Medical care data intelligent retrieval and evaluation system and method driven by artificial intelligence

Abstract

The invention discloses an artificial intelligence-driven medical care data intelligent retrieval and evaluation system and method, and relates to the technical field of artificial intelligence and medical information processing. Artificial intelligence technologies such as natural language processing and a large language model are deeply integrated into medical care data management, the natural language processing technology carries out preprocessing such as part-of-speech tagging and syntactic analysis on an input query instruction, and the accuracy of large language model understanding is improved. Aiming at the medical field, the big language model optimizes training data selection and model training, collects a large amount of professional medical data and adopts transfer learning and fine tuning technologies to enhance the processing ability of medical knowledge; the problems that in traditional medical data processing, cross-source data formats are not uniform, retrieval efficiency is low, and evaluation subjectivity is high are solved.

Description

Technical Field

[0001] The present invention relates to the technical fields of artificial intelligence and medical information processing, focusing on the application of artificial intelligence technologies such as natural language processing and large language models in the processing of medical care data. Specifically, it covers the intelligent management of medical record data in hospital information systems, the intelligent semantic retrieval of medical literature databases, the automated quantitative assessment of nursing quality, and the development of clinical decision support systems based on the deep integration of multi-source medical data, and specifically relates to an intelligent retrieval and evaluation system and method for medical care data driven by artificial intelligence. Background Art

[0002] The statements in this section only provide background information related to the present disclosure and may not constitute prior art.

[0003] In today's medical care field, the acquisition and analysis of data are crucial for improving the quality of medical services and promoting the development of medical research. In terms of data retrieval, although the internal medical record systems of hospitals have been computerized, the formats of medical record data between different departments and hospitals are not unified, lacking integration and correlation, making it difficult to query across departments and hospitals; the retrieval of medical literature databases is based on keyword matching, making it difficult to understand semantics and context. When retrieving complex problems, the results are either too numerous and require a large amount of screening, or important literature is missed. The assessment of nursing quality mainly relies on manually filling out questionnaires, self-reports by nurses, and simple indicator statistics, which is highly subjective and has inconsistent standards, and cannot comprehensively and objectively reflect the actual quality.

[0004] In clinical decision support, the evaluation of treatment plans is mostly based on doctors' personal experience and limited case analysis, lacking comprehensive analysis and scientific evaluation of a large amount of clinical data. Facing complex diseases, it is difficult to quickly judge the best plan, which easily leads to poor treatment or waste of resources. Moreover, although the amount of medical care data has increased explosively, different types of data are isolated from each other, and the potential value cannot be fully explored, and it cannot provide comprehensive and in-depth support for medical decision-making, nursing optimization, and medical research. These problems seriously restrict the development of the medical care industry, and innovative technical solutions are urgently needed to achieve the efficient retrieval and scientific evaluation of medical care data. Summary of the Invention

[0005] The objective of the present invention is to provide an artificial intelligence-driven intelligent retrieval and evaluation system and method for medical care data, which deeply integrates cutting-edge artificial intelligence technologies such as natural language processing and large language models with the medical care data management process, and effectively solves problems faced in the traditional medical data processing process, such as inconsistent cross-source data formats, low retrieval efficiency, and strong subjectivity in evaluation. The scenarios cover the management of medical record data in hospital information systems, the intelligent retrieval of medical literature databases, the construction of an automated evaluation model for nursing quality, and the development of a clinical decision support system based on the integration of multi-source medical data, etc. By constructing an intelligent system integrating data retrieval, extraction, evaluation, and storage, the efficient integration, in-depth mining, and scientific application of medical care data are realized, providing technical support for improving the quality of medical services, optimizing the nursing process, assisting clinical decision-making, and promoting medical research.

