A Medical Information Data Processing System and Method Based on Agent Learning Model for Spondyloarthritis

Abstract

This application discloses a medical information data processing system and method for spondyloarthritis based on an intelligent agent learning model. The system includes: a pre-consultation intelligent agent module for generating an initial medical record in response to the patient's pre-consultation interaction data; a medical data processing intelligent agent module for acquiring the initial medical record and the patient's heterogeneous medical test data, and extracting medical entity information data related to spondyloarthritis from the initial medical record and heterogeneous medical test data using a BERT-BiLSTM-CRF hybrid model, and determining the entity information links formed by the medical entity information data using an EMR-Graph entity relationship graph model; and a medical reasoning intelligent agent module for performing reasoning analysis on the entity information links using an LSTM-TPP model to obtain the initial reasoning probability between the patient and spondyloarthritis. This module can fully consider the link relationship between medical entity information related to spondyloarthritis during the disease probability reasoning process, thereby effectively improving the accuracy of disease probability reasoning.

Description

Technical Field

[0001] This application generally relates to the field of medical artificial intelligence technology, specifically to a medical information data processing system and method for spondyloarthritis based on an agent learning model, and more particularly to a multi-agent collaborative data processing system and method for the diagnosis and treatment of spondyloarthritis. Background Technology

[0002] Axial spondyloarthritis (axSpA) is a chronic inflammatory arthritis characterized primarily by arthritis of the axial spine.

[0003] With the development of medical platforms, technologies have emerged that can provide remote interactive medical platforms for axSpA patients. However, given the heterogeneity and unstructured nature of medical data, these technologies typically perform inference analysis only on a specific type of medical data. For example, they may perform in-depth analysis of MRI or CT scans of a patient's sacroiliac joints or spine to achieve relatively accurate probabilistic inference. However, while probabilistic inference based on single-type medical data has high accuracy, it lacks correlation analysis between multiple types of medical data, which can easily lead to misdiagnosis of similar diseases. Summary of the Invention

[0004] In view of the above-mentioned defects or deficiencies in the existing technology, it is desirable to provide a medical information data processing system and method for spondyloarthritis based on an intelligent agent learning model, which can fully consider the link relationship between medical entity information related to spondyloarthritis during the process of disease probability reasoning, thereby effectively improving the accuracy of disease probability reasoning.

[0005] In a first aspect, embodiments of this application provide a medical information data processing system for spondylitis based on an intelligent agent learning model, comprising:

[0006] The pre-consultation intelligent agent module is used to generate initial medical records in response to patients' pre-consultation interaction data;

[0007] The medical data processing intelligent agent module is used to acquire the initial medical record and the patient's heterogeneous medical test data, and to extract medical entity information data related to spondyloarthritis from the initial medical record and the heterogeneous medical test data using the BERT-BiLSTM-CRF hybrid model, and to determine the entity information links formed by the medical entity information data using the EMR-Graph entity relationship graph model.

[0008] The medical reasoning intelligent agent module is used to perform reasoning analysis on the entity information link using the LSTM-TPP model to obtain the initial reasoning probability between the patient and spondyloarthritis.

[0009] In some embodiments, the medical data processing intelligent agent module is further configured to:

[0010] The medical entity information data related to spondyloarthritis extracted from the initial medical record and the heterogeneous medical test data are respectively vectorized to obtain the original entity representation of each medical entity information data.

[0011] Based on multiple preset entity relationships, the original entity representations of each medical entity information data are projected onto the corresponding relationship space;

[0012] The entity information link is generated by generating at least three medical entity information data based on the entity relationships between the medical entity information data.

[0013] In some embodiments, the medical reasoning agent module is further configured to:

[0014] Based on the entity information link formed by the medical entity information data, the patient's symptom evolution is predicted to obtain the probability of typical symptom evolution of spondyloarthritis.

[0015] Based on the evolution probability of typical symptoms of spondyloarthritis, the initial inference probability between the patient and spondyloarthritis is determined.

[0016] In some embodiments, the typical symptoms of spondyloarthritis include multiple symptoms, and the medical reasoning agent module is further configured to:

[0017] Based on the entity information link formed by the medical entity information data, the evolution prediction of each typical symptom is performed to obtain the symptom prediction probability corresponding to each typical symptom.

