AI-based medical auxiliary management method and system

Abstract

The embodiment of the invention discloses an AI-based medical auxiliary management method and system, and the method comprises the steps: obtaining a medical data set, extracting sample numerical data and sample text data, and processing the sample numerical data and the sample text data respectively to obtain a numerical sample set and a text sample set; user information is obtained, the user information comprises user identity information and user medical information, and priorities are divided for users based on the user information; performing model training based on the numerical sample set, the text sample set and the user medical information to obtain an AI medical assistance model; and obtaining user physiological data information and user physical condition description information by using a user terminal, and inputting the information into the AI medical auxiliary model to obtain a user medical analysis report. According to the invention, a personalized and comprehensive medical analysis report is provided for the user based on the AI medical assistance model, and efficient medical assistance is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data processing, and particularly relates to an AI-based medical auxiliary management method and system. BACKGROUND

[0002] Currently, the medical industry is facing many challenges, including uneven distribution of medical resources, heavy workload of medical staff, relatively high misdiagnosis rate and missed diagnosis rate, and other problems. With the aging of the population and the increase in the number of chronic disease patients, these problems have become more prominent. Traditional medical management methods often rely on the experience and intuition of doctors, which to some extent limits the quality and efficiency of medical services. In recent years, artificial intelligence (AI) technology has made significant progress, and its application in the medical field has become increasingly widespread, providing new possibilities and solutions for medical auxiliary management. By analyzing a large amount of medical data, AI can assist doctors in diagnosis, treatment decision-making, and patient management. However, existing AI in medical assistance can only achieve universal functions and cannot provide personalized and comprehensive medical auxiliary management for users. SUMMARY

[0003] Therefore, the main purpose of the present application is to provide an AI-based medical auxiliary management method and system that can comprehensively, accurately, and objectively monitor the health status of users and provide personalized medical auxiliary management for users.

[0004] In a first aspect, an AI-based medical auxiliary management method is provided, which includes the following steps: S1: Obtain a medical data set, extract sample numerical data and sample text data, and process the sample numerical data and the sample text data respectively to obtain a numerical sample set and a text sample set; S2: Obtain user information, including user identity information and user medical information, and divide the user's priority based on the user information; S3: Perform model training based on the numerical sample set, the text sample set, and the user medical information to obtain an AI medical auxiliary model; S4: Obtain user physiological data information and user physical condition description information using a user terminal, input the AI medical auxiliary model, and obtain a user medical analysis report.

[0005] Optionally, in the above-mentioned method embodiments of the present application, the method further includes, The processing of the sample numerical data and the sample text data respectively to obtain the numerical sample set and the text sample set specifically includes: The processing manner of the sample numerical data is data cleaning and data conversion, the data cleaning specifically includes filling missing values, identifying abnormal values and deleting duplicate data, the data conversion is standardization and normalization processing of the data, and finally the standardized numerical data with uniform format is obtained, and then the numerical sample set is obtained. The processing manner of the sample text data is to remove meaningless mood words and stop words in the text data, and then perform word segmentation processing to segment the continuous text data into single words to form a sample word sequence corresponding to each sample text data, and then the text sample set is obtained.

[0006] Optionally, in the above-mentioned method embodiments of the present application, the method further comprises, The model training based on the numerical sample set, the text sample set and the user medical information obtains an AI medical auxiliary model, specifically comprising: S31: processing the user medical data information obtained in S2 to extract text data, and processing the text data to obtain a user word sequence corresponding to the user; S32: calculating the similarity between the user word sequence and each sample word sequence in the text sample set, putting the sample text data corresponding to the sample word sequence with a similarity greater than a preset threshold into a first sample set, and putting the sample text data corresponding to the sample word sequence with a similarity less than or equal to the preset threshold into a second sample set; S33: presetting a generative large language model, using the first sample set to supervise the training of the generative large language model to obtain a first model, using the second sample set to unsupervise the training of the first model to obtain a second model, using cross-validation to optimize the parameters of the second model to obtain a medical auxiliary language model; S34: dividing the numerical sample set according to a preset proportion to obtain a training set and a validation set, using the training set to train a preset neural network model, and using the validation set to test the performance of the neural network model to obtain a medical auxiliary judgment model; S35: setting the AI medical auxiliary model with a multi-branch structure, wherein the first branch is the medical auxiliary language model obtained based on S33, and the second branch is the medical auxiliary judgment model obtained based on S34.