[0006] The technical solution of the present invention is as follows: An artificial intelligence-driven intelligent retrieval and evaluation system for medical care data, comprising: A user interface module, which integrates advanced natural language interaction interface optimization technology. It can not only efficiently receive natural language query instructions input by users, but also use the real-time feedback optimization algorithm of natural language processing to analyze and understand user instructions in real time, providing instant interaction feedback to users; when presenting retrieval and evaluation results, it adopts a visual data presentation optimization strategy, and according to the data type and user needs, displays them in various intuitive forms such as charts, graphs, and text summaries, facilitating users to quickly understand and analyze. A query processing module, which is built with a high-precision database type recognition algorithm and can quickly and accurately judge the target database type; when facing a structured medical database, with the help of a large language model fine-tuned with medical domain expertise and combined with semantic understanding enhancement technology, it converts natural language query statements into optimized structured query instructions, such as optimizing the logic and refining query conditions for complex queries; if the target is an unstructured medical literature database, it uses the query instruction optimization technology of natural language processing to directly pass the optimized query instructions to the data retrieval module to ensure the accuracy and efficiency of retrieval. A data retrieval module, which uses an innovative heterogeneous data fusion retrieval algorithm. This algorithm breaks the barrier between structured medical databases and unstructured medical literature databases by establishing a unified data indexing mechanism, realizing the rapid retrieval of relevant data. Combined with intelligent caching technology, it caches frequently retrieved data to further improve the retrieval speed and reduce the response time. The data extraction module uses a large language model based on the sequence-to-sequence model of deep learning combined with the attention mechanism to accurately extract structured data from unstructured data. During the extraction process, a pre-trained language model in the medical field is utilized, combined with data augmentation techniques, to improve the accuracy and integrity of the extraction. For different types of unstructured data, such as medical literature, medical record texts, etc., targeted extraction strategies are adopted to ensure that key information is not omitted; The evaluation module, based on professional standards in the medical field, clinical experience, and the specific needs of users, uses a large language model combined with the Analytic Hierarchy Process (AHP) and the fuzzy comprehensive evaluation method to construct a multi-dimensional quantitative evaluation model. The weights of each evaluation index are determined through AHP, and then the fuzzy comprehensive evaluation method is used to comprehensively evaluate the extracted structured data; The data storage module connects each module and configures a real-time update mechanism. It adopts distributed storage technology and data encryption algorithms to ensure the security and scalability of the data. The retrieved data, structured data, and evaluation results are classified and stored, and data compression technology is used to reduce the storage space occupancy. Through the data synchronization mechanism, real-time synchronization with various medical data sources is achieved to ensure the timeliness and consistency of the data.

[0007] Furthermore, the query processing module includes: If it is a structured medical database, the query processing module uses a large language model to convert natural language query statements into corresponding structured query instructions; If it is an unstructured medical literature database, the query instructions are directly passed to the data retrieval module.

[0008] Furthermore, the data retrieval module searches for relevant data from the medical database or medical literature database according to the query instructions.

[0009] Furthermore, for unstructured retrieval results, the data extraction module uses a large language model to extract key structured data from them.

[0010] Furthermore, the evaluation module evaluates the extracted structured data according to professional standards in the medical field, clinical experience, and the specific needs of users, using a large language model.

[0011] Furthermore, the evaluation results are presented in the form of quantitative scores or grades.

[0012] Furthermore, it also includes: The data storage module connects each module and configures a real-time update mechanism for storing retrieved data, structured data, and evaluation results.

[0013] The present invention also proposes an intelligent retrieval and evaluation method for medical care data driven by artificial intelligence. Based on the above-mentioned intelligent retrieval and evaluation system for medical care data driven by artificial intelligence, it includes: Step S1: The user interface module uses natural language interaction interface optimization technology and real-time feedback optimization algorithm to receive the natural language query instruction input by the user and immediately give the user interactive feedback; Step S2: The query processing module determines the target database type and generates the corresponding query instruction; Step S3: The data retrieval module uses heterogeneous data fusion retrieval algorithm and intelligent caching technology to quickly retrieve data from the structured medical database and the unstructured medical literature database based on the query instruction; Step S4: The data extraction module adopts a large language model that combines a sequence-to-sequence model based on deep learning with an attention mechanism, and uses a pre-trained language model in the medical field and data augmentation technology to accurately extract structured data from unstructured data; Step S5: The evaluation module uses a multi-dimensional quantitative evaluation model constructed based on the analytic hierarchy process and the fuzzy comprehensive evaluation method to comprehensively evaluate the structured data based on medical professional standards.

[0014] Further, the step S2 includes: If it is a structured medical database, use a large language model fine-tuned with medical field expertise and semantic understanding enhancement technology to convert the natural language query statement into an optimized structured query instruction; If it is an unstructured medical literature database, use the query instruction optimization technology of natural language processing to directly pass the optimized query instruction to the data retrieval module.

[0015] Further, the retrieved data, structured data and evaluation results are also stored through the data storage module.