[0018] For each typical symptom, when the predicted probability of the corresponding symptom is greater than or equal to a preset threshold, the evolution probability of the typical symptom is determined as a first probability value.

[0019] Based on the sum of the evolution probabilities of multiple typical symptoms, the initial inference probability between the patient and spondyloarthritis is determined.

[0020] In some embodiments, the medical reasoning agent module is further configured to:

[0021] When the sum of the evolution probabilities of the typical symptoms is greater than or equal to the inference threshold, the initial inference probability is determined as the third probability value; when the sum of the evolution probabilities of the typical symptoms is less than the inference threshold, the initial inference probability is determined as the fourth probability value.

[0022] In some embodiments, the medical reasoning agent module is further configured to:

[0023] The target inference probability, obtained by correcting the initial inference probability by the doctor, is used to optimize the LSTM-TPP model.

[0024] In some embodiments, the medical reasoning agent module is further configured to:

[0025] Based on the entity information links formed by the medical entity information data and the target inference probability, updated questionnaire information is generated;

[0026] The pre-diagnosis intelligent agent module is also used for:

[0027] The updated questionnaire information is provided to the patient, and an initial medical record corresponding to the updated questionnaire information is generated in response to the patient's pre-consultation interaction data.

[0028] Secondly, embodiments of this application provide a method for processing medical information data related to spondylitis based on an intelligent agent learning model, including:

[0029] Initial medical records are generated in response to patient pre-consultation interaction data;

[0030] The initial medical record and the patient's heterogeneous medical testing data were obtained, and the BERT-BiLSTM-CRF hybrid model was used to extract medical entity information data related to spondyloarthritis from the initial medical record and the heterogeneous medical testing data.

[0031] The entity information links formed by the medical entity information data are determined using the EMR-Graph entity relationship graph model;

[0032] The LSTM-TPP model was used to perform inference analysis on the entity information link to obtain the initial inference probability between the patient and spondyloarthritis.

[0033] In some embodiments, it also includes:

[0034] The target inference probability, obtained by correcting the initial inference probability by the doctor, is used to optimize the LSTM-TPP model.

[0035] In some embodiments, it also includes:

[0036] Based on the initial inference probability, updated questionnaire information is generated; and

[0037] The updated questionnaire information is provided to the patient, and an initial medical record corresponding to the updated questionnaire information is generated in response to the patient's pre-consultation interaction data.

[0038] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in embodiments of this application.

[0039] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in embodiments of this application.

[0040] Fifthly, embodiments of this application provide a computer program product, including a computer program, characterized in that, when the computer program is executed by a processor, it implements the method described in embodiments of this application.

[0041] This application provides a medical information data processing system and method for spondylitis based on an intelligent agent learning model. By constructing an EMR-Graph entity relationship graph model, medical information data related to spondylitis is embedded into the EMR-Graph entity relationship graph model. Entity information links are generated based on the pairwise entity relationships between medical entity information data. This effectively overcomes the limitation on the number of entities expressed using triples or quadruples in traditional medical graphs, allowing the entity information links for subsequent reasoning analysis to include more medical entity information data and relationship data, thus improving the accuracy of subsequent reasoning analysis. Simultaneously, it enriches the types of entity data and relationship types carried by the EMR-Graph entity relationship graph model, providing possibilities for its application in various medical analysis scenarios.

[0042] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0043] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0044] Figure 1 A flowchart illustrating a medical information data processing method for spondyloarthritis based on an intelligent agent learning model, according to an embodiment of this application, is shown.

[0045] Figure 2 A schematic diagram of a medical information data processing method for spondyloarthritis based on an agent learning model provided in an embodiment of this application is shown.

[0046] Figure 3 A schematic diagram of a four-node entity information link provided in an embodiment of this application is shown;

[0047] Figure 4 This paper illustrates a schematic diagram of the structure of a medical information data processing system for spondyloarthritis based on an agent learning model, according to an embodiment of this application. Detailed Implementation

[0048] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0049] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0050] To further illustrate the technical solutions provided in the embodiments of this application, a detailed description is provided below in conjunction with the accompanying drawings and specific implementation methods. Although the embodiments of this application provide method operation instruction steps as shown in the following embodiments or drawings, the method may include more or fewer operation instruction steps based on conventional or non-creative effort. In steps where there is no logically necessary causal relationship, the execution order of these steps is not limited to the execution order provided in the embodiments of this application. In actual processing or when the device executes the method, it may be executed sequentially or in parallel according to the method shown in the embodiments or drawings.

