A method and system for constructing a knowledge base for brain disease nursing

Abstract

The present invention relates to the technical field of database construction, and in particular to a method and system for constructing a knowledge base for brain disease nursing. The method comprises the following steps: acquiring multi-source brain disease data; performing structured data screening on the multi-source brain disease data to generate structured brain disease data and unstructured brain disease data; performing multi-modal data structured stratification on the unstructured brain disease data to generate multi-modal structured stratified data; performing data preprocessing on the multi-modal structured stratified data and the structured brain disease data to generate a standard structured brain disease data set; performing context analysis on the standard structured brain disease data set to generate a semantic vector set of brain disease information; performing redundant relationship pruning processing on the semantic vector set of brain disease information to generate spatial semantic optimized data. The present invention improves the construction quality and practicality of the brain disease nursing knowledge base by integrating multi-source data, structured and unstructured information, real-time dynamic mapping of nursing needs and system feedback evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of database construction, and in particular to a method and system for constructing a brain disease nursing knowledge base. Background Art

[0002] Early brain disease nursing mainly relied on clinical experience and traditional teaching, lacking systematic knowledge integration, which led to the problems of insufficient information and unscientific decision-making for nursing staff when dealing with brain disease patients. With the emergence of electronic health records (EHR) and clinical decision support systems (CDSS), the acquisition and application of nursing knowledge began to develop in the direction of digitalization and intelligence. With the popularization of network technology, many hospitals and research institutions began to explore how to transform brain disease nursing knowledge into shareable digital resources. The initial establishment of the knowledge base mainly relied on the manual input of experts and the collation of literature, but due to the slow knowledge update, the application effect was limited. The rapid development of artificial intelligence technologies such as machine learning and natural language processing has brought new opportunities for the construction of brain disease nursing knowledge base. By analyzing a large amount of clinical data and literature, researchers can automatically extract and organize the best practices and latest research results of brain disease nursing, significantly improving the efficiency and accuracy of knowledge base construction. However, the value of unstructured data (such as doctor-patient communication records, imaging data, etc.) in brain disease nursing is often overlooked. At the same time, in the previous knowledge base construction, there was a lack of in-depth analysis of the semantic relationship of brain disease information, which led to the low construction quality and practicality of brain disease nursing knowledge base. Summary of the invention

[0003] Based on this, it is necessary to provide a method and system for constructing a brain disease nursing knowledge base to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for constructing a brain disease nursing knowledge base is provided, the method comprising the following steps:

[0005] Step S1: Acquire multi-source encephalopathy data; perform structured data screening on the multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data; perform multi-modal data structured stratification on the unstructured encephalopathy data to generate multi-modal structured stratified data; perform data preprocessing on the multi-modal structured stratified data and structured encephalopathy data to generate a standard structured encephalopathy data set;

[0006] Step S2: Perform context analysis on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information; perform redundant relationship pruning on the semantic vector set of encephalopathy information to generate spatial semantic optimization data; perform collaborative reasoning modeling on the spatial semantic optimization data to generate a distributed encephalopathy collaborative association network; perform dynamic mapping of encephalopathy nursing needs on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate dynamic nursing needs data;

[0007] Step S3: Acquire the user's real-time encephalopathy physiological monitoring data; perform abnormal monitoring data screening on the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data; perform nursing implementation data collection on dynamic nursing demand data according to the user's real-time encephalopathy abnormal physiological monitoring data to obtain the user's real-time encephalopathy physiological monitoring data after implementation; perform dynamic change data screening on the user's real-time encephalopathy physiological monitoring data after implementation and the user's real-time encephalopathy abnormal physiological monitoring data to generate personalized encephalopathy nursing change data;

[0008] Step S4: Conducting brain disease nursing feedback evaluation on the personalized brain disease nursing change data to generate brain disease nursing feedback evaluation data; constructing a database on the brain disease nursing feedback evaluation data to generate a brain disease nursing knowledge base.

[0009] The present invention obtains multi-source encephalopathy data. Multi-source data acquisition provides a comprehensive information foundation, which can analyze encephalopathy from different angles and enhance the depth and breadth of research. Structured data screening is performed on multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data. The extraction of structured data makes subsequent analysis more efficient, convenient for computer processing and analysis, and the value of unstructured data is mined through subsequent processing. Multi-modal data structured stratification is performed on unstructured encephalopathy data to generate multi-modal structured stratified data. The stratification of unstructured data helps to identify the relationship between different data modalities and improve the flexibility of data integration and analysis. Data preprocessing is performed on multi-modal structured stratified data and structured encephalopathy data to generate a standard structured encephalopathy data set. Preprocessing ensures the consistency and accuracy of the data, lays a good foundation for subsequent analysis, and reduces errors caused by data quality issues. Context analysis is performed on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information. Context analysis helps to extract implicit information in the data, generate a semantic vector set to capture encephalopathy characteristics, and facilitate subsequent deeper analysis. The redundant relationship pruning process is performed on the semantic vector set of encephalopathy information to generate spatial semantic optimization data. Pruning improves the availability of data, reduces noise, and improves the efficiency of subsequent analysis and modeling. Collaborative reasoning modeling is performed on the spatial semantic optimization data to generate a distributed encephalopathy collaborative association network. The collaborative association network can reveal the correlation between different factors of encephalopathy and provide data support for precision medicine. Dynamic mapping provides personalized nursing needs, can adjust the nursing plan according to the specific situation of the patient, and improve the pertinence and effectiveness of nursing. Real-time monitoring provides the latest information on the current patient status and supports timely nursing decisions. Timely identification of abnormal data can prompt nursing staff to respond quickly. The collection of post-implementation data provides a basis for subsequent effect evaluation and adjustment of nursing strategies. Dynamically changing data helps to evaluate the effectiveness of nursing measures and supports the adjustment and optimization of personalized nursing strategies. Feedback evaluation provides a reference for the effectiveness of subsequent nursing measures and promotes continuous improvement. The construction of the knowledge base provides valuable data resources for future research and practice, and promotes the accumulation and sharing of knowledge in the field of encephalopathy nursing. Therefore, the present invention improves the construction quality and practicality of the brain disease nursing knowledge base by integrating multi-source data, structured and unstructured information, real-time dynamic mapping of nursing needs and system feedback evaluation.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: Acquire multi-source brain disease data;

[0012] Step S12: screening the multi-source encephalopathy data by structured data to generate structured encephalopathy data and unstructured encephalopathy data;

[0013] Step S13: dividing the unstructured encephalopathy data into data types to generate text-type encephalopathy data and image-type encephalopathy data; performing multimodal data structured stratification on the text-type encephalopathy data and image-type encephalopathy data to generate multimodal structured stratified data;

[0014] Step S14: Integrate the multimodal structured hierarchical data and the structured encephalopathy data to generate a structured encephalopathy data set; perform data preprocessing on the structured encephalopathy data set to generate a standard structured encephalopathy data set, wherein the data preprocessing includes data cleaning, data missing value filling and data standardization.

[0015] The present invention can integrate multiple information types and improve the comprehensiveness and representativeness of data by acquiring encephalopathy data from different sources. This provides a multi-angle perspective for subsequent analysis and helps to better understand the complexity of encephalopathy. Multi-source encephalopathy data is screened for structured data to generate structured encephalopathy data and unstructured encephalopathy data, making data processing more efficient. The extraction of structured data facilitates subsequent calculation and analysis, and at the same time, the retention of unstructured data also provides the possibility for subsequent deep mining. Data types are divided for unstructured encephalopathy data to generate text type encephalopathy data and image type encephalopathy data, which can better organize and manage data and help to apply specific analysis methods for different data types. Multimodal data structured stratification is performed on text type and image type encephalopathy data so that different modal data can be understood and analyzed under the same framework. This stratification can more effectively capture the intrinsic relationship in the data and help to improve the accuracy of data analysis. Multimodal structured stratified data is integrated with structured encephalopathy data to generate a unified structured encephalopathy data set, providing a systematic data platform and providing a comprehensive and neat foundation for subsequent analysis and research. Data preprocessing was performed on the structured encephalopathy dataset, including data cleaning, data missing value filling, and data standardization. This step ensures the consistency and quality of the data and improves the reliability of the data. Through cleaning and standardization, noise and bias in the data can be eliminated, and the effectiveness and comparability of subsequent analysis can be enhanced. A standardized encephalopathy dataset was constructed by acquiring, screening, dividing, and integrating encephalopathy data. Its beneficial effects are: improving the quality and effectiveness of the data, enhancing the integration ability of multimodal data, and laying a solid foundation for subsequent analysis and research.

[0016] Preferably, multimodal data structured stratification of text-type encephalopathy data and image-type encephalopathy data includes:

[0017] Performing feature classification indexing on text-type encephalopathy data and image-type encephalopathy data to obtain a text-type feature index and an image-type feature index;

[0018] The text type feature index is used to extract word features from the text type encephalopathy data to obtain the encephalopathy word vector; the image type feature index is used to extract embedding features from the image type encephalopathy data to obtain the encephalopathy embedding vector;

[0019] Matching feature points of brain disease word vectors and brain disease embedding vectors to generate brain disease feature point matching data; hierarchically fusing feature maps of brain disease word vectors and brain disease embedding vectors based on the brain disease feature point matching data to generate a multi-dimensional feature hierarchical table;

[0020] A standardized matrix was constructed for text-type encephalopathy data and imaging-type encephalopathy data according to the multidimensional feature hierarchy table to generate multimodal structured hierarchical data.