[0016] Compared with the existing technology, the beneficial effects of the present invention are: 1. In terms of data retrieval, through the transformation of natural language queries combined with large language models, the present invention can achieve accurate retrieval for both structured and unstructured data. For example, when querying materials related to complex diseases, it can accurately screen out the cases and literature that meet the requirements, greatly reducing the retrieval time of medical staff and researchers and improving work efficiency. For the evaluation of nursing quality, based on the evaluation system of objective data and professional standards, it avoids the interference of subjective factors and can comprehensively and accurately reflect the quality of nursing work. Through in-depth analysis of nursing data, it can accurately identify the advantages and disadvantages in nursing work, providing a strong basis for optimizing nursing processes and improving the professional level of nursing staff, thereby enhancing the nursing experience and rehabilitation effect of patients. In clinical decision-making support, by integrating a large amount of clinical data and scientific evaluation methods, it provides doctors with more reliable treatment plan options. In the face of complex diseases, the system can quickly analyze the applicability of different treatment plans for specific patients, weigh the pros and cons of various plans, assist doctors in formulating personalized and efficient treatment plans, improving the treatment success rate and reducing waste of medical resources. From the perspective of overall data utilization, the system breaks the gap between different types of medical data, integrates scattered data for in-depth mining, provides rich data support for medical research, helps to discover new medical laws and treatment methods, and promotes the continuous innovation and development of the medical nursing industry.

[0017] 2. The present invention can unify the data formats of medical nursing data from different sources, achieve efficient and accurate data retrieval, avoid redundancy and omission of retrieval results; establish an objective and comprehensive nursing quality evaluation system, reduce the influence of subjective factors; improve the level of clinical decision-making support, and provide scientific and effective treatment plan options for complex diseases through comprehensive analysis and evaluation of a large amount of clinical data; break the barriers between different types of medical data, fully explore the potential value of data, and promote the overall development of the medical nursing industry.

[0018] 3. In the data retrieval link, the present invention uses heterogeneous data fusion algorithms to achieve accurate retrieval of structured and unstructured data through data marking and unified indexing mechanisms. In terms of evaluation, a quantitative evaluation system based on multi-dimensional indicators is constructed.

[0019] 4. In the evaluation of nursing quality, the present invention uses the analytic hierarchy process to determine the index weights and combines the fuzzy comprehensive evaluation method to achieve dynamic evaluation, ensuring the scientific objectivity of the evaluation results. The scenarios cover the management of medical record data in hospital information systems, the intelligent retrieval of medical literature databases, the construction of an automated evaluation model for nursing quality, and the development of a clinical decision-making support system based on the integration of multi-source medical data.

[0020] 5. The present invention realizes the efficient integration, in-depth mining, and scientific application of medical care data by constructing an intelligent system integrating data retrieval, extraction, evaluation, and storage, providing technical support for improving the quality of medical services, optimizing the nursing process, assisting clinical decision-making, and promoting medical research. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a schematic block diagram of an artificial intelligence-driven intelligent retrieval and evaluation system for medical care data; Figure 2 is an architecture diagram of a large language model; Figure 3 is a flowchart of an artificial intelligence-driven intelligent retrieval and evaluation method for medical care data. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] It should be noted that relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variation thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0023] The features and performance of the present invention will be further described in detail below in conjunction with the embodiments.

[0024] Embodiment 1 Please refer to Figure 1 , an artificial intelligence-driven intelligent retrieval and evaluation system for medical care data, comprising: A user interface module (i.e., the "interface" in the figure), which is used to receive natural language query instructions and present the retrieval and evaluation results; that is, the user interface module serves as an interactive window to receive natural language query instructions input by the user. For example, medical staff input "Find the best nursing plan and related research results for elderly patients with diabetes and kidney complications", and at the same time present the retrieval and evaluation results to the user in an intuitive and easy-to-understand manner; A query processing module (i.e., the "query processor" in the figure), connected to the user interface module, which is used to determine the type of target database and generate corresponding query instructions; that is, after receiving a natural language query, the query processing module will determine the type of database where the target data is located; The data retrieval module (i.e., the "search engine" in the figure), which is connected to the query processing module, is used to retrieve data from a structured medical database and an unstructured medical literature database; The data extraction module (i.e., the "extractor" in the figure), which is connected to the data retrieval module, extracts structured data from unstructured data using a large language model; The evaluation module (i.e., the "shield" in the figure), which is connected to the data extraction module, conducts multi-dimensional quantitative evaluation of the structured data based on medical professional standards; The data storage module (i.e., the "storage" in the figure), which is connected to each module and is configured with a real-time update mechanism, is used to store the retrieved data, structured data, and evaluation results; that is, the data storage module is used to store the retrieved data, extracted structured data, and evaluation results, and has a data update mechanism that can obtain update information of medical data sources in real time or regularly to ensure the timeliness of the data.