[0051] It should be noted that the acquisition or use of data in the embodiments of this application requires the user's consent. The relevant data can only be obtained after the user's authorization, and the acquisition or use of the data complies with the provisions of relevant laws and regulations.

[0052] Please refer to Figure 1 , Figure 1 This illustration shows a flowchart of a medical information data processing method for spondylitis based on an intelligent agent learning model, according to an embodiment of this application. Figure 1 As shown, the method includes:

[0053] Step 101: Generate an initial medical record in response to the patient's pre-consultation interaction data.

[0054] It should be noted that the pre-consultation interaction data consists of the questions and responses provided by the patient based on pre-set questionnaire information. Optionally, the pre-set questionnaire information can be questionnaire information with fixed questions, or it can be questionnaire information generated based on the target inference probability obtained by the doctor after subsequent revision; this application does not make any specific limitations.

[0055] Optionally, the pre-consultation interaction data includes, but is not limited to, patient data, symptom data, and medication data. Patient data may include the patient's basic information and past medical history; symptom data may be user-perceived data, such as pain intensity, duration of pain, and duration of morning stiffness; medication data may be medication data determined by the user based on the doctor's prior diagnosis or medication data for other diseases, etc., without specific limitations in this application.

[0056] In one feasible embodiment, such as Figure 2 As shown, after obtaining the pre-consultation interaction data submitted by the patient in response to the questionnaire information, the pre-consultation intelligent agent module generates an initial medical record based on the pre-consultation interaction data according to the preset medical record generation rules.

[0057] Optionally, an initial medical record can be generated based on the mapping relationship between question data and response data in the pre-consultation interaction data. For example, the question data in the pre-consultation interaction data includes question entries such as patient name, gender, age, and past medical history, while the response data includes the response information for the corresponding question entries, such as name, gender, age, and no past medical history. The initial medical record then includes the question data and its corresponding response data from the pre-consultation interaction data, presented according to a preset positional relationship.

[0058] Step 102: Obtain the initial medical records and heterogeneous medical test data of the patient, and use the BERT-BiLSTM-CRF hybrid model to extract medical entity information data related to spondyloarthritis from the initial medical records and heterogeneous medical test data.

[0059] It should be noted that heterogeneous medical testing data refers to different types of medical testing data, including but not limited to blood test data and image test data. For example, ... Figure 2 As shown, heterogeneous medical testing data can be obtained from medical tests performed on patients after a pre-consultation interaction. The medical tests performed by the patient can be fixed tests, such as blood tests and image analysis, or they can include relevant tests recommended by the pre-consultation intelligent agent module based on the initial medical record.

[0060] In one feasible embodiment, before extracting medical entity information data related to spondyloarthritis from initial medical records and heterogeneous medical testing data using the BERT-BiLSTM-CRF hybrid model, the method further includes: constructing an initial BERT-BiLSTM-CRF hybrid model, training the initial BERT-BiLSTM-CRF hybrid model using labeled samples related to spondyloarthritis, and obtaining a BERT-BiLSTM-CRF hybrid model for extracting medical entity information data related to spondyloarthritis from initial medical records and heterogeneous medical testing data.

[0061] The BERT-BiLSTM-CRF hybrid model comprises an embedding layer (BERT, Bidirectional Encoder Representation from Transformers), a bidirectional long short-term memory network (BiLSTM), and a conditional random field (CRF). The BERT embedding layer transforms text data into word embedding vectors, the BiLSTM network represents the hidden states of word relationships, and the CRF predicts medical entity information related to spondyloarthritis in the input data.

[0062] In a feasible embodiment, the CRF decoding layer employs the following prediction function:

[0063]

[0064] Where y = (y1,...,y L ) represents the predicted label sequence, X represents the input data sequence, such as initial medical records and text-based medical test data, and W represents the input data sequence. crf Let h be the label weight matrix. t For the hidden states of word context relationships obtained based on the BiLSTM context encoding layer, b crf For the label bias vector, Trans(y) t-1 ,y t () represents the label transfer score.