[0021] The present invention can effectively organize and manage different types of data by establishing feature classification indexes, which is convenient for subsequent feature extraction and analysis. Feature indexing provides a clear direction for model training and feature selection, and improves the efficiency of data processing. Extracting brain disease word vectors can convert text information into numerical representations that can be used for machine learning and deep learning, retain the semantic information of text data, and help improve the accuracy of subsequent analysis. Embedded feature extraction converts image data into vector representations, which can effectively capture important features in images and improve the understanding and processing capabilities of subsequent analysis models for image features. Feature point matching can identify the correlation between text data and image data, helping to reveal different manifestations of encephalopathy and their mutual influence. Through feature point matching, an important basis can be provided for subsequent feature fusion, enhancing the effectiveness of multimodal analysis. Hierarchical fusion of feature maps can combine data of different modalities to form a multidimensional feature representation, so that data can be comprehensively analyzed at a higher level. This fusion helps to capture the complex relationship between data, improve the performance and generalization ability of the model, and promote the progress of deep learning applications. The construction of a standardized matrix uniformly formats data from different sources to ensure the comparability and consistency of the data. Through standardization, the noise and deviation in the data can be reduced, laying a more solid foundation for subsequent analysis and modeling. By structuring and stratifying multimodal data of text-type encephalopathy data and imaging-type encephalopathy data, the overall process effectively improves the efficiency and quality of data processing. The beneficial effects of each step include strengthening data organization and management, improving the accuracy of feature extraction and matching, promoting the effective fusion of different modal data, and finally generating standardized multimodal structured stratified data.

[0022] Preferably, step S2 comprises the following steps:

[0023] Step S21: performing semantic recognition on a standard structured encephalopathy data set to generate a structured encephalopathy semantic vector; performing context analysis on the standard structured encephalopathy data set based on the structured encephalopathy semantic vector to generate a encephalopathy information semantic vector set;

[0024] Step S22: performing hierarchical semantic relationship modeling on the semantic vector set of brain disease information to generate spatial semantic modeling data; performing redundant relationship pruning processing on the spatial semantic modeling data to generate spatial semantic optimization data;

[0025] Step S23: constructing a cross-node knowledge association network for the spatial semantic optimization data to generate a cross-node brain disease association network; performing collaborative reasoning modeling on the cross-node brain disease association network to generate a distributed brain disease collaborative association network;

[0026] Step S24: Based on the distributed brain disease collaborative association network, the standard structured brain disease data set is predicted to evolve complex diseases, and disease evolution prediction data is generated; the brain disease nursing needs are dynamically mapped to the disease evolution prediction data, and dynamic nursing needs data is generated.

[0027] The present invention converts raw data into vector representations with a higher level of abstraction through semantic recognition, which helps to capture implicit information and relationships in the data. Context analysis enables the semantic vector set of encephalopathy information to reflect the background and characteristics of encephalopathy-related data, enhances the interpretability and usability of the data, and provides a rich information basis for subsequent analysis. Hierarchical semantic relationship modeling can identify and construct multi-level and complex relationships between data, providing a clear structure for subsequent data analysis and decision-making. Redundant relationship pruning optimizes the semantic model, reduces information redundancy, improves the efficiency of the model, and ensures that subsequent analysis focuses on truly important semantic relationships. The cross-node knowledge association network integrates knowledge from different data sources, promotes interaction and information sharing between data, and enhances the understanding of the complexity and diversity of encephalopathy. Collaborative reasoning modeling further enhances the reasoning ability of the model, and can more effectively reveal the potential mechanisms and influencing factors of encephalopathy progression. Complex disease evolution prediction can provide a scientific and quantitative analysis of encephalopathy progression, help identify trends and patterns of disease changes, and provide a basis for clinical decision-making. The generation of dynamic nursing demand data enables the nursing plan to be adjusted in a timely manner according to the actual changes in the patient's condition, ensure the personalization and effectiveness of medical services, and improve nursing quality and patient satisfaction. Through comprehensive analysis and modeling of standard structured encephalopathy datasets, step S2 improves the scientificity and flexibility of encephalopathy care in many aspects. The beneficial effects of each sub-step include improving the accuracy of semantic recognition and context analysis, optimizing the efficiency of data models, enhancing the relevance of knowledge networks, and providing reliable disease evolution prediction and dynamic nursing needs mapping.

[0028] Preferably, performing redundant relationship pruning processing on spatial semantic modeling data includes:

[0029] Perform semantic node recognition on the spatial semantic modeling data to obtain spatial semantic node recognition data; perform adjacent node relationship density calculation on the spatial semantic modeling data according to the spatial semantic node recognition data to obtain adjacent node relationship density data; perform redundant relationship node analysis on the spatial semantic node recognition data using the adjacent node relationship density data to generate redundant relationship node data;

[0030] Score the spatial semantic node recognition data for relationship importance to generate relationship semantic importance score data; sort the relationship semantic importance score data by pruning priority through redundant relationship node data to generate pruning priority score data; calculate the information entropy value of redundant relationship node data by pruning priority score data to obtain the redundant relationship node information entropy value;

[0031] The information entropy value of the redundant relationship node is compared with the preset standard information entropy value threshold. When the information entropy value of the redundant relationship node is less than the preset standard information entropy value threshold, the corresponding redundant relationship node data is redundantly pruned to generate spatial semantic optimization data.

[0032] The present invention provides clear basic data for subsequent analysis through semantic node identification, so that the relationship between different semantics can be clearly identified and analyzed. Relationship density calculation can quantify the degree of association between nodes, which helps to identify important semantic relationships in the data and ensure that subsequent processing focuses on the most relevant nodes. Redundant relationship node analysis can effectively identify redundant or low-value nodes in the data, thereby reducing unnecessary information interference. Relationship importance scoring provides a quantitative evaluation for nodes, so that subsequent pruning decisions are based on data-driven basis, improving the scientific nature of pruning. Pruning priority sorting ensures that the least important nodes are pruned first, effectively reducing the complexity and computational cost of data processing. The calculation of information entropy provides a quantitative standard for the value of redundant relationship nodes, further improving the effectiveness of pruning decisions and ensuring that important information is retained. Threshold comparison provides a clear decision-making standard for pruning, ensuring the systematicness and consistency of the pruning process. By effectively removing low-value redundant nodes, the generated spatial semantic optimization data will be more concise and clear, thereby improving the accuracy and efficiency of subsequent analysis and model training. Through redundant relationship pruning processing, the quality of spatial semantic modeling data has been significantly improved. The beneficial effects of each sub-step include effectively identifying and analyzing important semantic nodes, reducing redundant information, ensuring the scientific nature of pruning decisions, and ultimately generating more optimized spatial semantic data.

[0033] Preferably, step S23 includes the following steps:

[0034] Step S231: performing cross-node association mining on the spatial semantic optimization data to generate cross-node association data; performing semantic aggregation on the cross-node association data to generate a preliminary cross-node knowledge association network;

[0035] Step S232: performing node association relationship weight distribution on the cross-node knowledge association network to generate cross-node association relationship weights; performing multi-layer relationship weight optimization on the cross-node knowledge association network according to the cross-node association relationship weights to generate a multi-layer weighted association network;

[0036] Step S233: performing connection enhancement on the multi-layer weighted association network to generate a cross-node brain disease association network;

[0037] Step S234: Perform multi-node information joint reasoning analysis on the cross-node brain disease association network to generate multi-node information joint reasoning data; perform distributed reasoning node optimization on the cross-node brain disease association network based on the multi-node information joint reasoning data to generate a distributed brain disease collaborative association network.

[0038] The present invention can reveal the deep connection between nodes through association mining, and provide basic data for subsequent knowledge discovery and data analysis. Semantic aggregation ensures that the constructed association network is more accurate and comprehensive by integrating relevant information, which helps to improve the performance and interpretation ability of data in the model. Weight allocation gives quantitative standards to the relationship of each node, so that important nodes and relationships in the network can be emphasized, and the accuracy of analysis is improved. Multi-layer relationship weight optimization can improve the flexibility and adaptability of the network, so that relationships at different levels are properly reflected, and the overall performance of the network is enhanced. The connection enhancement operation effectively improves the connectivity and information flow of the network by strengthening the correlation between nodes in the network, which helps to improve the efficiency of reasoning and analysis. This enhancement also improves the modeling ability of the network for complex relationships, so that the actual situation in the field of encephalopathy can be better reflected. Multi-node information joint reasoning analysis can deeply explore the interaction between each node, helping to better understand the complexity and dynamic changes of encephalopathy. Distributed reasoning node optimization ensures the efficiency and reliability of the network when processing large amounts of data, helps to achieve real-time analysis and decision support, and promotes the progress of the field of encephalopathy care. Through the implementation of step S23, the cross-node encephalopathy association network has been significantly optimized and enhanced. These operations not only improve the relevance and analytical capabilities of the data, but also provide a more reliable basis for subsequent reasoning and decision-making. Ultimately, the generated distributed brain disease collaborative association network will greatly improve the understanding and response capabilities of brain disease care needs.

[0039] Preferably, performing multi-node information joint reasoning analysis on a cross-node brain disease association network includes:

[0040] Aggregating node data of the cross-node brain disease association network to generate cross-node brain disease association aggregated data; extracting key features of the cross-node brain disease association aggregated data to obtain cross-node brain disease association key feature data, wherein the key feature extraction includes mean calculation, median calculation and variance calculation;

[0041] By using the cross-node correlation key feature data, the Pearson coefficient between nodes of the cross-node brain disease correlation network is calculated to obtain the cross-node node similarity data; the correlation analysis is performed on the cross-node node similarity data to generate a node similarity matrix;

[0042] According to the node similarity matrix, the rule-based reasoning relationship of the cross-node brain disease association network is established to generate a rule-based reasoning set; based on the rule-based reasoning set, information reasoning is performed on the cross-node brain disease association network to generate a reasoning result set; the reasoning result set is logically deduced to generate multi-node information joint reasoning data.