[0025] In this embodiment, it should be noted that the large language model can adopt existing models. The architecture of the large language model adopted in this embodiment is as Figure 2 shown.

[0026] In this embodiment, specifically, the query processing module includes: If it is a structured medical database, the query processing module uses a large language model to convert a natural language query statement into a corresponding structured query instruction; If it is an unstructured medical literature database, the query instruction is directly passed to the data retrieval module; That is, if it is a structured medical database, such as a hospital's case database, etc., the query processing module uses a large language model to convert a natural language query into a corresponding structured query statement, such as an SQL query statement, for accurate retrieval; If retrieving unstructured data resources such as a medical literature database, the query instruction is directly passed to the data retrieval module.

[0027] In this embodiment, specifically, the data retrieval module searches for relevant data from a medical database or a medical literature database according to the query instruction; That is, the data retrieval module searches for relevant data from various medical data sources according to the query instruction. These data sources include medical literature databases, case data in hospital information systems, drug databases, etc.

[0028] In this embodiment, specifically, for unstructured retrieval results, the data extraction module uses a large language model to extract key structured data from them. For example, it extracts research methods, experimental results, conclusions, etc. from medical literature, and extracts patient basic information, symptom manifestations, diagnosis results, treatment processes and effects, etc. from cases.

[0029] In this embodiment, specifically, the evaluation module evaluates the extracted structured data using a large language model based on professional standards in the medical field, clinical experience, and the specific needs of the user; It should be noted that in the clinical decision-making scenario, factors such as the effectiveness, safety, and cost-effectiveness of different treatment plans will be comprehensively considered; in the evaluation of nursing quality, the implementation of nursing measures, patient satisfaction, nursing effects, etc. will be evaluated; in medical research, the reliability, innovation, and research value of research data will be evaluated, and the evaluation results will be presented in the form of quantified scores or grades.

[0030] Please refer to Figure 3 , this embodiment also proposes an intelligent retrieval and evaluation method for medical nursing data driven by artificial intelligence. Based on the above-mentioned intelligent retrieval and evaluation system for medical nursing data driven by artificial intelligence, it includes: Step S1: Receive a natural language query instruction input by the user through the user interface module; Step S2: Determine the target database type through the query processing module and generate a corresponding query instruction; Step S3: The data retrieval module retrieves data from the structured medical database and the unstructured medical literature database based on the query instruction; Step S4: The data extraction module extracts structured data from the unstructured data using a large language model; Step S5: Perform multi-dimensional quantitative evaluation on the structured data through the evaluation module based on medical professional standards.

[0031] In this embodiment, specifically, the said Step S2 includes: If it is a structured medical database, the query processing module uses a large language model to convert the natural language query statement into a corresponding structured query instruction; If it is an unstructured medical literature database, the query instruction is directly passed to the data retrieval module.

[0032] In this embodiment, specifically, the retrieved data, structured data, and evaluation results are also stored through the data storage module.

[0033] Embodiment Two Taking the clinical decision support scenario as an example, Embodiment Two further illustrates the intelligent retrieval and evaluation system and method for medical nursing data driven by artificial intelligence proposed in Embodiment One.

[0034] Suppose an endocrinologist is treating a patient with type 2 diabetes and cardiovascular complications.

[0035] The doctor inputs a natural language query through the user interface module of this system: "Provide the latest treatment plan and relevant clinical research results for this 55-year-old type 2 diabetic patient with a history of hypertension and cardiovascular complications."

[0036] After receiving the query, the query processing module determines that it is necessary to retrieve the hospital's internal case database and medical literature database. For the hospital case database, the query processing module uses a large language model to convert the natural language query into an SQL query statement. The generated SQL query statement may be similar to: "SELECT * FROM case table WHERE disease diagnosis LIKE '%type 2 diabetes%' AND complications LIKE '%cardiovascular complications%' AND patient age = 55 AND past medical history LIKE '%hypertension%'", and sends this query statement to the data retrieval module. For the medical literature database, the natural language query is directly passed to the data retrieval module.