[0065] Therefore, this application utilizes a trained BERT-BiLSTM-CRF hybrid model to effectively extract medical entity information data of patients from initial medical records and heterogeneous medical testing data, such as medical entity information data of patients describing symptoms (morning stiffness, pain, etc.) in initial medical records, numerical information of blood tests in heterogeneous medical testing data, or test descriptions provided by doctors for image detection, etc. This application does not make any specific limitations.

[0066] In a preferred embodiment, this application also utilizes other existing analysis models to extract medical entity information data from heterogeneous medical testing data. For example, it uses image analysis models to extract information data such as soft tissue changes and bone changes in patient images, so as to use multi-channel and multi-type medical entity information data for subsequent entity relationship graph construction, thereby improving the comprehensiveness and reliability of medical entity information data in the patient-related entity relationship graph.

[0067] Step 103: Use the EMR-Graph entity relationship graph model to determine the entity information links formed by medical entity information data.

[0068] It should be noted that the EMR-Graph entity relationship graph model is a medical knowledge graph for spondylitis built based on medical information data related to spondylitis. It is trained by a hybrid method combining the Electronic Medical Record (EMR) system and the Graph knowledge graph. The EMR-Graph entity relationship graph model consists of nodes composed of medical information data entities related to spondylitis, including but not limited to direct medical entity information data such as patients, symptoms, examinations, behaviors, medications, or attribute information, as well as indirect medical entity information data. The edges between nodes are formed by the relationships between these entities, such as the manifestation relationship between patients and symptoms, or the usage relationship between patients and medications.

[0069] In a feasible embodiment, the medical entity information data related to spondyloarthritis extracted from the initial medical records and heterogeneous medical test data are respectively vectorized to obtain the original entity representation of each medical entity information data. Based on a number of preset entity relationships, the original entity representation of each medical entity information data is projected onto the corresponding relationship space, and at least three medical entity information data are used to generate entity information links according to the entity relationships between the medical entity information data.

[0070] For example, such as Figure 3 As shown, the EMR-Graph entity relationship model presents a four-node entity information link related to a patient. The four nodes include the patient, drug A, morning stiffness, and nocturnal pain. The relationship between the patient and drug A is one of usage; the relationship between the patient and morning stiffness is one of performance; the relationship between drug A and morning stiffness is one of efficacy (relief); and the relationship between morning stiffness and nocturnal pain is one of association.

[0071] Therefore, this embodiment of the application constructs an EMR-Graph entity relationship graph model, embedding medical-like information data related to spondyloarthritis into the EMR-Graph entity relationship graph model, and generating entity information links based on the pairwise entity relationships between medical entity information data. This effectively overcomes the limitation on the number of entities expressed using triples or quadruples in traditional medical graphs, allowing the entity information links for subsequent reasoning analysis to include more medical entity information data and relationship data, thus improving the accuracy of subsequent reasoning analysis. Simultaneously, it enriches the entity data types and relationship types carried by the EMR-Graph entity relationship graph model, providing possibilities for applying the EMR-Graph entity relationship graph model to various medical analysis scenarios.

[0072] In a preferred embodiment, such as Figure 2The medical data processing intelligent agent module shown can extract medical entity information data related to spondyloarthritis, and can also extract the corresponding time nodes of the medical entity information data. For example, it can extract the time when the patient experiences morning stiffness, the time when the patient describes morning stiffness in the pre-consultation interaction data, the time of medication use, and the corresponding symptom relief time. When the medical entity information data includes time information, the EMR-Graph entity relationship graph model can further form entity information links based on the time relationships of the medical entity information data. That is, the entity information links can further represent the time relationships between the patient's medical entity information data, so that when the LSTM-TPP model is used for inference analysis of the entity information links, it can effectively extrapolate the patient's symptom evolution over time, providing more dimensions of data information for subsequent symptom evolution prediction and inference analysis.

[0073] Step 104: Use the LSTM-TPP model to perform inference analysis on the entity information link to obtain the initial inference probability between the patient and spondyloarthritis.

[0074] The LSTM-TPP model is a hybrid model of Long Short-Term Memory (LSTM) network and Temporal Point Processes (TPP). It uses the LSTM model to predict future symptom occurrence events by analyzing medical entity information data and temporal relationships in entity information links, and combines the TPP model to analyze discrete symptom occurrence events and frequencies, thereby determining the evolution probability of typical symptoms of spondyloarthritis.