[0043] The present invention can effectively integrate the information of each node through data aggregation, provide a more comprehensive perspective, reduce data redundancy, and improve data utilization efficiency. This integration helps to provide a more reliable data basis for subsequent analysis and reasoning, thereby improving the accuracy of the analysis. By extracting key features, important information related to brain disease can be extracted to help identify the main trends and changes between nodes. These statistical features (mean, median, variance) provide important quantitative indicators for subsequent similarity calculation and reasoning analysis, and enhance the explanatory power and predictive ability of the model. Pearson coefficient calculation can quantify the linear correlation between nodes, identify nodes with similar features, and reveal the potential association between them. This similarity analysis provides a scientific basis for establishing a relationship network, which helps to more accurately understand the intrinsic connection between different brain disease data. The node similarity matrix provides an intuitive tool for analyzing the relationship between nodes, which helps to identify key nodes and their influence. This analysis can provide an effective basis for subsequent reasoning and decision support, which is particularly important in medical decision-making. The establishment of a rule-based reasoning set can form a data-driven reasoning model, so that the network can perform effective reasoning and deduction based on existing data. The establishment of such rules provides a systematic method for subsequent reasoning analysis, which helps to improve the reliability and consistency of reasoning. The combination of information reasoning and logical deduction can transform the information in the network into useful reasoning results, support decision-making and the formulation of action plans, and the generated multi-node information joint reasoning data can provide a basis for clinical practice, helping medical staff to better understand the changes in patients' conditions and formulate corresponding nursing measures.

[0044] Preferably, step S24 includes the following steps:

[0045] Step S241: performing fuzzy reasoning on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate a encephalopathy fuzzy reasoning result data set; performing identical result data screening on the encephalopathy fuzzy reasoning result data set to obtain identical reasoning result data;

[0046] Step S242: Divide the data set into a data set with the same inference result data to generate a model training set and a model test set; perform model training on the model training set using a long short-term memory neural network algorithm to generate a complex disease evolution prediction pre-model; perform model optimization iteration on the complex disease evolution prediction pre-model using the model test set to generate a complex disease evolution prediction model;

[0047] Step S243: Import the same inference result data into the complex disease evolution prediction model to perform complex disease evolution prediction and generate disease evolution prediction data; dynamically map brain disease nursing needs to the disease evolution prediction data and generate dynamic nursing needs data.

[0048] The present invention can handle uncertainty and ambiguity through fuzzy reasoning, provide a more flexible reasoning method, and enable the model to better adapt to complex clinical scenarios. The screening of the same result data helps to remove noise and redundant information, ensure the accuracy and consistency of subsequent analysis, and enhance the reliability of the prediction model. Data set partitioning provides an effective basis for model training and testing, ensuring the generalization ability and prediction accuracy of the model. Long short-term memory neural network (LSTM) can effectively capture long-term dependencies in time series data and is suitable for the prediction of complex disease evolution. The optimization iteration of the model continuously improves the prediction accuracy, and the complex disease evolution prediction model finally formed can provide a more accurate analysis of the disease evolution trend. The prediction of complex disease evolution provides important decision-making support for clinicians, helping them to foresee potential changes in the disease and take corresponding intervention measures. Dynamic nursing demand mapping can reflect the demand for nursing resources and services for changes in the disease in real time, help optimize the nursing process, and improve the pertinence and effectiveness of nursing services. In addition, dynamic nursing demand data can provide a reference for hospital management, ensure the rational allocation and use of medical resources, and enhance the overall nursing experience of patients. Through the implementation of step S24, the distributed brain disease collaborative association network not only improves the ability to predict the evolution of complex diseases, but also provides a data-driven dynamic adjustment plan for clinical care. This method based on fuzzy reasoning and deep learning effectively combines data analysis with clinical needs, ultimately achieving personalized and efficient nursing services, and providing patients with better medical insurance and nursing experience.

[0049] Preferably, step S3 comprises the following steps:

[0050] Step S31: Acquire real-time encephalopathy physiological monitoring data of the user;

[0051] Step S32: Screening the abnormal monitoring data of the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data;

[0052] Step S33: Generate a personalized encephalopathy nursing plan for the dynamic nursing demand data according to the user's real-time abnormal encephalopathy physiological monitoring data to obtain a personalized encephalopathy nursing plan;

[0053] Step S34: Implement the nursing plan on the user's real-time brain disease physiological monitoring data through a personalized brain disease nursing plan, and synchronously collect the user's real-time brain disease physiological monitoring data after implementation; dynamically screen the user's real-time brain disease physiological monitoring data and the user's real-time brain disease abnormal physiological monitoring data after implementation to generate personalized brain disease nursing change data.

[0054] The present invention can ensure that the nursing team can grasp the patient's physiological state at any time through the acquisition of real-time monitoring data, and provide a basis for timely nursing decisions. Through continuous monitoring, the physiological changes of encephalopathy patients in different time periods can be captured, which is helpful to establish a comprehensive record of the development of the disease. By screening abnormal data, the nursing team can more quickly identify the patient's potential risks, timely discover changes in the disease, and prevent the disease from worsening. This step enhances the effectiveness of monitoring, so that nursing resources can be concentrated on patients who need special attention, and improves the pertinence of nursing. Personalized nursing plans can be tailored according to the patient's specific physiological state and needs, improving the nursing effect and patient satisfaction. This method ensures the applicability and effectiveness of nursing measures, helps to improve the patient's recovery rate and reduce the risk of complications. Data collection after implementation can provide direct feedback on the effect of nursing measures, and provide data basis for the adjustment and optimization of subsequent nursing plans. The analysis of dynamically changing data helps to identify the effectiveness and adaptability of nursing measures, so that the nursing process can be flexibly adjusted to adapt to the changing needs of patients. This real-time feedback mechanism not only enhances the patient's sense of security, but also improves the responsiveness and decision-making efficiency of the nursing team. Through the implementation of step S3, real-time monitoring is combined with personalized care, which not only improves the accuracy and timeliness of care, but also enhances the personalized experience of patients. The entire process optimizes the care plan in a data-driven way, ensuring that patients receive the most suitable support for their disease development throughout the care process. This patient-centered care model not only improves patient satisfaction.

[0055] In this specification, a system for constructing a brain disease nursing knowledge base is provided, which is used to execute the above-mentioned method for constructing a brain disease nursing knowledge base. The system for constructing a brain disease nursing knowledge base includes:

[0056] The structured stratification module is used to obtain multi-source encephalopathy data; perform structured data screening on multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data; perform multi-modal data structured stratification on unstructured encephalopathy data to generate multi-modal structured stratified data; perform data preprocessing on multi-modal structured stratified data and structured encephalopathy data to generate a standard structured encephalopathy data set;

[0057] The relational network modeling module is used to perform context analysis on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information; perform redundant relation pruning on the semantic vector set of encephalopathy information to generate spatial semantic optimization data; perform collaborative reasoning modeling on the spatial semantic optimization data to generate a distributed encephalopathy collaborative association network; perform dynamic mapping of encephalopathy nursing needs on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate dynamic nursing needs data;

[0058] The dynamic change detection module is used to obtain the user's real-time encephalopathy physiological monitoring data; perform abnormal monitoring data screening on the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data; perform nursing implementation data collection on the dynamic nursing demand data according to the user's real-time encephalopathy abnormal physiological monitoring data to obtain the user's real-time encephalopathy physiological monitoring data after implementation; perform dynamic change data screening on the user's real-time encephalopathy physiological monitoring data after implementation and the user's real-time encephalopathy abnormal physiological monitoring data to generate personalized encephalopathy nursing change data;

[0059] The database construction module is used to conduct brain disease nursing feedback evaluation on personalized brain disease nursing change data to generate brain disease nursing feedback evaluation data; and to construct a database for the brain disease nursing feedback evaluation data to generate a brain disease nursing knowledge base.

[0060] The beneficial effect of the present invention is that by acquiring multi-source encephalopathy data, different types of information can be collected to provide a comprehensive patient view. The separation of structured data and unstructured data is helpful for subsequent processing, making the information more orderly and easy to analyze. By multimodal structured stratification of unstructured data, potential valuable information such as doctor-patient communication records, imaging data, etc. can be mined. Data preprocessing generates a standard structured encephalopathy data set, which improves data quality and ensures the reliability of subsequent analysis. By generating a semantic vector set of encephalopathy information through context analysis, the understanding of encephalopathy information is enhanced to ensure relevance. The optimized spatial semantic data reduces information redundancy and improves data processing efficiency. The generated distributed encephalopathy collaborative association network promotes the identification of relationships between data and supports intelligent decision-making and reasoning. Dynamic mapping based on collaborative association networks enables nursing needs to reflect patient status in real time and improve nursing adaptability. Obtaining real-time physiological monitoring data of encephalopathy from users provides an important basis for personalized nursing and improves nursing quality. Screening out abnormal monitoring data helps to promptly detect changes in the patient's condition and ensure rapid response. Collecting nursing implementation data based on abnormal data can evaluate the effectiveness of nursing measures. Personalized nursing change data is generated by dynamically changing data screening to provide data support for the formulation of individualized nursing strategies. Feedback evaluation of personalized nursing change data helps to understand the effect of nursing measures and promote continuous improvement. Establishing a knowledge base for brain disease nursing allows experience and data to be preserved and shared for a long time, providing a reference for future nursing work. By building a knowledge base, nursing experience can be accumulated, the inheritance and application of knowledge can be promoted, and the level of nursing expertise can be improved. Therefore, the present invention improves the construction quality and practicality of the knowledge base for brain disease nursing by integrating multi-source data, structured unstructured information, real-time dynamic mapping of nursing needs and system feedback evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A schematic diagram of the steps of a method for constructing a knowledge base for brain disease nursing is shown;