[0007] Optionally, in the above-mentioned method embodiments of the present application, the method further comprises, The user terminal is used to collect user physiological data information and user body condition description information, input the AI medical auxiliary model, and obtain a user medical analysis report, specifically comprising: input the user physiological data information into the medical auxiliary judgment model of the second branch, and obtain the user medical analysis report by combining the output result of the medical auxiliary language model and the output result of the medical auxiliary judgment model.

[0008] Optionally, in the above-mentioned method embodiments of the present application, the method further comprises, The user physiological data information and the user body condition description information obtained by the user terminal specifically include: The user physiological data information is collected by a collection module of the user terminal, and the user body condition description information is input by an input module of the user terminal. When collecting the user physiological data, the collection module collects according to a preset collection period, for a high-priority user, the preset collection period is T1, for a low-priority user, the preset collection period is T2, and T1 < T2.

[0009] In a second aspect, the embodiments of the present application provide an AI-based medical auxiliary management system, which is used to implement the above-mentioned AI-based medical auxiliary management method, and the system comprises: A plurality of user terminals, each user terminal comprising a collection submodule, an input submodule, a transmission submodule and a display submodule, the collection submodule being used to collect user physiological data information, the input submodule being used to input user body condition description information, the transmission submodule being used to upload user medical analysis reports to a regional medical auxiliary module according to a preset frequency, and receive regional medical analysis reports sent by the regional medical auxiliary module, and the display submodule being used to display the user medical analysis reports and the regional medical analysis reports. A plurality of regional medical auxiliary modules, each regional medical auxiliary module comprising a storage submodule, an analysis submodule and a transmission submodule, the storage submodule being used to store user medical analysis reports uploaded by each user terminal, the analysis submodule being used to comprehensively analyze the user medical analysis reports in the storage submodule to obtain regional medical analysis reports, and the transmission submodule being used to receive the user medical analysis reports uploaded by each user terminal, and send the regional medical analysis reports to each user terminal.

[0010] In a third aspect, the embodiments of the present application provide an AI-based medical auxiliary management device, which comprises a memory and a processor, the memory and the processor being coupled, the memory storing program instructions, and the program instructions being executed by the processor to enable the device to perform the above-mentioned AI-based medical auxiliary management method.

[0011] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, characterized in that, comprising a computer program, when the computer program runs on an electronic device, makes the electronic device execute the AI-based medical auxiliary management method.

[0012] The AI-based medical auxiliary management method is adopted, in the model training stage, by acquiring the medical information of the user to join the model training process, so that the model obtained by training is more targeted to the personal health status of the user, thereby realizing personalized medical auxiliary management and improving the performance of the model.

[0013] The AI medical auxiliary model with a multi-branch structure is set, the collected physiological data information of the user is analyzed by using the medical auxiliary judgment model, the input user physical condition description information is analyzed by using the medical auxiliary language model, the objective physiological data information and the subjective physical condition description information of the user are combined, the health status of the user can be more comprehensively and accurately analyzed, and more effective medical advice can be provided for the user.

[0014] The AI-based medical auxiliary management system is adopted, by setting a plurality of regional medical auxiliary modules, the health status of the user in the current region is monitored, the generation of regional diseases can be found in time, and more comprehensive medical advice can be provided for the user.

[0015] The technical solutions of the present application will be described in further detail below with reference to the drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0016] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings in which:

[0017] Figure 1 is a flowchart of the AI-based medical auxiliary management method provided by an exemplary embodiment of the present application.

[0018] Figure 2 is a structural diagram of the AI-based medical auxiliary management system provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0019] Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, and not all embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.

[0020] It should be noted that the relative arrangement, numerical expressions and numerical values of the components and steps set forth in these embodiments do not limit the scope of the present application unless otherwise specifically stated.

[0021] Those skilled in the art can understand that the terms "first", "second", etc. in the embodiments of the present application are only used to distinguish different steps, devices or modules, and do not represent any specific technical meaning, nor do they represent the inevitable logical order between them.