[0075] In one feasible embodiment, the evolution of symptoms of patients can be predicted based on the information link formed by medical entity information data, the evolution probability of typical symptoms of spondyloarthritis can be obtained, and the initial inference probability between the patient and spondyloarthritis can be determined based on the typical evolution probability of spondyloarthritis.

[0076] In other words, in this embodiment of the application, the LSTM-TPP model is used to predict the future symptom evolution of the patient based on the information link formed by the patient's historical medical entity information data, and the evolution probability of typical symptoms of spondyloarthritis is obtained. That is, the probability prediction of the patient developing typical symptoms of spondyloarthritis based on the natural development of the information link formed by the current medical entity information data. Then, based on the probability of the patient developing typical symptoms of spondyloarthritis, the initial inference probability between the patient and spondyloarthritis is determined.

[0077] Optionally, the initial probability between a patient and spondyloarthritis is a binary probability value. For example, when the initial inferred probability between a patient and the probability of spondyloarthritis is 1, it means that the patient has a high probability of developing spondyloarthritis in the future. When the initial inferred probability between a patient and the probability of spondyloarthritis is 0, it means that although the patient has similar symptoms, he or she is unlikely to develop spondyloarthritis.

[0078] In one feasible embodiment, based on the information link formed by medical entity information data, the patient's symptom evolution is predicted to obtain the evolution probability of typical symptoms of spondyloarthritis. This includes: based on the entity information link formed by medical entity information data, the evolution prediction is performed on each typical symptom to obtain the symptom prediction probability corresponding to each typical symptom; for each typical symptom, when its corresponding symptom prediction probability is greater than a preset threshold, the evolution probability of the typical symptom is determined as a first probability value; based on the sum of the evolution probabilities of multiple typical symptoms, the initial inference probability between the patient and spondyloarthritis is determined.

[0079] For example, the evolution of each typical symptom can be predicted using the following formula, specifically expressed as follows:

[0080]

[0081] Where, λ s (t|H0) represents the predicted evolution of symptom s within the symptom prediction time window t, formed by the entity information link based on medical entity information data. s represents typical symptoms, such as morning stiffness or sacroiliac joint pain. K represents the number of clinical stages of symptom development (which can be determined based on the typical course of spondyloarthritis; for example, in sacroiliac joint pain, k=1 represents bone marrow edema, k=2 represents cartilage destruction, and k=3 represents bony ankylosis). s,k μ represents the tendency of symptom s to progress in stage k. s,k σ represents the typical progression time of symptom s in stage k. s,k The individual variability in symptom progression time (e.g., genetic heterogeneity, differences in treatment response, or comorbidity effects), t is the predicted time span from the present to the onset of symptoms, such as 90 days, and H0 is the entity information link formed by medical entity information data.

[0082] It should be noted that w s,k Based on the symptom-specific weight matrix W s The prediction weights h0 determined by the LSTM model are used to obtain the following, which can be specifically expressed as:

[0083] w s,k =W s ·h0+b s .

[0084] Among them, b sThis is the weight bias term for symptom s.

[0085] Furthermore, the symptom prediction probability corresponding to each typical symptom can be calculated using the following formula:

[0086]

[0087] Among them, P s Let λ be the symptom prediction probability corresponding to symptom s. s (t|H0) represents the predicted evolution of symptom s within the symptom prediction time window t, based on the entity information link formed from medical entity information data.

[0088] Specifically, based on the entity information links formed by medical entity information data, the evolution prediction of each typical symptom can be performed to obtain the symptom prediction probability corresponding to each typical symptom. For example, the symptom prediction probability corresponding to morning stiffness, sacroiliac joint pain, enthesitis, etc., are all typical symptom prediction probabilities. Then, it is determined whether the symptom prediction probability corresponding to each typical symptom is greater than or equal to a preset threshold. If it is greater than or equal to, it means that the patient will develop the symptom in the future; if it is less than, it means that the patient will not develop the symptom in the future. The evolution probability corresponding to the typical symptom with a prediction probability greater than or equal to the preset threshold is set as the first probability value, for example, 1. It can be understood that other prediction probabilities less than the preset threshold are set as the first probability value. The evolution probability corresponding to a typical symptom with a preset threshold is set as a second probability value, such as 0. Then, the sum of the evolution probabilities of multiple typical symptoms is calculated, and the initial inference probability between the patient and spondyloarthritis is determined based on the sum of the evolution probabilities. For example, when the sum of the evolution probabilities is greater than or equal to the inference threshold, it is determined that the patient may have multiple typical symptoms at the same time. At this time, the initial inference probability between the patient and spondyloarthritis is determined as a third probability value, such as 1, that is, the patient is more likely to have spondyloarthritis. When the sum of the evolution probabilities is less than the inference threshold, it is determined that the patient may have typical symptoms but only similar to the symptoms of spondyloarthritis. At this time, the initial inference probability between the patient and spondyloarthritis is determined as a fourth probability value, such as 0, that is, the patient is more likely not to have spondyloarthritis.