[0062] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0063] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0064] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0065] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0066] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0067] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0068] To achieve this, please refer to Figures 1 to 3 , a method for constructing a brain disease nursing knowledge base, the method comprising the following steps:

[0069] Step S1: Acquire multi-source encephalopathy data; perform structured data screening on the multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data; perform multi-modal data structured stratification on the unstructured encephalopathy data to generate multi-modal structured stratified data; perform data preprocessing on the multi-modal structured stratified data and structured encephalopathy data to generate a standard structured encephalopathy data set;

[0070] Step S2: Perform context analysis on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information; perform redundant relationship pruning on the semantic vector set of encephalopathy information to generate spatial semantic optimization data; perform collaborative reasoning modeling on the spatial semantic optimization data to generate a distributed encephalopathy collaborative association network; perform dynamic mapping of encephalopathy nursing needs on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate dynamic nursing needs data;

[0071] Step S3: Acquire the user's real-time encephalopathy physiological monitoring data; perform abnormal monitoring data screening on the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data; perform nursing implementation data collection on dynamic nursing demand data according to the user's real-time encephalopathy abnormal physiological monitoring data to obtain the user's real-time encephalopathy physiological monitoring data after implementation; perform dynamic change data screening on the user's real-time encephalopathy physiological monitoring data after implementation and the user's real-time encephalopathy abnormal physiological monitoring data to generate personalized encephalopathy nursing change data;

[0072] Step S4: Conducting brain disease nursing feedback evaluation on the personalized brain disease nursing change data to generate brain disease nursing feedback evaluation data; constructing a database on the brain disease nursing feedback evaluation data to generate a brain disease nursing knowledge base.

[0073] The present invention obtains multi-source encephalopathy data. Multi-source data acquisition provides a comprehensive information foundation, which can analyze encephalopathy from different angles and enhance the depth and breadth of research. Structured data screening is performed on multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data. The extraction of structured data makes subsequent analysis more efficient, convenient for computer processing and analysis, and the value of unstructured data is mined through subsequent processing. Multi-modal data structured stratification is performed on unstructured encephalopathy data to generate multi-modal structured stratified data. The stratification of unstructured data helps to identify the relationship between different data modalities and improve the flexibility of data integration and analysis. Data preprocessing is performed on multi-modal structured stratified data and structured encephalopathy data to generate a standard structured encephalopathy data set. Preprocessing ensures the consistency and accuracy of the data, lays a good foundation for subsequent analysis, and reduces errors caused by data quality issues. Context analysis is performed on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information. Context analysis helps to extract implicit information in the data, generate a semantic vector set to capture encephalopathy characteristics, and facilitate subsequent deeper analysis. The redundant relationship pruning process is performed on the semantic vector set of encephalopathy information to generate spatial semantic optimization data. Pruning improves the availability of data, reduces noise, and improves the efficiency of subsequent analysis and modeling. Collaborative reasoning modeling is performed on the spatial semantic optimization data to generate a distributed encephalopathy collaborative association network. The collaborative association network can reveal the correlation between different factors of encephalopathy and provide data support for precision medicine. Dynamic mapping provides personalized nursing needs, can adjust the nursing plan according to the specific situation of the patient, and improve the pertinence and effectiveness of nursing. Real-time monitoring provides the latest information on the current patient status and supports timely nursing decisions. Timely identification of abnormal data can prompt nursing staff to respond quickly. The collection of post-implementation data provides a basis for subsequent effect evaluation and adjustment of nursing strategies. Dynamically changing data helps to evaluate the effectiveness of nursing measures and supports the adjustment and optimization of personalized nursing strategies. Feedback evaluation provides a reference for the effectiveness of subsequent nursing measures and promotes continuous improvement. The construction of the knowledge base provides valuable data resources for future research and practice, and promotes the accumulation and sharing of knowledge in the field of encephalopathy nursing. Therefore, the present invention improves the construction quality and practicality of the brain disease nursing knowledge base by integrating multi-source data, structured and unstructured information, real-time dynamic mapping of nursing needs and system feedback evaluation.

[0074] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of a method for constructing a brain disease nursing knowledge base of the present invention. In this example, the method for constructing a brain disease nursing knowledge base includes the following steps:

[0075] Step S1: Acquire multi-source encephalopathy data; perform structured data screening on the multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data; perform multi-modal data structured stratification on the unstructured encephalopathy data to generate multi-modal structured stratified data; perform data preprocessing on the multi-modal structured stratified data and structured encephalopathy data to generate a standard structured encephalopathy data set;

[0076] In an embodiment of the present invention, potential sources of multi-source encephalopathy data are determined, such as hospital information systems (HIS), clinical databases, public health data sets, scientific research literature, etc. Relevant data is extracted from these data sources using an API interface or database query. API can obtain real-time data, and database queries can extract historical data and structured data. The acquired data is divided into two categories: structured and unstructured. Structured data usually exists in a table form, while unstructured data includes text, images and other formats. Use specific rules (such as data types and formats) to identify and filter out structured data. For unstructured data, classify and identify for subsequent processing. Unstructured data of different sources and types (such as images, clinical texts, and biomarkers) are integrated to form a multimodal data set. Create a hierarchical data structure, and store image data, text data, and biomarker data at different levels for subsequent analysis and processing. Clean all data, including removing duplicate records and invalid data, to ensure data accuracy and consistency. Standardize numerical data to ensure that all numerical data use the same unit or format, such as uniformly converting age into years. De-noise the image data to improve image quality and prepare for subsequent analysis. For missing values ​​in the data set, use the mean, median or interpolation method to fill in to ensure data integrity.

[0077] Step S2: Perform context analysis on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information; perform redundant relationship pruning on the semantic vector set of encephalopathy information to generate spatial semantic optimization data; perform collaborative reasoning modeling on the spatial semantic optimization data to generate a distributed encephalopathy collaborative association network; perform dynamic mapping of encephalopathy nursing needs on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate dynamic nursing needs data;

[0078] In an embodiment of the present invention, by comprehensively understanding the standard structured encephalopathy data set, key features, variables and their interrelationships in the data are identified. Contextual information in the data, such as medical record text, treatment records and patient feedback, is extracted by natural language processing (NLP) technology. Keyword extraction, topic modeling and other technologies are used to identify the main themes and semantic content related to encephalopathy. The extracted contextual information is converted into semantic vectors using word embedding technology (such as Word2Vec, GloVe or BERT) for subsequent analysis and processing. The similarity between semantic vectors is analyzed to identify redundant or highly similar semantic information. The semantic vectors are clustered using a clustering algorithm (such as K-means, hierarchical clustering) to identify similar vectors. According to a preset similarity threshold, redundant semantic vectors are removed and representative semantic vectors are retained to generate spatial semantic optimization data. This process can be performed by calculating the cosine similarity between vectors. Based on the optimized spatial semantic data, a collaborative reasoning model is constructed, taking into account the relationship and mutual influence between each semantic. A graph neural network (GNN) or a Bayesian network is used to model the collaborative relationship between different encephalopathy information. In the collaborative reasoning model, each brain disease information is used as a node, and the relationship between nodes is used as an edge to construct a distributed brain disease collaborative association network. According to the node information in the distributed brain disease collaborative association network, the nursing needs of different brain disease patients are identified and analyzed. The information propagation mechanism in the network is used to derive the nursing needs association between different nodes. A dynamic nursing needs mapping model is created to generate dynamic nursing needs data through real-time data input and update. This process requires continuous updating of the model to adapt to new patient data and changes in needs.

[0079] Step S3: Acquire the user's real-time encephalopathy physiological monitoring data; perform abnormal monitoring data screening on the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data; perform nursing implementation data collection on dynamic nursing demand data according to the user's real-time encephalopathy abnormal physiological monitoring data to obtain the user's real-time encephalopathy physiological monitoring data after implementation; perform dynamic change data screening on the user's real-time encephalopathy physiological monitoring data after implementation and the user's real-time encephalopathy abnormal physiological monitoring data to generate personalized encephalopathy nursing change data;

[0080] In an embodiment of the present invention, by selecting appropriate physiological monitoring devices, such as smart bracelets, electrocardiogram (ECG) monitors, electroencephalogram (EEG) devices, etc., these devices can collect the user's physiological data in real time. The collected physiological monitoring data is transmitted to a central database or cloud platform in real time by means of Bluetooth, Wi-Fi or mobile data to ensure the timeliness and accuracy of the data. Use machine learning or statistical analysis techniques (such as control charts, Z-scores, etc.) to monitor abnormal conditions of real-time physiological data. Define abnormal standards, such as physiological parameters such as heart rate, blood pressure, and body temperature that are beyond the normal range. Screen the monitored physiological data, extract abnormal physiological monitoring data, and record the user's real-time abnormal state. According to the user's abnormal physiological monitoring data, evaluate their nursing needs, including first aid measures, nursing interventions, and follow-up observations. According to the evaluation results, implement corresponding nursing measures, and record the user's physiological monitoring data during the implementation process. Record the user's physiological monitoring data after implementation to form real-time encephalopathy physiological monitoring data after implementation. Compare the user's real-time encephalopathy physiological monitoring data after implementation with the user's real-time encephalopathy abnormal physiological monitoring data previously screened, and analyze the changes in the two sets of data. According to the comparison results, the data reflecting the changes in the user's physiological state are screened out to generate personalized brain disease care change data. Focus on the improvement or deterioration of physiological indicators, such as heart rate returning to normal, blood pressure changes, etc.

[0081] Step S4: Conducting brain disease nursing feedback evaluation on the personalized brain disease nursing change data to generate brain disease nursing feedback evaluation data; constructing a database on the brain disease nursing feedback evaluation data to generate a brain disease nursing knowledge base.