[0022] It should also be understood that in the embodiments of the present application, "a plurality of" can mean two or more, and "at least one" can mean one, two or more.

[0023] It should also be understood that for any component, data or structure mentioned in the embodiments of the present application, it can be understood as one or more in general without explicit limitation or in the context of the opposite indication given by the preceding or following text. Embodiment one

[0024] As shown in Figure 1 The present embodiment provides an AI-based medical auxiliary management method, comprising the following steps: S1: Obtain a medical data set, extract sample numerical data and sample text data, and process the sample numerical data and sample text data respectively to obtain a numerical sample set and a text sample set.

[0025] The medical data set can provide case, symptom and physiological data, etc. for training a medical model, which can be obtained in various ways, such as from public medical data set websites such as MIMIC-III, eICU, UCI machine learning library, etc. It can also be obtained by cooperating with hospitals or research institutions, obtaining desensitized medical data, or applying for the use of government health department data through legal channels.

[0026] The medical data set usually contains numerical data and text data, and the text data usually includes: 1. Patient medical record: including doctor's notes, medical history records, diagnosis reports, etc.

[0027] 2. Laboratory report: describes the results and interpretation of various laboratory tests.

[0028] 3. Prescription information: doctor's drug prescription and medication guidance.

[0029] 4. Imaging report: report of imaging examinations such as X-ray, CT, MRI, etc.

[0030] 5. Surgical record: record of surgical procedure, instruments used, surgical results, etc.

[0031] 6. Patient feedback: The patient's description of the treatment process or health condition.

[0032] Numerical data usually includes: 1. Physiological parameters: such as heart rate, blood pressure, body temperature, blood oxygen saturation, etc.

[0033] 2. Laboratory test results: such as hemoglobin levels, blood glucose levels, cholesterol levels, etc. in blood tests.

[0034] 3. Image data: Although images themselves are image data, they may contain numerical data such as tumor size, location coordinates, etc.

[0035] 4. Drug dosage: The dosage of drugs used by the patient.

[0036] 5. Time series data: For example, electrocardiogram (ECG) and electroencephalogram (EEG) data, which usually record changes over time in numerical form.

[0037] Since the processing methods for text data and numerical data are different, it is necessary to extract text data and numerical data from the medical data set for processing.

[0038] For text data, the usual processing method is to remove meaningless words, stop words, etc. in the text data, and then use a word segmentation tool to segment the text, divide the continuous text into individual words, and form a sample word sequence corresponding to each text, and then obtain a text sample set. In the field, commonly used word segmentation tools include ICTCLAS of the Chinese Academy of Sciences, LTP of Harbin Institute of Technology, and Jieba word segmentation, etc. Chinese word segmentation tools can all achieve word segmentation of preprocessed corpus text, and users can select according to actual needs.

[0039] For numerical data, the usual processing method is data cleaning and data conversion, data cleaning specifically includes filling missing values, identifying outliers, and deleting duplicate data, data conversion is standardization and normalization of data, and finally standardized numerical data with uniform format is obtained, and then a numerical sample set is obtained.

[0040] Other ways of processing text data and numerical data can also be used, as long as a sample set that can be used for model training is obtained, which is not limited in the present application.

[0041] S2: Obtain user information, including user identity information and user medical information, and divide the user's priority based on the user information.

[0042] The user information can include user identity information and user medical information, and the user identity information can specifically include information such as gender, age, occupation, and the like of the user, and the user medical information can specifically include information such as medical records and physical examination reports of the user.

[0043] Due to different health conditions of different users, the demand for medical assistance is also different, for example, people with a history of chronic diseases or past medical history and people with older age or poor physical condition will have more demand for medical assistance than young or healthy people, or users in some high-risk industries will have more demand for medical assistance due to health risks than users in other low-risk industries. Therefore, the user priority is divided by comprehensively considering the user's identity information and medical information, and high-priority users and low-priority users are obtained.

[0044] S3: training a model based on the numerical sample set, the text sample set, and the user medical information to obtain an AI medical assistance model.

[0045] The model training process specifically includes S31: processing the user medical data information obtained in S2 to extract text data, and processing the text data to obtain a user word sequence corresponding to the user.