[0089] In some embodiments, the target inference probability is further optimized in response to the doctor's correction of the initial probability.

[0090] In other words, in this embodiment, the prediction obtained by the medical inference agent module is an initial inference probability. This initial inference probability needs to be corrected by the doctor to obtain the final target inference probability. In other words, the initial inference probability is only used to assist the doctor's diagnosis; the final diagnosis must be made by the doctor based on the initial inference probability. Then, after obtaining the target inference probability obtained by the doctor's correction of the initial inference probability, the LSTM-TPP model is further optimized through feedback to improve the reliability of predicting the evolution of subsequent typical symptoms, thereby improving the reliability of the initial inference probability.

[0091] In one feasible embodiment, updated questionnaire information can be generated based on the entity information links and target inference probabilities formed by medical entity information data, and the updated questionnaire information can be provided to the patient. In response to the patient's pre-consultation interaction data, an initial medical record corresponding to the updated questionnaire information can be generated.

[0092] In other words, after each final diagnosis by the doctor and the resulting target inference probability, the medical inference agent module can use the LSTM-TPP model to predict the evolution of subsequent symptoms based on the disease progression stage and the entity information links formed by the already generated medical entity information data. It can also generate questionnaire information related to the subsequent symptoms, i.e., generate updated questionnaire information, and provide updated questionnaire information to the patient. This responds to the patient's pre-consultation interaction data to generate an initial medical record corresponding to the updated questionnaire information, thereby realizing an iterative process of generating a new round of medical records and initial inference based on the patient's new pre-consultation interaction data.

[0093] Therefore, the medical information data processing method for spondyloarthritis based on an intelligent agent learning model provided in this application extracts medical entity information data related to spondyloarthritis from initial medical records and heterogeneous medical test data using a BERT-BiLSTM-CRF hybrid model, improving the accuracy of obtaining medical entity information data. The EMR-Graph entity relationship graph model is used to determine the entity information links formed by the medical entity information data, effectively enhancing the correlation between entity information. This provides entity information data with a longer time span and more types for subsequent inference analysis of entity information links using the LSTM-TPP model, improving the reliability of the LSTM-TPP model's prediction results. Simultaneously, the LSTM-TPP model fully considers the irregular time interval characteristics and multimodal nature of disease symptoms, and uses an improved log-Gaussian mixture model to fit complex time distributions, achieving reliable prediction of the evolution of typical symptoms.

[0094] It should be noted that although the operation of the method of the present invention is described in a specific order in the accompanying drawings, this does not require or imply that the operations must be performed in that specific order, or that all the operations shown must be performed in order to achieve the desired result.

[0095] Figure 3 This illustration shows a schematic diagram of a medical information data processing system for spondylitis based on an intelligent agent learning model, according to an embodiment of this application. Figure 3 As shown, the medical information data processing system 10 for spondylitis based on an intelligent agent learning model includes:

[0096] The pre-consultation intelligent agent module 11 is used to generate an initial medical record in response to the patient's pre-consultation interaction data;

[0097] The medical data processing intelligent agent module 12 is used to acquire the initial medical record and the heterogeneous medical test data of the patient, and to extract medical entity information data related to spondyloarthritis from the initial medical record and the heterogeneous medical test data using the BERT-BiLSTM-CRF hybrid model, and to determine the entity information links formed by the medical entity information data using the EMR-Graph entity relationship graph model.

[0098] The medical reasoning agent module 13 is used to perform reasoning analysis on the entity information link using the LSTM-TPP model to obtain the initial reasoning probability between the patient and spondyloarthritis.

[0099] In some embodiments, the medical data processing intelligent agent module 12 is further configured to:

[0100] The medical entity information data related to spondyloarthritis extracted from the initial medical record and the heterogeneous medical test data are respectively vectorized to obtain the original entity representation of each medical entity information data.