[0082] In an embodiment of the present invention, by determining indicators for evaluating personalized encephalopathy nursing change data, these indicators may include: improvement of patient physiological indicators (such as heart rate, blood pressure, body temperature, etc.), improvement of patient self-perception and quality of life (collected through questionnaires), and the effect of nursing intervention (such as the timeliness and effectiveness of nursing measures). Use statistical analysis tools (such as SPSS, R, Python) to evaluate personalized nursing change data: use descriptive statistics to analyze the change trend of each indicator. Use inferential statistical methods (such as t-test, variance analysis) to test the significant difference before and after nursing measures. According to the analysis results, generate encephalopathy nursing feedback evaluation data, including evaluation results, improvement suggestions and follow-up nursing plans. Design a structured database to store encephalopathy nursing feedback evaluation data. Database design should include the following elements: Table structure: Create a table to store basic patient information, nursing change data, feedback evaluation results, etc. Relationship design: Ensure that there is a reasonable association between tables to facilitate data retrieval and analysis. Use data import tools or scripts (such as ETL tools) to migrate data into the database to ensure data integrity and consistency. Regularly maintain and update the database to ensure that it reflects the latest nursing feedback evaluation information. Extract valuable information and patterns from the constructed database, including: common nursing interventions and their effects, nursing problems and suggestions mentioned in patient feedback, and the correlation between nursing changes and patient physiological indicators. Organize the extracted information into a knowledge base, including: organize knowledge items by topic or category (such as drug care, rehabilitation training, lifestyle guidance, etc.). Link knowledge items with relevant literature, case studies, or clinical guidelines to form a complete knowledge system. Establish a knowledge base update mechanism to continuously supplement and revise the knowledge base content based on new nursing feedback and clinical practice.

[0083] Preferably, step S1 comprises the following steps:

[0084] Step S11: Acquire multi-source brain disease data;

[0085] Step S12: screening the multi-source encephalopathy data by structured data to generate structured encephalopathy data and unstructured encephalopathy data;

[0086] Step S13: dividing the unstructured encephalopathy data into data types to generate text-type encephalopathy data and image-type encephalopathy data; performing multimodal data structured stratification on the text-type encephalopathy data and image-type encephalopathy data to generate multimodal structured stratified data;

[0087] Step S14: Integrate the multimodal structured hierarchical data and the structured encephalopathy data to generate a structured encephalopathy data set; perform data preprocessing on the structured encephalopathy data set to generate a standard structured encephalopathy data set, wherein the data preprocessing includes data cleaning, data missing value filling and data standardization.

[0088] In an embodiment of the present invention, the sources of multi-source encephalopathy data are determined, including electronic medical records (EMR), hospital information systems (HIS), laboratory test data, imaging data (such as MRI, CT), doctor reports and patient monitoring data. Data interfaces (such as HL7, FHIR) or ETL tools are used to collect data from different data sources and uniformly transmit them to a central database or data warehouse. The structured features of the data are identified, and the formatted encephalopathy data (such as numerical laboratory test data, text-based doctor diagnosis records) are classified as structured data. The parts of the data without a fixed format (such as images, doctor's handwritten records, and patient self-reported symptoms) are classified as unstructured data. According to the screening results, the structured data and unstructured data are stored separately, and the contents such as doctor's handwritten records, patient self-reports, and unstructured diagnostic texts are extracted. Image data such as MRI and CT are extracted for image separation and annotation. The text-type encephalopathy data is processed using natural language processing (NLP) methods, including word segmentation, entity recognition, keyword extraction, etc. Image segmentation, edge detection and other computer vision technologies are used to process the images in layers (such as separating the image layers of different brain regions). Different contents (such as symptoms, time, and medication records) are layered and classified to generate a structured representation of the text. The images are structured and layered according to features such as lesion area, brain region, and lesion density. The layered text and image data are integrated to form multimodal structured layered data for subsequent data analysis and processing. The multimodal structured layered data are matched and mapped with the structured data to ensure the association between different data (such as the association between the text description of a specific disease and the corresponding image data). The integrated data is stored as a structured data set to form a unified encephalopathy data set. Duplicate data, outliers, and null values ​​are removed to ensure data quality. Mean filling, interpolation, or machine learning prediction are used to fill missing data to complete the data set. Numerical data are normalized or standardized to make the data on the same scale for easy analysis and modeling. After integration and standardization, a standard structured encephalopathy data set suitable for subsequent analysis is finally generated to ensure the uniformity and consistency of different data sources.

[0089] Preferably, multimodal data structured stratification of text-type encephalopathy data and image-type encephalopathy data includes:

[0090] Performing feature classification indexing on text-type encephalopathy data and image-type encephalopathy data to obtain a text-type feature index and an image-type feature index;

[0091] The text type feature index is used to extract word features from the text type encephalopathy data to obtain the encephalopathy word vector; the image type feature index is used to extract embedding features from the image type encephalopathy data to obtain the encephalopathy embedding vector;

[0092] Matching feature points of brain disease word vectors and brain disease embedding vectors to generate brain disease feature point matching data; hierarchically fusing feature maps of brain disease word vectors and brain disease embedding vectors based on the brain disease feature point matching data to generate a multi-dimensional feature hierarchical table;

[0093] A standardized matrix was constructed for text-type encephalopathy data and imaging-type encephalopathy data according to the multidimensional feature hierarchy table to generate multimodal structured hierarchical data.

[0094] In an embodiment of the present invention, a text feature classification table is generated by identifying the main features in the text data (such as disease name, time stamp, symptom description, medication record, etc.). An index is generated for each feature to facilitate the rapid location of the feature position and its corresponding context content. The main feature areas in the image data (such as images of specific brain areas, density features of lesion areas, boundary shapes, etc.) are identified. A position index is generated for the features in the image to ensure the relevance of the image and the text description. An index mapping table is created to uniformly manage the text and image feature indexes, thereby facilitating subsequent feature extraction and matching. The text data is segmented and tagged using natural language processing (NLP) methods, and relevant word features are extracted from the data using feature indexes. The extracted text features are converted into numerical brain disease word vectors using word vector models (such as Word2Vec, BERT) to facilitate subsequent multimodal data fusion. The image data is embedded and extracted using a deep learning model (such as CNN), and image features (such as shape, density, and boundaries) are converted into embedded vectors. The extracted image embedding features are digitized to form brain disease embedding vectors for matching and fusion with text word vectors. The feature similarity between the encephalopathy word vector and the encephalopathy embedding vector is calculated using methods such as cosine similarity or Euclidean distance. A feature matching threshold is set to pair text features and image features with similarity above the threshold. The matching results are recorded in a feature point matching table, including the text feature position, image feature position and similarity score of each pair of matching features. Based on the feature point matching data, similar feature points are clustered to form feature groups. A hierarchical feature graph is constructed based on the feature group, and feature points at each level are associated with their semantic and spatial relationships. The hierarchical feature graph is organized into a multidimensional feature hierarchy table to store the hierarchy and positional relationship of multimodal features. Based on the multidimensional feature hierarchy table, the structure of the standardized matrix is ​​designed (e.g., rows and columns represent different feature types and hierarchies). The encephalopathy word vector and encephalopathy embedding vector are filled into the matrix according to the hierarchical relationship. A unified structured representation of text and image data is achieved through a standardized matrix to facilitate subsequent analysis and processing.

[0095] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0096] Step S21: performing semantic recognition on a standard structured encephalopathy data set to generate a structured encephalopathy semantic vector; performing context analysis on the standard structured encephalopathy data set based on the structured encephalopathy semantic vector to generate a encephalopathy information semantic vector set;

[0097] Step S22: performing hierarchical semantic relationship modeling on the semantic vector set of brain disease information to generate spatial semantic modeling data; performing redundant relationship pruning processing on the spatial semantic modeling data to generate spatial semantic optimization data;

[0098] Step S23: constructing a cross-node knowledge association network for the spatial semantic optimization data to generate a cross-node brain disease association network; performing collaborative reasoning modeling on the cross-node brain disease association network to generate a distributed brain disease collaborative association network;

[0099] Step S24: Based on the distributed brain disease collaborative association network, the standard structured brain disease data set is predicted to evolve complex diseases, and disease evolution prediction data is generated; the brain disease nursing needs are dynamically mapped to the disease evolution prediction data, and dynamic nursing needs data is generated.

[0100] In an embodiment of the present invention, the descriptive information in the structured encephalopathy data set is analyzed by using natural language processing (NLP) technology (such as named entity recognition, dependency analysis, etc.), and semantic features containing key contents such as condition, symptoms, and treatment plans are extracted. The extracted semantic features are converted into structured encephalopathy semantic vectors through word embedding models (such as Word2Vec, BERT) to represent different condition information. In the encephalopathy data, a context window (such as the 5 words before and after) is used to identify relevant semantic words to ensure the correlation between different semantic vectors. Based on the results of the context analysis, the semantic vectors are aggregated using a clustering method (such as K-means) to form a semantic vector set of encephalopathy information, which contains a variety of context information. The semantic vector set of encephalopathy information is layered according to different categories such as condition, symptoms, and treatment to construct a hierarchical semantic model. The semantic vector set is represented as nodes and edges using a graph structure, and the relationship between different semantic concepts is represented by edges. The data of the hierarchical semantic model is converted into a graph structure for subsequent spatial semantic optimization and network construction. Redundancy analysis is performed on the spatial semantic modeling data to detect repeated and unnecessary semantic associations. Remove low-correlation edges, optimize the spatial semantic modeling graph structure, and generate spatial semantic optimization data to reduce data redundancy and improve data efficiency. Associate spatial semantic optimization data on different nodes, connect scattered semantic information into an overall network through knowledge graph technology (such as RDF and OWL), generate a cross-node brain disease association network, and include cross-platform association of disease information to achieve distributed reasoning. Define reasoning rules based on the cross-node brain disease association network, and incorporate information such as diagnosis and treatment information, disease development trends, and nursing needs into the reasoning process. Use logical reasoning and machine learning methods to generate a distributed brain disease collaborative association network through cross-node collaborative reasoning to connect the associations between different disease data. Use time series analysis models (such as LSTM, ARIMA) to predict the evolution of complex diseases on standard structured brain disease data sets. Generate disease evolution prediction data based on the prediction results, covering the disease progression trend. According to the disease evolution prediction data, dynamically match the predicted disease change trend with nursing needs to generate dynamic nursing needs data, including specific nursing measures, frequency, and matters of concern, to provide data support for personalized care.