[0046] The processing manner of the text data is the same as the processing manner of the sample text data described above.

[0047] S32: calculating the similarity between the user word sequence and each sample word sequence in the text sample set, and putting the sample text data corresponding to the sample word sequence with a similarity greater than a preset threshold into a first sample set, and putting the sample text data corresponding to the sample word sequence with a similarity less than or equal to the preset threshold into a second sample set.

[0048] The similarity between the word sequences can be calculated by using the TF-IDF algorithm to calculate the word weight corresponding to each word in each word sequence, and calculating the cosine similarity between any two word sequences based on the word weight, and taking the cosine similarity as the similarity between the two word sequences, that is, the similarity between the two text data.

[0049] S33: presetting a generative large language model, using the first sample set to supervise the training of the generative large language model to obtain a first model, using the second sample set to unsupervise the training of the first model to obtain a second model, and using cross-validation to optimize the parameters of the second model to obtain a medical assistance language model.

[0050] The pre-set generative large language model refers to a general generative large language model trained by a basic training sample. Generally, it only generates a text sequence as a result based on the prediction probability of the language unit according to the knowledge contained in the text input and the large model parameters, and the result is usually inaccurate. The purpose of supervised training is to enable the model to learn the knowledge of a specific task, so as to provide more accurate results in actual application. For generative models, supervised training can also help the model generate text that meets the requirements of a specific field or task. Since the first sample set has high similarity with the user's medical information text data, supervised training of the pre-trained generative large language model using the first sample set can enable the generative large language model to learn medical data that matches the user's health condition, thereby improving the professionalism of the large language model and realizing personalized medical assistance for each user.

[0051] Then the first model is unsupervised trained by using the second sample set to obtain the second model.

[0052] The purpose of unsupervised training is to enable the model to identify patterns in the input data without relying on externally provided labels. In the unsupervised training phase, the model learns from a large amount of unlabeled text data, thereby generating coherent and natural text and being able to understand and generate multiple languages and styles. Since the second sample set has low similarity with the user's medical information text data, unsupervised training of the first model using the second sample set can enable the generative large language model to learn more extensive medical knowledge, thereby improving the comprehensiveness of the large language model.

[0053] Further, the second model is parameter optimized by using cross-validation to obtain a medical assistance language model.

[0054] Cross-validation is a technique for evaluating the performance of machine learning models. It involves dividing the dataset into several mutually exclusive subsets, using different subsets for training and testing to estimate the model's generalization ability. Since large language models are usually trained on very large datasets, standard cross-validation methods may not be applicable or efficient, so the following strategies can be used to optimize the performance of large language models: 1. Stratified Cross-Validation: If the dataset has an unbalanced class distribution, stratified cross-validation can ensure that each fold has representative samples from each class. This is particularly important for classification tasks.

[0055] 2. Time Series Cross-Validation: For time-dependent data such as stock prices or social media data, the dataset needs to be divided in chronological order to avoid future information leakage into the past.

[0056] 3. Continual Learning: Large language models can adopt a continual learning strategy, which involves fine-tuning the model on new data that continuously flows in, to maintain its adaptability to new information.

[0057] 4. Transfer Learning: Transfer learning is applied to large language models, which involves using the model weights trained on a specific task as a starting point and then fine-tuning them on a new task.

[0058] 5. Model Ensembling: Combining the prediction results of multiple models can improve the overall performance of the model. This can be achieved by training multiple models and averaging their outputs when making predictions.

[0059] 6. Hyperparameter Optimization: Methods such as grid search, random search, or Bayesian optimization are used to find the best hyperparameter settings.

[0060] 7. Early Stopping: During training, when the performance on the validation set no longer improves, training is stopped to prevent overfitting.

[0061] In practical applications, various strategies can be combined to optimize the performance of large language models according to actual needs.

[0062] Supervised fine-tuning is derived from the fine-tuning idea in deep learning. First, a large language model is pre-trained to obtain better parameter results on some general data sets. Then, using cross-entropy loss, the model is fine-tuned on the labeled data set of a specific task, so that the model's parameters are adjusted based on the initial parameters to minimize the loss function of the supervised task. Unsupervised fine-tuning is to further train the pre-trained model in the absence of labeled training sets, in order to try to let the model learn the structure and patterns of the data on its own. In unsupervised fine-tuning, the model does not learn from human-labeled data, but through self-supervised tasks such as masked language modeling and next-word prediction, to further understand and process data and language patterns related to specific tasks. This method can improve the performance of the model on specific tasks without relying on a large amount of labeled data, especially suitable for situations where labeled data is scarce or costly.