[0037] After the data retrieval module retrieves data from the corresponding data sources, the returned medical literature is a large amount of unstructured text, and although the case data is structured, it contains a lot of redundant information.

[0038] The data extraction module uses a large language model to extract structured data such as the core conclusions of the research, treatment methods, and experimental data from the medical literature; and extracts key information such as the treatment process and treatment effects of similar patients from the case data.

[0039] The evaluation module evaluates the extracted structured data using a large language model based on medical professional standards, such as diabetes treatment guidelines and cardiovascular disease treatment specifications. Evaluate the effectiveness (such as blood glucose control effect, improvement of cardiovascular symptoms), safety (such as risk of drug side effects), and cost-effectiveness (treatment cost, length of hospital stay, etc.) of different treatment plans.

[0040] Compare the advantages and disadvantages of different drug combination treatment plans, and after comprehensively considering various factors, generate a quantitative evaluation result and recommended plan for the doctor.

[0041] The user interface module presents the evaluation results to the doctor, including a detailed introduction of different treatment plans, evaluation scores and recommended orders, as well as key information of relevant clinical research. Based on this information and combined with the actual situation of the patient, the doctor formulates a more scientific and reasonable treatment plan.

[0042] Example 3 Taking the nursing quality assessment scenario as an example, Example 3 further illustrates the artificial intelligence-driven intelligent retrieval and evaluation system and method for medical nursing data proposed in Example 1.

[0043] The nursing department of a certain hospital wants to evaluate the nursing quality of a certain department.

[0044] The staff of the nursing department enter in the user interface module of the system: "Evaluate the nursing quality of the internal medicine ward in the past month, with a focus on patient satisfaction and the implementation of nursing measures."

[0045] The query processing module converts this natural language query into a structured query instruction for the hospital nursing record database and patient feedback database, and sends it to the data retrieval module.

[0046] The data retrieval module retrieves the nursing records and feedback information of all patients in the internal medicine ward in the past month from the database.

[0047] The data extraction module uses a large language model to extract the implementation of nursing measures from the nursing records, such as the number of times of administering medications on time, the frequency of vital sign monitoring, etc.; and extracts patient satisfaction-related information from the patient feedback, such as evaluations of nursing attitude and nursing effect.

[0048] The evaluation module evaluates the extracted data using a large language model according to the nursing quality evaluation criteria. Calculate the integrity score of the implementation of nursing measures and the patient satisfaction score according to the set weights, and comprehensively obtain the nursing quality score of this department. For items with lower scores, further analyze the reasons. For example, the non-standard implementation of some nursing measures may be due to insufficient nurse training, etc.

[0049] The user interface module presents the evaluation results to the nursing department staff in the form of charts and detailed reports. The report includes the overall nursing quality score, the scores of each indicator, existing problems and improvement suggestions. Based on these results, the nursing department formulates targeted improvement measures, such as strengthening nurse training, optimizing the nursing process, etc., so as to improve the nursing quality.

[0050] It should be noted that the core of the present invention lies in constructing an intelligent processing system for medical care data driven by artificial intelligence, with key protection for cross-modal intelligent retrieval technology (linking the structured medical record library SQL conversion and unstructured literature library retrieval through natural language instructions, and precisely disassembling complex medical problems with a dynamic mapping algorithm), multi-dimensional intelligent evaluation system (constructing a quantitative evaluation model based on large language models integrated with medical professional standards, covering scenarios such as clinical decision-making and nursing quality and outputting standardized results, and fine-tuning with domain knowledge to ensure standardization), full-process data processing closed-loop ("retrieval - extraction - evaluation - storage" integrated architecture, realizing structured extraction of unstructured data through large language models, combining real-time updates to ensure data timeliness, and forming a technical barrier for heterogeneous data conversion and incremental updates), and scenario-based intelligent applications (generating personalized treatment plan recommendations based on big data analysis, dynamically monitoring nursing quality and positioning problems). The core protects the innovative methods of natural language processing and large language models in medical data retrieval, evaluation, and application, as well as the full-process intelligent system architecture, providing an efficient and objective data intelligent processing solution for the medical care industry.