[0101] Preferably, performing redundant relationship pruning processing on spatial semantic modeling data includes:

[0102] Perform semantic node recognition on the spatial semantic modeling data to obtain spatial semantic node recognition data; perform adjacent node relationship density calculation on the spatial semantic modeling data according to the spatial semantic node recognition data to obtain adjacent node relationship density data; perform redundant relationship node analysis on the spatial semantic node recognition data using the adjacent node relationship density data to generate redundant relationship node data;

[0103] Score the spatial semantic node recognition data for relationship importance to generate relationship semantic importance score data; sort the relationship semantic importance score data by pruning priority through redundant relationship node data to generate pruning priority score data; calculate the information entropy value of redundant relationship node data by pruning priority score data to obtain the redundant relationship node information entropy value;

[0104] The information entropy value of the redundant relationship node is compared with the preset standard information entropy value threshold. When the information entropy value of the redundant relationship node is less than the preset standard information entropy value threshold, the corresponding redundant relationship node data is redundantly pruned to generate spatial semantic optimization data.

[0105] In an embodiment of the present invention, the nodes in the spatial semantic modeling data are analyzed by using a feature extraction algorithm, the features of each node are identified and classified into different types of semantic nodes, and spatial semantic node identification data is generated for further analysis of adjacent node relationships. Based on the spatial semantic node identification data, the adjacency relationship between the nodes is determined. A density calculation formula (such as adjacency degree or similarity index) is used to quantify the density of adjacent node relationships, and adjacent node relationship density data is generated to evaluate the closeness and association of the nodes. Using the adjacent node relationship density data, redundant relationship nodes, that is, node relationships with low density are identified by setting a density threshold. Node relationships with low density values ​​are extracted and identified as redundant node relationships to form redundant relationship node data, which provides input for subsequent pruning processing. The spatial semantic node identification data is scored for importance, the importance of each node relationship is evaluated, and the impact of the relationship on the overall model is calculated. Relationship semantic importance scoring data is generated based on the scoring results, so that node relationships with high importance are preferentially retained in redundant nodes. Priority is sorted according to importance based on the redundant relationship node data and the relationship semantic importance scoring data. The node relationships are sorted to generate pruning priority scoring data to ensure that low-importance node relationships are removed preferentially. According to the pruning priority scoring data, the information entropy value of the redundant relationship node data is calculated, the redundancy of each node relationship is quantified, and the information entropy value of each redundant relationship node is obtained to evaluate the redundancy of the relationship. The information entropy value of the redundant relationship node is compared with the preset standard information entropy value threshold. If the information entropy value of the redundant relationship node is lower than the standard threshold, it is judged as a redundant relationship node. The redundant relationship nodes that meet the conditions are pruned to remove the node relationships with low redundancy and generate optimized spatial semantic data. Finally, the spatial semantic optimization data is output to provide an efficient structured semantic model for downstream applications.

[0106] Preferably, step S23 includes the following steps:

[0107] Step S231: performing cross-node association mining on the spatial semantic optimization data to generate cross-node association data; performing semantic aggregation on the cross-node association data to generate a preliminary cross-node knowledge association network;

[0108] Step S232: performing node association relationship weight distribution on the cross-node knowledge association network to generate cross-node association relationship weights; performing multi-layer relationship weight optimization on the cross-node knowledge association network according to the cross-node association relationship weights to generate a multi-layer weighted association network;

[0109] Step S233: performing connection enhancement on the multi-layer weighted association network to generate a cross-node brain disease association network;

[0110] Step S234: Perform multi-node information joint reasoning analysis on the cross-node brain disease association network to generate multi-node information joint reasoning data; perform distributed reasoning node optimization on the cross-node brain disease association network based on the multi-node information joint reasoning data to generate a distributed brain disease collaborative association network.

[0111] In an embodiment of the present invention, a cross-node association relationship analysis is performed on the spatial semantic optimization data to mine the potential associations between different nodes and generate cross-node association relationship data. This process usually uses a graph mining algorithm or a similarity-based matching algorithm to identify the relationship between different nodes. The cross-node association relationship data is aggregated to form a preliminary cross-node knowledge association network. During the aggregation process, the association relationships are classified and layered according to the similarity of the node semantic content, so that the network can preliminarily reflect the distribution and association of knowledge. Weights are assigned to the association relationships in the cross-node knowledge association network. The weight assignment is based on parameters such as the strength of the association relationship and the importance of the node, and the cross-node association relationship weight data is generated through a weight assignment algorithm (such as a weight calculation based on the edge frequency or the number of interactions). Based on the weight assignment data, the cross-node knowledge association network is multi-level weighted optimized so that the network can better reflect the priority of different association relationships. A multi-layer weighted association network with hierarchical characteristics is generated through a hierarchical optimization strategy (such as a multi-layer network decomposition algorithm). In the multi-layer weighted association network, the association connection between nodes is strengthened to improve the connectivity and robustness of the network. There are many strategies for connection enhancement, such as adding redundant paths to ensure the effectiveness of information transmission between nodes, or enhancing connection algorithms based on node similarity to ensure that the network can maintain stable node associations under different circumstances. The network after connection enhancement is tested to ensure the integrity and stability of the network structure, and finally a cross-node brain disease association network is obtained. Based on the cross-node brain disease association network, multi-node joint reasoning analysis is performed. Joint reasoning improves the network's association reasoning ability by integrating multiple node information and generating multi-node information joint reasoning data. The joint reasoning process usually uses methods based on logistic regression, neural networks, etc., and derives higher-level association relationships through collaborative analysis of information between nodes. Based on the multi-node information joint reasoning data, the reasoning nodes in the network are distributedly optimized. The optimization includes load distribution of reasoning nodes, data flow adjustment, and correlation control to ensure the efficiency of reasoning in a large-scale network environment. The optimized distributed reasoning node data is integrated into the entire association network, and finally a distributed brain disease collaborative association network is generated, forming a collaborative association system with high scalability and real-time performance.

[0112] Preferably, performing multi-node information joint reasoning analysis on a cross-node brain disease association network includes:

[0113] Aggregating node data of the cross-node brain disease association network to generate cross-node brain disease association aggregated data; extracting key features of the cross-node brain disease association aggregated data to obtain cross-node brain disease association key feature data, wherein the key feature extraction includes mean calculation, median calculation and variance calculation;

[0114] By using the cross-node correlation key feature data, the Pearson coefficient between nodes of the cross-node brain disease correlation network is calculated to obtain the cross-node node similarity data; the correlation analysis is performed on the cross-node node similarity data to generate a node similarity matrix;

[0115] According to the node similarity matrix, the rule-based reasoning relationship of the cross-node brain disease association network is established to generate a rule-based reasoning set; based on the rule-based reasoning set, information reasoning is performed on the cross-node brain disease association network to generate a reasoning result set; the reasoning result set is logically deduced to generate multi-node information joint reasoning data.

[0116] In an embodiment of the present invention, cross-node brain disease associated aggregated data is generated by selecting associated nodes from a cross-node brain disease associated network and integrating the data of these nodes. The aggregation process is based on the attributes and data structure of the nodes to ensure the integrity of the aggregated data. Feature analysis is performed on the cross-node brain disease associated aggregated data to extract key features. Key feature extraction includes the following steps: calculating the mean of each node data to obtain a data center trend. Extracting median information to ensure anti-interference to outliers. Calculating the variance of the node data to evaluate the discreteness and fluctuation characteristics of the data. Finally, cross-node brain disease associated key feature data containing mean, median and variance are obtained to reflect the basic feature differences between nodes. Using the cross-node brain disease associated key feature data, the Pearson coefficient is calculated between different nodes of the cross-node brain disease associated network to obtain cross-node node similarity data. The Pearson coefficient reflects the linear relationship between node data, and the numerical range is between -1 and 1, which is used to measure the similarity of node data. Based on the calculated node similarity data, a node similarity matrix is ​​established to matrix the similarity relationship of each node. The similarity matrix is ​​used to systematically represent the similarity association between nodes, which is convenient for subsequent analysis and reasoning. According to the node similarity matrix, the rule-based reasoning relationship is constructed for the cross-node brain disease association network. The establishment of the rule-based reasoning set is based on preset rules, such as similarity threshold setting or specific node association logic, so as to systematically deduce the reasoning relationship between nodes. Based on the rule-based reasoning set, the information reasoning process is executed to generate the reasoning result set. During the reasoning process, the node data is deduced and analyzed through the rule relationship to obtain the potential association information of each node in the association network. The logical deduction strategy is applied to the reasoning result set to further refine the associated reasoning relationship and exclude illogical reasoning results. After logical deduction, the final multi-node information joint reasoning data is obtained as an important reasoning basis in the cross-node brain disease association network. This data reflects the deep information association between multiple nodes and provides key reasoning support for the distributed brain disease collaborative association network.