[0063] S34: Divide the numerical sample set according to the preset proportion to obtain a training set and a validation set, train the preset neural network model using the training set, and test the performance of the neural network model using the validation set to obtain a medical auxiliary judgment model.

[0064] A preset proportion of the numerical sample set can be randomly selected as training data, and the remaining as test data. The preset neural network model can be a common convolutional neural network or a recurrent neural network, etc. The neural network model can be trained and tested using some conventional training and testing methods in the art, which are not limited in the present application.

[0065] Since the numerical sample set contains a large amount of medical data, the medical auxiliary judgment model trained based on the numerical sample set can predict the user's risk of illness.

[0066] S35: Set an AI medical auxiliary model with a multi-branch structure, wherein the first branch is a medical auxiliary language model obtained based on S33, and the second branch is a medical auxiliary judgment model obtained based on S34.

[0067] S4: Obtain user physiological data information and user body condition description information using a user terminal, input the AI medical auxiliary model, and obtain a user medical analysis report.

[0068] The user terminal is provided with a collection module, an input module and a display module.

[0069] The collection module includes a plurality of modules, such as a body temperature collection module, an EMG electromyogram signal collection module, an electrocardiogram signal collection module, a blood pressure collection module, etc. The types of the collection module can be set based on user identity information and user priority. For high-priority users or users with medical history, in addition to setting regular collection modules, some dedicated collection modules such as blood glucose collection modules and blood oxygen collection modules can also be set. The number of collection modules can also be increased or decreased as needed.

[0070] The input module supports text input and voice input. The user can describe his or her current physical condition through text input or voice input.

[0071] The display module is used for visual display of the user medical analysis report.

[0072] After the AI medical auxiliary model obtains the information transmitted by the user terminal, the user body condition description information is input into the medical auxiliary language model of the first branch, and the user physiological data information is input into the medical auxiliary judgment model of the second branch, so as to comprehensively analyze the user's physical condition by combining the subjective description information and the objective physiological data information, and output the user medical analysis report.

[0073] In addition, when collecting physiological data of the user, the collection module collects the physiological data according to a preset collection period. For a user with high priority, the preset collection period is T1, and for a user with low priority, the preset collection period is T2, and T1 < T2. In addition, the user can actively collect physiological data using the collection module. For example, when the user feels unwell, the user can collect physiological data and describe the physical condition at any time, and the AI medical assistance model can analyze the health condition of the user, so that the user can determine whether to seek medical treatment in time according to the medical analysis report of the user. Embodiment Two

[0074] As shown in Figure 2 The present embodiment provides an AI-based medical assistance management system for implementing the AI-based medical assistance management method described above, comprising the following modules: A plurality of user terminals, each user terminal comprising a collection sub-module, an input sub-module, a transmission sub-module and a display sub-module, the collection sub-module being configured to collect physiological data information of a user, the input sub-module being configured to input physical condition description information of the user, the transmission sub-module being configured to upload a medical analysis report of the user to a regional medical assistance module at a preset frequency, and receive a regional medical analysis report sent by the regional medical assistance module, and the display sub-module being configured to display the medical analysis report of the user and the regional medical analysis report.

[0075] A plurality of regional medical assistance modules, each regional medical assistance module comprising a storage sub-module, an analysis sub-module and a transmission sub-module, the storage sub-module being configured to store the medical analysis report of the user uploaded by each user terminal, the analysis sub-module being configured to comprehensively analyze the medical analysis report of the user in the storage sub-module to obtain a regional medical analysis report, and the transmission sub-module being configured to receive the medical analysis report of the user uploaded by each user terminal, and send the regional medical analysis report to each user terminal.