Claims

1. An intelligent retrieval and evaluation system for medical care data driven by artificial intelligence, characterized in that, Including: A user interface module that uses input preprocessing technology of natural language processing to accurately identify key information in natural language query instructions and presents the retrieval evaluation results through an intuitive visual interaction interface; A query processing module that connects to the user interface module and uses a built-in database type recognition algorithm to determine the target database type; For A structured medical database, using a large language model combined with the preprocessing results of natural language processing to convert natural language query statements into optimized structured query instructions and passing them to the data retrieval module; for an unstructured medical literature database, directly passing the query instructions optimized by natural language processing to the data retrieval module; A data retrieval module that connects to the query processing module and uses a heterogeneous data fusion retrieval algorithm to quickly retrieve data from a structured medical database and an unstructured medical literature database according to the query instructions; A data extraction module that connects to the data retrieval module and uses a large language model based on a sequence-to-sequence model of deep learning combined with an attention mechanism to accurately extract structured data from unstructured data; An evaluation module that connects to the data extraction module and uses an evaluation model constructed by a large language model combined with the analytic hierarchy process and fuzzy comprehensive evaluation method to perform multi-dimensional quantitative evaluation on the structured data based on medical professional standards, clinical experience, and user needs; A data storage module that connects to each module and is configured with a real-time update mechanism, and uses distributed storage technology and data encryption algorithms to store retrieval data, structured data, and evaluation results.

2. The artificial intelligence-driven intelligent retrieval and evaluation system for medical care data according to claim 1, wherein The query processing module includes: If it is a structured medical database, the query processing module uses a large language model to convert natural language query statements into corresponding structured query instructions; If it is an unstructured medical literature database, directly pass the query instructions to the data retrieval module.

3. The artificial intelligence-driven intelligent retrieval and evaluation system for medical care data according to claim 1, wherein The data retrieval module searches for relevant data from a medical database or a medical literature database according to the query instructions.

4. The artificial intelligence-driven intelligent retrieval and evaluation system for medical care data according to claim 1, wherein For unstructured retrieval results, the data extraction module uses a large language model to extract key structured data from them.

5. The artificial intelligence-driven intelligent retrieval and evaluation system for medical care data according to claim 1, characterized in that, The evaluation module evaluates the extracted structured data using a large language model according to professional standards, clinical experience, and user needs in the medical field.

6. The artificial intelligence-driven intelligent retrieval and evaluation system for medical care data according to claim 5, wherein The evaluation results are presented in the form of quantitative scores or grades.

7. The artificial intelligence-driven intelligent retrieval and evaluation system for medical care data according to claim 1, wherein It also includes: A data storage module that connects to each module and is configured with a real-time update mechanism for storing retrieval data, structured data, and evaluation results.

8. An intelligent retrieval and evaluation method for medical care data driven by artificial intelligence, characterized in that, The artificial intelligence-driven intelligent retrieval and evaluation system for medical care data according to any one of claims 1-7 includes: Step S1: Receive a natural language query instruction input by the user through the user interface module using natural language processing and input preprocessing technology; Step S2: Determine the target database type through the query processing module using a database type recognition algorithm, and generate corresponding query instructions using a large language model combined with the preprocessing results of natural language processing; Step S3: The data retrieval module uses a heterogeneous data fusion retrieval algorithm based on the query instructions to retrieve data from a structured medical database and an unstructured medical literature database; Step S4: The data extraction module extracts structured data from unstructured data using a large language model based on a sequence-to-sequence model with an attention mechanism in deep learning; Step S5: The evaluation module uses an evaluation model constructed based on the analytic hierarchy process and the fuzzy comprehensive evaluation method to conduct multi-dimensional quantitative evaluation of the structured data based on medical professional standards; the data storage module uses distributed storage technology and data encryption algorithms to store and retrieve data, structured data, and evaluation results.

9. The intelligent retrieval and evaluation method for medical care data driven by artificial intelligence according to claim 8, characterized in that The said step S2 includes: If it is a structured medical database, the query processing module uses a large language model to convert natural language query statements into corresponding structured query instructions; If it is an unstructured medical literature database, the query instructions are directly passed to the data retrieval module.

10. The artificial intelligence-driven intelligent retrieval and evaluation method for medical care data according to claim 8, characterized in that, The retrieval data, structured data, and evaluation results are also stored through the data storage module.

Images