[0117] Preferably, step S24 includes the following steps:

[0118] Step S241: performing fuzzy reasoning on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate a encephalopathy fuzzy reasoning result data set; performing identical result data screening on the encephalopathy fuzzy reasoning result data set to obtain identical reasoning result data;

[0119] Step S242: Divide the data set into a data set with the same inference result data to generate a model training set and a model test set; perform model training on the model training set using a long short-term memory neural network algorithm to generate a complex disease evolution prediction pre-model; perform model optimization iteration on the complex disease evolution prediction pre-model using the model test set to generate a complex disease evolution prediction model;

[0120] Step S243: Import the same inference result data into the complex disease evolution prediction model to perform complex disease evolution prediction and generate disease evolution prediction data; dynamically map brain disease nursing needs to the disease evolution prediction data and generate dynamic nursing needs data.

[0121] In an embodiment of the present invention, fuzzy reasoning is performed on a standard structured encephalopathy data set based on a distributed encephalopathy collaborative association network to generate a fuzzy reasoning result data set for encephalopathy. A fuzzy logic system is used to define a fuzzy set of key features (such as fuzzy states such as "mild", "moderate", and "severe" of symptom severity). The feature data is reasoned according to fuzzy rules to generate fuzzy reasoning results covering different disease states. Data with consistent reasoning values ​​are extracted from the fuzzy reasoning result data set to form a data set with the same reasoning result. The fuzzy reasoning result data is clustered according to the reasoning value, and high-frequency reasoning results are screened out to ensure the stability and consistency of the reasoning results. The same reasoning result data set is randomly divided into a model training set and a model test set to ensure that both the training and test sets contain representative data features. The model training set is trained using an LSTM neural network model. The disease time series data is input into the LSTM network, model parameters such as the learning rate and time step are set, and the model weights are adjusted using the feedforward and backpropagation algorithms. The error value is gradually reduced through multiple rounds of training to generate a complex disease evolution prediction pre-model. The model test set is input into the complex disease evolution prediction pre-model to calculate the prediction accuracy. Adjust model parameters (such as learning rate, number of hidden layer nodes, etc.) according to the test results, optimize the model structure through multiple rounds of iterations until the preset accuracy threshold is reached on the test set, and generate the final complex disease evolution prediction model. Import the same inference result data into the optimized complex disease evolution prediction model. Use the LSTM model to perform time series prediction on the evolution trend of the disease and generate disease evolution prediction data. The model will output the state of disease change in the future. Perform trend analysis on the disease evolution prediction data to identify the key change nodes of disease deterioration or improvement. Based on the type of disease change, convert the trend analysis results into corresponding nursing needs (for example, the trend of disease deterioration indicates enhanced monitoring, and stable disease indicates routine care), and generate nursing resource allocation recommendations, including the required nursing staff and nursing frequency.

[0122] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0123] Step S31: Acquire real-time encephalopathy physiological monitoring data of the user;

[0124] Step S32: Screening the abnormal monitoring data of the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data;

[0125] Step S33: Generate a personalized encephalopathy nursing plan for the dynamic nursing demand data according to the user's real-time abnormal encephalopathy physiological monitoring data to obtain a personalized encephalopathy nursing plan;

[0126] Step S34: Implement the nursing plan on the user's real-time brain disease physiological monitoring data through a personalized brain disease nursing plan, and synchronously collect the user's real-time brain disease physiological monitoring data after implementation; dynamically screen the user's real-time brain disease physiological monitoring data and the user's real-time brain disease abnormal physiological monitoring data after implementation to generate personalized brain disease nursing change data.

[0127] In an embodiment of the present invention, physiological monitoring data of encephalopathy is collected in real time from the physiological monitoring device of the user, and the data types include heart rate, blood oxygen, brain waves, respiratory rate, etc. The latest monitoring data is obtained through the data acquisition interface and transmitted to the central processing system to ensure the stability and real-time performance of data transmission. The collected physiological monitoring data is formatted and converted into a unified structured format to facilitate subsequent analysis and processing. The processed data is stored in a real-time data monitoring database to support continuous monitoring and analysis. According to medical research and the characteristics of the disease, a normal value range and an alarm threshold (such as the normal range of heart rate, blood oxygen level, etc.) are set for each monitoring indicator. Physiological data is monitored in real time, and data exceeding the threshold is automatically screened out, and its changes are recorded. For monitoring data with abnormal fluctuations, a user's real-time abnormal physiological monitoring data set for encephalopathy is generated, including data timestamps and abnormal types. Based on the user's abnormal monitoring data, matching analysis is performed with preset dynamic nursing demand data. Assess the user's current nursing needs and analyze nursing intervention measures (such as drug adjustment, environmental optimization, mental support, etc.). According to the results of the nursing needs analysis, a personalized encephalopathy nursing plan that meets the user's current health status is generated. The nursing plan includes the content of nursing measures, implementation frequency, nursing staff recommendations, and corresponding observation indicators and emergency measures. The personalized nursing plan is transmitted to the nursing team or equipment to start the implementation process of the nursing measures. Targeted adjustments are made to relevant indicators in the real-time physiological monitoring data (for example, adjusting the monitoring frequency or increasing drug dosage monitoring according to the plan). During the implementation of the nursing plan, the user's brain disease physiological monitoring data is continuously collected to observe the nursing effect. The real-time data after implementation is compared and analyzed with the abnormal data before the nursing plan to monitor the dynamic changes in the nursing effect. By comparing the real-time data after implementation with the abnormal monitoring data, the data features reflecting the nursing effect are extracted to generate personalized brain disease nursing change data, including nursing effect evaluation data, change trend data, and potential adjustment suggestions, to provide feedback reference for follow-up care.

[0128] In this specification, a system for constructing a brain disease nursing knowledge base is provided, which is used to execute the above-mentioned method for constructing a brain disease nursing knowledge base. The system for constructing a brain disease nursing knowledge base includes:

[0129] The structured stratification module is used to obtain multi-source encephalopathy data; perform structured data screening on multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data; perform multi-modal data structured stratification on unstructured encephalopathy data to generate multi-modal structured stratified data; perform data preprocessing on multi-modal structured stratified data and structured encephalopathy data to generate a standard structured encephalopathy data set;

[0130] The relational network modeling module is used to perform context analysis on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information; perform redundant relation pruning on the semantic vector set of encephalopathy information to generate spatial semantic optimization data; perform collaborative reasoning modeling on the spatial semantic optimization data to generate a distributed encephalopathy collaborative association network; perform dynamic mapping of encephalopathy nursing needs on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate dynamic nursing needs data;

[0131] The dynamic change detection module is used to obtain the user's real-time encephalopathy physiological monitoring data; perform abnormal monitoring data screening on the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data; perform nursing implementation data collection on the dynamic nursing demand data according to the user's real-time encephalopathy abnormal physiological monitoring data to obtain the user's real-time encephalopathy physiological monitoring data after implementation; perform dynamic change data screening on the user's real-time encephalopathy physiological monitoring data after implementation and the user's real-time encephalopathy abnormal physiological monitoring data to generate personalized encephalopathy nursing change data;

[0132] The database construction module is used to conduct brain disease nursing feedback evaluation on personalized brain disease nursing change data to generate brain disease nursing feedback evaluation data; and to construct a database for the brain disease nursing feedback evaluation data to generate a brain disease nursing knowledge base.

[0133] The beneficial effect of the present invention is that by acquiring multi-source encephalopathy data, different types of information can be collected to provide a comprehensive patient view. The separation of structured data and unstructured data is helpful for subsequent processing, making the information more orderly and easy to analyze. By multimodal structured stratification of unstructured data, potential valuable information such as doctor-patient communication records, imaging data, etc. can be mined. Data preprocessing generates a standard structured encephalopathy data set, which improves data quality and ensures the reliability of subsequent analysis. By generating a semantic vector set of encephalopathy information through context analysis, the understanding of encephalopathy information is enhanced to ensure relevance. The optimized spatial semantic data reduces information redundancy and improves data processing efficiency. The generated distributed encephalopathy collaborative association network promotes the identification of relationships between data and supports intelligent decision-making and reasoning. Dynamic mapping based on collaborative association networks enables nursing needs to reflect patient status in real time and improve nursing adaptability. Obtaining real-time physiological monitoring data of encephalopathy from users provides an important basis for personalized nursing and improves nursing quality. Screening out abnormal monitoring data helps to promptly detect changes in the patient's condition and ensure rapid response. Collecting nursing implementation data based on abnormal data can evaluate the effectiveness of nursing measures. Personalized nursing change data is generated by dynamically changing data screening to provide data support for the formulation of individualized nursing strategies. Feedback evaluation of personalized nursing change data helps to understand the effect of nursing measures and promote continuous improvement. Establishing a knowledge base for brain disease nursing allows experience and data to be preserved and shared for a long time, providing a reference for future nursing work. By building a knowledge base, nursing experience can be accumulated, the inheritance and application of knowledge can be promoted, and the level of nursing expertise can be improved. Therefore, the present invention improves the construction quality and practicality of the knowledge base for brain disease nursing by integrating multi-source data, structured unstructured information, real-time dynamic mapping of nursing needs and system feedback evaluation.