[0076] In addition to being related to the user's own physical condition, the probability of the user's illness may also have regional characteristics. For example, some regional diseases are mainly caused by natural and social environments, such as iodine deficiency, water-borne fluorosis, endemic acute barium poisoning, etc., or are epidemic diseases such as influenza, H1N1, mycoplasma pneumonia, etc. In addition, mass food poisoning and industry diseases also occur in certain specific regions. Therefore, the regional medical assistance module is provided to monitor the health condition of the users in the current region, and if a regional disease is found, medical advice can be provided to the users in the current region, including treatment advice and prevention advice, etc.

[0077] In the setting of the regional medical auxiliary module, it can be divided according to the geographical position, for example, a plurality of regional medical auxiliary modules are set in a city, which are used for monitoring different regions in the city, or it can be divided according to the industry type, for example, one or more regional medical auxiliary modules are set in the fire fighting system, the education system, the environmental sanitation system and the like, and the present application does not make specific limitation, which can be flexibly divided according to the actual demand of the user. Embodiment three

[0078] The embodiment of the present application also provides an AI-based medical auxiliary management device, including a memory and a processor, the memory and the processor are coupled; the memory stores program instructions, the program instructions are executed by the processor, so that the device executes the AI-based medical auxiliary management method. Embodiment four

[0079] The embodiment of the present application also provides a computer readable storage medium, including a computer program, when the computer program runs on an electronic device, so that the electronic device executes the AI-based medical auxiliary management method.

[0080] The basic principles of the present disclosure are described above in combination with specific embodiments, but it should be pointed out that the advantages, advantages, effects and the like mentioned in the present disclosure are only examples and cannot be considered as the must-have of each embodiment of the present disclosure. In addition, the above-mentioned specific details are only for the purpose of example and for the purpose of understanding, and the above-mentioned details do not limit the present disclosure to the above-mentioned specific details.

[0081] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other. For system embodiments, since they basically correspond to method embodiments, they are described more simply, and the relevant parts are referred to the part of the method embodiment.

[0082] The block diagram of the device, apparatus, equipment, system involved in the present disclosure is only an illustrative example and is not intended to require or imply that the connection, arrangement and configuration shown in the block diagram must be connected, arranged and configured. As a person skilled in the art will recognize, these devices, apparatuses, equipment and systems can be connected, arranged and configured in any way. Words such as "include", "contain", "have" and the like are open words, which means "including but not limited to", and can be used interchangeably. The words "or" and "and" used herein mean the word "and / or", and can be used interchangeably unless the context clearly indicates otherwise. The word "such as" used herein means the phrase "such as but not limited to", and can be used interchangeably.

[0083] The methods and apparatus of this disclosure can be implemented in a number of ways. For example, the methods and apparatus of this disclosure can be implemented using software, hardware, firmware, or any combination of these methods. The order of any steps described above is merely exemplary and the steps of the methods of this disclosure need not be performed in the order described unless otherwise specified. Furthermore, in some embodiments, the methods of this disclosure can also be implemented as processor-controlled operations stored in a recording medium, which include machine-readable instructions for implementing the methods according to this disclosure. Thus, this disclosure also covers a recording medium storing a program for executing the methods according to this disclosure.

[0084] It is also important to note that the methods of this disclosure can be implemented by a device, apparatus, or system, or any combination of these, as appropriate. Also, the various methods described herein can be implemented by a computer program product, which can be stored on a recording medium, and which includes machine-readable instructions for implementing the methods according to this disclosure. Thus, this disclosure also covers a recording medium storing a program for executing the methods according to this disclosure.

[0085] The above description is given for the sake of example and description. Additionally, the description is not intended to limit the embodiments of this disclosure to one form or particular form disclosed. Although various example aspects and embodiments have been discussed above, those of ordinary skill in the art will readily appreciate that some modifications, permutations, additions, and sub-combinations of the methods and apparatuses described in this disclosure can be used without departing from the scope of the disclosure.

Claims

1. An AI-based medical assistant management method, characterized by, The method comprises the following steps: S1: obtaining a medical data set, extracting sample numerical data and sample text data, and processing the sample numerical data and the sample text data respectively to obtain a numerical sample set and a text sample set; S2: obtaining user information, the user information including user identity information and user medical information, and dividing the priority of the user based on the user information; S3: training a model based on the numerical sample set, the text sample set, and the user medical information to obtain an AI medical assistance model; S4: obtaining user physiological data information and user physical condition description information using a user terminal, inputting the AI medical assistance model, and obtaining a user medical analysis report.