[0101] Based on multiple preset entity relationships, the original entity representations of each medical entity information data are projected onto the corresponding relationship space;

[0102] The entity information link is generated by generating at least three medical entity information data based on the entity relationships between the medical entity information data.

[0103] In some embodiments, the medical reasoning agent module 13 is further configured to:

[0104] Based on the entity information link formed by the medical entity information data, the patient's symptom evolution is predicted to obtain the probability of typical symptom evolution of spondyloarthritis.

[0105] Based on the evolution probability of typical symptoms of spondyloarthritis, the initial inference probability between the patient and spondyloarthritis is determined.

[0106] In some embodiments, the typical symptoms of spondyloarthritis include multiple symptoms, and the medical reasoning agent module 13 is further used for:

[0107] Based on the entity information link formed by the medical entity information data, the evolution prediction of each typical symptom is performed to obtain the symptom prediction probability corresponding to each typical symptom.

[0108] For each typical symptom, when the predicted probability of the corresponding symptom is greater than or equal to a preset threshold, the evolution probability of the typical symptom is determined as a first probability value.

[0109] Based on the sum of the evolution probabilities of multiple typical symptoms, the initial inference probability between the patient and spondyloarthritis is determined.

[0110] In some embodiments, the medical reasoning agent module 13 is further configured to:

[0111] When the sum of the evolution probabilities of the typical symptoms is greater than or equal to the inference threshold, the initial inference probability is determined as the third probability value; when the sum of the evolution probabilities of the typical symptoms is less than the inference threshold, the initial inference probability is determined as the fourth probability value.

[0112] In some embodiments, the medical reasoning agent module 13 is further configured to:

[0113] The target inference probability, obtained by correcting the initial inference probability by the doctor, is used to optimize the LSTM-TPP model.

[0114] In some embodiments, the medical reasoning agent module 13 is further configured to:

[0115] Based on the entity information links formed by the medical entity information data and the target inference probability, updated questionnaire information is generated;

[0116] The pre-diagnosis intelligent agent module 11 is also used for:

[0117] The updated questionnaire information is provided to the patient, and an initial medical record corresponding to the updated questionnaire information is generated in response to the patient's pre-consultation interaction data.

[0118] It should be understood that the modules or modules described in the medical information data processing system 10 for spondyloarthritis based on an intelligent agent learning model are similar to those in the reference system. Figure 1The steps in the described method correspond accordingly. Therefore, the operations and features described above for the method are also applicable to the spondyloarthritis-related medical information data processing system 10 based on an intelligent agent learning model and its included modules, and will not be repeated here. The spondyloarthritis-related medical information data processing system 10 based on an intelligent agent learning model can be pre-implemented in the browser or other secure applications of an electronic device, or it can be loaded into the browser or other secure applications of an electronic device through download or other means. The corresponding modules in the spondyloarthritis-related medical information data processing system 10 based on an intelligent agent learning model can cooperate with the modules in the electronic device to implement the solution of the embodiments of this application.

[0119] The division of modules or units mentioned in the detailed description above is not mandatory. In fact, according to the embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0120] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operational instructions of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two connected blocks may actually be executed substantially in parallel, or they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified functions or operational instructions, or using a combination of dedicated hardware and computer instructions.

[0121] In another aspect, this application also provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments, or may exist independently and not assembled into the electronic device. The aforementioned computer-readable storage medium stores one or more programs, which are used by one or more processors to execute the spinal arthritis-related medical information data processing method of this application based on an intelligent agent learning model.

