease-inducing cause data into a second feature vector; Input the first feature vector into a first prediction model to obtain weight values ​​corresponding to each item of data in the first feature vector that affect changes in each item of data in the second feature vector; input the second feature vector into a second prediction model to obtain weight values ​​corresponding to each item of data in the second feature vector that affect changes in each item of data in the first feature vector; The predicted value of the severity of the disease of the target user is determined by using the weight value of the first feature vector and its various data items, the weight value of the second feature vector and its various data items, and the medical imaging data.

2. The method according to claim 1, characterized in that The step of obtaining a physical sign data comparison table corresponding to the causal data and the medical history data, and determining whether the physical sign data of the target user is within a normal data range according to the physical sign data comparison table, includes: Determine the disease type based on the causative agent data and the medical history data, and determine a physical sign data comparison table corresponding to the disease type, wherein the physical sign data comparison table includes the normal data range of each physical sign data corresponding to the disease type; The target user's vital sign data is compared with the vital sign data in the vital sign data comparison table. If the target user's vital sign data is not within the normal data range of the corresponding vital sign data in the vital sign data comparison table, it is determined that the target user's vital sign data is not within the normal data range.

3. The method according to claim 2, characterized in that The determining of the disease type based on the causative disease data and the medical history data includes: Extracting entity keywords from the causative disease data and the medical history data, wherein the entity keywords at least include disease name, drug name, and examination result; Determine a connection relationship between entity keywords based on the semantic relationship between the entity keywords, wherein the connection relationship at least includes a cause, a symptom, and a complication; The disease type is determined based on the logical relationship between the entity keywords and the connection relationship.

4. The method according to claim 1, characterized in that The step of converting the target user's vital sign data, medical history data, and medical test data into a first feature vector, and converting the target user's disease-causing data into a second feature vector includes: Performing data preprocessing on the target user's vital sign data, medical history data, medical test data, and induced disease cause data, wherein the data preprocessing at least includes: abnormal data filtering, data standardization, and normalization; Input the target user's physical sign data, medical history data, medical test data, and induced disease cause data that have undergone data preprocessing into the feature extraction model to obtain data features corresponding to the physical sign data, medical history data, medical test data, and induced disease cause data; Combining the target user's vital sign data, medical history data, and data features corresponding to the medical test to obtain the first feature vector; The data feature corresponding to the disease-inducing cause data of the target user is determined as a second feature vector.

5. The method according to claim 4, characterized in that Before inputting the first feature vector into a first prediction model to obtain weight values ​​corresponding to each item of data in the first feature vector that affect changes in each item of data in the second feature vector; and inputting the second feature vector into a second prediction model to obtain weight values ​​corresponding to each item of data in the second feature vector that affect changes in each item of data in the first feature vector, the method further comprises: Determine a first sample feature vector according to the physical sign data, medical history data and medical test data in the sample data, and determine a second sample feature vector according to the induced disease cause data in the corresponding sample data; Using the second sample feature vector as a sample label of the first sample feature vector to perform model training to obtain the first prediction model; The first sample feature vector is used as a sample label of the second sample feature vector to perform model training to obtain the second prediction model.

6. The method according to claim 5, characterized in that The step of using the second sample feature vector as a sample label of the first sample feature vector to perform model training to obtain the first prediction model includes: Inputting the first sample feature vector into the first prediction model to obtain a first prediction weight value of the influence of each data in the first sample feature vector on the second sample feature vector; Determine, based on the data relationship between the data in the preset knowledge base, the influence value of each data in the first sample feature vector on the change of each data in the second sample feature vector; Determine an average value of all influence values ​​corresponding to the same item of data in the first sample feature vector as a first actual weight value; Calculate a first model loss value according to the first predicted weight value and the first actual weight value; When the loss value of the first model is less than a preset value, the training of the first prediction model is stopped.

7. The method according to claim 6, characterized in that The step of using the first sample feature vector as the sample label of the second sample feature vector to perform model training to obtain the second prediction model includes: Inputting the second sample feature vector into the second prediction model to obtain a second prediction weight value of the influence of each data in the second sample data on the first sample data; the initial model parameters of the first prediction model are the model parameters of the second prediction model; Calculating a second actual weight value according to a numerical relationship between the first sample feature vector and the second sample feature vector; Calculate a second model loss value according to the second predicted weight value and the second actual weight value; When the loss value of the second model is less than a preset value, the training of the second prediction model is stopped.

8. The method according to any one of claims 4 to 7, characterized in that: The step of determining the predicted value of the severity of illness of the target user by using the first feature vector and the weight values ​​of its various data items, the second feature vector and the weight values ​​of its various data items, and the medical imaging data includes: Determine a first disease severity prediction value based on the first feature vector and the weight values ​​of its various data items, and the second feature vector and the weight values ​​of its various data items; determine a second disease severity prediction value based on the medical imaging data; The predicted illness severity value of the target user is determined by using the first illness severity predicted value and the second illness severity predicted value.

9. The method according to claim 8, characterized in that The step of determining a first severity prediction value according to the first feature vector and the weight values ​​of its various data items, and the step of determining a second severity prediction value according to the medical imaging data comprises: Inputting the first feature vector and the weight values ​​of its various data items, and the second feature vector and the weight values ​​of its various data items into a first illness severity prediction model to obtain a first illness severity prediction value; The medical imaging data is input into a second illness severity prediction model to obtain a second illness severity prediction value.

10. An emergency and critical illness auxiliary assessment system based on big data analysis, the system implements the method according to claim 1, characterized in that: The system comprises: The acquisition module is used to obtain the target user's medical history data, physical sign data, and induced disease cause data; The acquisition module is further used to obtain a comparison table of physical sign data corresponding to the induced disease cause data and the medical history data; A determination module, used to determine whether the target user's vital sign data is within a normal data range according to the vital sign data comparison table; The acquisition module is further configured to acquire the medical examination data and medical imaging data of the target user if the vital sign data of the target user is not within the normal data range; A conversion module, used to convert the target user's vital sign data, medical history data and the medical test data into a first feature vector, and convert the target user's induced disease cause data into a second feature vector; A prediction module, configured to input the first feature vector into a first prediction model to obtain weight values ​​corresponding to each item of data in the first feature vector that affect changes in each item of data in the second feature vector; and input the second feature vector into a second prediction model to obtain weight values ​​corresponding to each item of data in the second feature vector that affect changes in each item of data in the first feature vector; The determination module is further used to determine the predicted value of the severity of the disease of the target user through the weight value of the first feature vector and its various data, the weight value of the second feature vector and its various data, and the medical imaging data.

Images