2. The medical aid management method according to claim 1, wherein The processing of the sample numerical data and the sample text data respectively to obtain the numerical sample set and the text sample set comprises: The processing of the sample numerical data comprises data cleaning and data conversion, the data cleaning specifically comprises filling missing values, identifying abnormal values, and deleting duplicate data, the data conversion is standardization and normalization processing of the data, and finally the standardized numerical data with uniform format is obtained, and then the numerical sample set is obtained; The processing of the sample text data comprises removing meaningless modal particles and stop words in the text data, and then performing word segmentation processing to divide the continuous text data into single words to form a sample word sequence corresponding to each sample text data, and then the text sample set is obtained.

3. The medical aid management method according to claim 2, wherein The model training based on the numerical sample set, the text sample set, and the user medical information to obtain the AI medical assistance model comprises: S31: processing the user medical data information obtained in S2 to extract text data, and processing the text data to obtain a user word sequence corresponding to the user; S32: calculating the similarity between the user word sequence and each sample word sequence in the text sample set, putting the sample text data corresponding to the sample word sequence with a similarity greater than a preset threshold into a first sample set, and putting the sample text data corresponding to the sample word sequence with a similarity less than or equal to the preset threshold into a second sample set; S33: presetting a generative large language model, performing supervised training on the generative large language model using the first sample set to obtain a first model, performing unsupervised training on the first model using the second sample set to obtain a second model, and performing parameter optimization on the second model using cross-validation to obtain a medical assistance language model; S34: dividing the numerical sample set according to a preset proportion to obtain a training set and a validation set, training a preset neural network model using the training set, testing the performance of the neural network model using the validation set, and obtaining a medical assistance judgment model; S35: setting the AI medical assistance model with a multi-branch structure, wherein a first branch is the medical assistance language model obtained based on S33, and a second branch is the medical assistance judgment model obtained based on S34.

4. The medical aid management method according to claim 3, wherein The user terminal collects user physiological data information and user physical condition description information, inputs an AI medical auxiliary model, and obtains a user medical analysis report, specifically including: The user physical condition description information is input into the medical auxiliary language model of the first branch, the user physiological data information is input into the medical auxiliary judgment model of the second branch, and the user medical analysis report is obtained by combining the output results of the medical auxiliary language model and the output results of the medical auxiliary judgment model.

5. The medical aid management method according to claim 4, wherein The user terminal obtains user physiological data information and user physical condition description information, specifically including: The user physiological data information is collected by the collection module of the user terminal, and the user physical condition description information is input by the input module of the user terminal. When collecting the user physiological data, the collection module collects according to a preset collection period, for a high-priority user, the preset collection period is T1, for a low-priority user, the preset collection period is T2, and T1 < T2.

6. An AI-based medical assistance management system for implementing the AI-based medical assistance management method according to claims 1-5, characterized in that, The system includes: A plurality of user terminals, each user terminal including a collection submodule, an input submodule, a transmission submodule, and a display submodule, the collection submodule being configured to collect user physiological data information, the input submodule being configured to input user physical condition description information, the transmission submodule being configured to upload user medical analysis reports to a regional medical auxiliary module according to a preset frequency and receive regional medical analysis reports sent by the regional medical auxiliary module, and the display submodule being configured to display the user medical analysis reports and the regional medical analysis reports; A plurality of regional medical auxiliary modules, each regional medical auxiliary module including a storage submodule, an analysis submodule, and a transmission submodule, the storage submodule being configured to store user medical analysis reports uploaded by each user terminal, the analysis submodule being configured to comprehensively analyze the user medical analysis reports in the storage submodule to obtain regional medical analysis reports, and the transmission submodule being configured to receive user medical analysis reports uploaded by each user terminal and send regional medical analysis reports to each user terminal.

7. An AI-based medical assistant management device, characterized by, The device includes a memory and a processor coupled to the memory, the memory stores program instructions, and the program instructions are executed by the processor to cause the device to perform the method of any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer program, when executed on an electronic device, causes the electronic device to perform the method of any one of claims 1 to 5.

Images