[0122] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the foregoing disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A spondyloarthritis medical information data processing system based on an agent learning model, characterized in that the system The method comprises the following steps: a pre-consultation intelligent agent module is used to generate an initial medical record in response to pre-consultation interaction data of a patient; a medical data processing intelligent agent module is used to obtain the initial medical record and heterogeneous medical detection data of the patient, extract medical entity information data related to spondyloarthritis from the initial medical record and the heterogeneous medical detection data by using a BERT-BiLSTM-CRF hybrid model, and determine an entity information link formed by the medical entity information data by using an EMR-Graph entity relationship graph model; a medical reasoning intelligent agent module is used to perform reasoning analysis on the entity information link by using an LSTM-TPP model to obtain an initial reasoning probability between the patient and spondyloarthritis. The medical data processing intelligent agent module is further used to: vectorize the medical entity information data related to spondyloarthritis extracted from the initial medical record and the heterogeneous medical detection data respectively to obtain original entity representations of each medical entity information data; project the original entity representations of each medical entity information data to a corresponding relationship space based on a plurality of preset entity relationships; generate the entity information link according to the entity relationships between at least three medical entity information data; The medical reasoning intelligent agent module is further used to: perform symptom evolution prediction on the patient based on the entity information link formed by the medical entity information data to obtain a typical symptom evolution probability of spondyloarthritis; determine the initial reasoning probability between the patient and spondyloarthritis based on the typical symptom evolution probability of spondyloarthritis.

2. The spondyloarthritis medical information data processing system based on an agent learning model according to claim 1, characterized by, The typical symptoms of spondyloarthritis include a plurality of typical symptoms, and the medical reasoning intelligent agent module is further used to: perform evolution prediction on each typical symptom based on the entity information link formed by the medical entity information data to obtain a symptom prediction probability corresponding to each typical symptom; for each typical symptom, when the corresponding symptom prediction probability is greater than or equal to a preset threshold value, determine that the evolution probability of the typical symptom is a first probability value; determine the initial reasoning probability between the patient and spondyloarthritis based on the sum of the evolution probabilities of a plurality of typical symptoms. 3.The spondyloarthritis medical information data processing system based on an agent learning model according to claim 2, wherein, The medical reasoning intelligent agent module is further used to: when the sum of the evolution probabilities of the typical symptoms is greater than or equal to a reasoning threshold value, determine that the initial reasoning probability is a third probability value, and when the sum of the evolution probabilities of the typical symptoms is less than the reasoning threshold value, determine that the initial reasoning probability is a fourth probability value. 4.The spondyloarthritis medical information data processing system based on an agent learning model according to claim 1, wherein, The medical reasoning intelligent agent module is further used to: perform feedback optimization on the LSTM-TPP model in response to a target reasoning probability obtained by a doctor modifying the initial reasoning probability.

5. The spondyloarthritis medical information data processing system based on an agent learning model according to claim 4, characterized in that, The medical reasoning intelligent agent module is further used to: generate updated questionnaire information based on the entity information link formed by the medical entity information data and the target reasoning probability; The pre-consultation intelligent agent module is further used to: provide the updated questionnaire information for the patient, and generate an initial medical record corresponding to the updated questionnaire information in response to pre-consultation interaction data of the patient. 6.A method for processing medical information data on spondyloarthritis based on an agent learning model, the method comprising: The method comprises the following steps: generating an initial medical record in response to pre-consultation interaction data of a patient; obtaining the initial medical record and heterogeneous medical detection data of the patient, and extracting medical entity information data related to spondyloarthritis from the initial medical record and the heterogeneous medical detection data by using a BERT-BiLSTM-CRF hybrid model; determining an entity information link formed by the medical entity information data by using an EMR-Graph entity relationship graph model; performing reasoning analysis on the entity information link by using an LSTM-TPP model to obtain an initial reasoning probability between the patient and spondyloarthritis; and respectively performing vector representation on the medical entity information data related to spondyloarthritis extracted from the initial medical record and the heterogeneous medical detection data to obtain original entity representation of each medical entity information data; projecting the original entity representation of each medical entity information data to a corresponding relationship space based on a plurality of preset entity relationships; generating the entity information link according to the entity relationship between at least three medical entity information data; performing symptom evolution prediction on the patient based on the entity information link formed by the medical entity information data to obtain a typical symptom evolution probability of spondyloarthritis; determining an initial reasoning probability between the patient and spondyloarthritis based on the typical symptom evolution probability of spondyloarthritis.

7. The spondyloarthritis medical information data processing method based on an agent learning model according to claim 6, characterized in that, Further comprising: performing feedback optimization on the LSTM-TPP model in response to a target reasoning probability obtained by a doctor correcting the initial reasoning probability. 8.The spondyloarthritis medical information data processing method based on the agent learning model according to claim 7, wherein, Further comprising: generating updated questionnaire information based on the initial reasoning probability; and providing the updated questionnaire information for the patient, and generating an initial medical record corresponding to the updated questionnaire information in response to pre-consultation interaction data of the patient.

Images