[0134] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0135] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for constructing a brain disease nursing knowledge base, characterized in that: The following steps are involved: Step S1: Acquire multi-source encephalopathy data; perform structured data screening on the multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data; Perform multimodal data structured stratification on unstructured brain disease data to generate multimodal structured stratified data; Preprocess the multimodal structured hierarchical data and structured encephalopathy data to generate a standard structured encephalopathy dataset; Step S2: Perform context analysis on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information; perform redundant relationship pruning on the semantic vector set of encephalopathy information to generate spatial semantic optimization data; Conduct collaborative reasoning modeling on spatial semantically optimized data to generate a distributed brain disease collaborative association network; Dynamic mapping of encephalopathy nursing needs to standard structured encephalopathy data sets is performed based on a distributed encephalopathy collaborative association network to generate dynamic nursing needs data; step S2 includes the following steps: Step S21: performing semantic recognition on a standard structured encephalopathy data set to generate a structured encephalopathy semantic vector; performing context analysis on the standard structured encephalopathy data set based on the structured encephalopathy semantic vector to generate a encephalopathy information semantic vector set; Step S22: performing hierarchical semantic relationship modeling on the semantic vector set of brain disease information to generate spatial semantic modeling data; performing redundant relationship pruning processing on the spatial semantic modeling data to generate spatial semantic optimization data; Step S23: constructing a cross-node knowledge association network for the spatial semantic optimization data to generate a cross-node brain disease association network; performing collaborative reasoning modeling on the cross-node brain disease association network to generate a distributed brain disease collaborative association network; Step S24: performing complex disease evolution prediction on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate disease evolution prediction data; dynamically mapping encephalopathy nursing needs on the disease evolution prediction data to generate dynamic nursing needs data; Step S3: Acquire the user's real-time encephalopathy physiological monitoring data; perform abnormal monitoring data screening on the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data; perform nursing implementation data collection on dynamic nursing demand data according to the user's real-time encephalopathy abnormal physiological monitoring data to obtain the user's real-time encephalopathy physiological monitoring data after implementation; perform dynamic change data screening on the user's real-time encephalopathy physiological monitoring data after implementation and the user's real-time encephalopathy abnormal physiological monitoring data to generate personalized encephalopathy nursing change data; Step S4: Conducting brain disease nursing feedback evaluation on the personalized brain disease nursing change data to generate brain disease nursing feedback evaluation data; constructing a database on the brain disease nursing feedback evaluation data to generate a brain disease nursing knowledge base.

2. The method for constructing a brain disease nursing knowledge base according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Acquire multi-source brain disease data; Step S12: screening the multi-source encephalopathy data by structured data to generate structured encephalopathy data and unstructured encephalopathy data; Step S13: dividing the unstructured encephalopathy data into data types to generate text-type encephalopathy data and image-type encephalopathy data; performing multimodal data structured stratification on the text-type encephalopathy data and image-type encephalopathy data to generate multimodal structured stratified data; Step S14: Integrate the multimodal structured hierarchical data and the structured encephalopathy data to generate a structured encephalopathy data set; perform data preprocessing on the structured encephalopathy data set to generate a standard structured encephalopathy data set, wherein the data preprocessing includes data cleaning, data missing value filling and data standardization.

3. The method for constructing a brain disease nursing knowledge base according to claim 2, characterized in that: The multimodal data structured stratification of text-type encephalopathy data and image-type encephalopathy data includes: Performing feature classification indexing on the text-type encephalopathy data and the image-type encephalopathy data to obtain a text-type feature index and an image-type feature index; The text type feature index is used to extract word features from the text type encephalopathy data to obtain the encephalopathy word vector; the image type feature index is used to extract embedding features from the image type encephalopathy data to obtain the encephalopathy embedding vector; Matching feature points of brain disease word vectors and brain disease embedding vectors to generate brain disease feature point matching data; hierarchically fusing feature maps of brain disease word vectors and brain disease embedding vectors based on the brain disease feature point matching data to generate a multi-dimensional feature hierarchical table; A standardized matrix was constructed for text-type encephalopathy data and imaging-type encephalopathy data according to the multidimensional feature hierarchy table to generate multimodal structured hierarchical data.

4. The method for constructing a brain disease nursing knowledge base according to claim 1, characterized in that: The redundant relationship pruning process for spatial semantic modeling data includes: Perform semantic node recognition on the spatial semantic modeling data to obtain spatial semantic node recognition data; perform adjacent node relationship density calculation on the spatial semantic modeling data according to the spatial semantic node recognition data to obtain adjacent node relationship density data; perform redundant relationship node analysis on the spatial semantic node recognition data using the adjacent node relationship density data to generate redundant relationship node data; Score the spatial semantic node recognition data for relationship importance to generate relationship semantic importance score data; sort the relationship semantic importance score data by pruning priority through redundant relationship node data to generate pruning priority score data; calculate the information entropy value of redundant relationship node data by pruning priority score data to obtain the redundant relationship node information entropy value; The information entropy value of the redundant relationship node is compared with the preset standard information entropy value threshold. When the information entropy value of the redundant relationship node is less than the preset standard information entropy value threshold, the corresponding redundant relationship node data is redundantly pruned to generate spatial semantic optimization data.

5. The method for constructing a brain disease nursing knowledge base according to claim 1, characterized in that: Step S23 includes the following steps: Step S231: performing cross-node association mining on the spatial semantic optimization data to generate cross-node association data; performing semantic aggregation on the cross-node association data to generate a preliminary cross-node knowledge association network; Step S232: performing node association relationship weight distribution on the cross-node knowledge association network to generate cross-node association relationship weights; performing multi-layer relationship weight optimization on the cross-node knowledge association network according to the cross-node association relationship weights to generate a multi-layer weighted association network; Step S233: enhancing the connection of the multi-layer weighted association network to generate a cross-node brain disease association network, and using an enhanced connection algorithm based on node similarity to ensure that the network can maintain stable node associations under different circumstances; Step S234: Perform multi-node information joint reasoning analysis on the cross-node brain disease association network to generate multi-node information joint reasoning data; perform distributed reasoning node optimization on the cross-node brain disease association network based on the multi-node information joint reasoning data to generate a distributed brain disease collaborative association network.

6. The method for constructing a brain disease nursing knowledge base according to claim 5, characterized in that: The multi-node information joint reasoning analysis of the cross-node brain disease association network includes: Aggregating node data of the cross-node brain disease association network to generate cross-node brain disease association aggregated data; extracting key features of the cross-node brain disease association aggregated data to obtain cross-node brain disease association key feature data, wherein the key feature extraction includes mean calculation, median calculation and variance calculation; By using the cross-node correlation key feature data, the Pearson coefficient between nodes of the cross-node brain disease correlation network is calculated to obtain the cross-node node similarity data; the correlation analysis is performed on the cross-node node similarity data to generate a node similarity matrix; According to the node similarity matrix, the rule-based reasoning relationship of the cross-node brain disease association network is established to generate a rule-based reasoning set; based on the rule-based reasoning set, information reasoning is performed on the cross-node brain disease association network to generate a reasoning result set; the reasoning result set is logically deduced to generate multi-node information joint reasoning data.

7. The method for constructing a brain disease nursing knowledge base according to claim 1, characterized in that: Step S24 includes the following steps: Step S241: performing fuzzy reasoning on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate a encephalopathy fuzzy reasoning result data set; performing identical result data screening on the encephalopathy fuzzy reasoning result data set to obtain identical reasoning result data; Step S242: Divide the data set into a data set with the same inference result data to generate a model training set and a model test set; perform model training on the model training set using a long short-term memory neural network algorithm to generate a complex disease evolution prediction pre-model; perform model optimization iteration on the complex disease evolution prediction pre-model using the model test set to generate a complex disease evolution prediction model; Step S243: Import the same inference result data into the complex disease evolution prediction model to perform complex disease evolution prediction and generate disease evolution prediction data; dynamically map brain disease nursing needs to the disease evolution prediction data and generate dynamic nursing needs data.

8. The method for constructing a brain disease nursing knowledge base according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: Acquire real-time encephalopathy physiological monitoring data of the user; Step S32: Screening the abnormal monitoring data of the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data; Step S33: Generate a personalized encephalopathy nursing plan for the dynamic nursing demand data according to the user's real-time abnormal encephalopathy physiological monitoring data to obtain a personalized encephalopathy nursing plan; Step S34: Implement the nursing plan on the user's real-time brain disease physiological monitoring data through a personalized brain disease nursing plan, and synchronously collect the user's real-time brain disease physiological monitoring data after implementation; dynamically screen the user's real-time brain disease physiological monitoring data and the user's real-time brain disease abnormal physiological monitoring data after implementation to generate personalized brain disease nursing change data.

9. A system for constructing a knowledge base for brain disease nursing, characterized in that: The method for constructing a brain disease nursing knowledge base according to claim 1 is used to execute the method for constructing a brain disease nursing knowledge base, wherein the system for constructing the brain disease nursing knowledge base comprises: The structured stratification module is used to obtain multi-source encephalopathy data; perform structured data screening on multi-source encephalopathy data to generate structured encephalopathy data and unstructured encephalopathy data; perform multi-modal data structured stratification on unstructured encephalopathy data to generate multi-modal structured stratified data; perform data preprocessing on multi-modal structured stratified data and structured encephalopathy data to generate a standard structured encephalopathy data set; The relational network modeling module is used to perform context analysis on the standard structured encephalopathy data set to generate a semantic vector set of encephalopathy information; perform redundant relation pruning on the semantic vector set of encephalopathy information to generate spatial semantic optimization data; perform collaborative reasoning modeling on the spatial semantic optimization data to generate a distributed encephalopathy collaborative association network; perform dynamic mapping of encephalopathy nursing needs on the standard structured encephalopathy data set based on the distributed encephalopathy collaborative association network to generate dynamic nursing needs data; The dynamic change detection module is used to obtain the user's real-time encephalopathy physiological monitoring data; perform abnormal monitoring data screening on the user's real-time encephalopathy physiological monitoring data to obtain the user's real-time encephalopathy abnormal physiological monitoring data; perform nursing implementation data collection on the dynamic nursing demand data according to the user's real-time encephalopathy abnormal physiological monitoring data to obtain the user's real-time encephalopathy physiological monitoring data after implementation; perform dynamic change data screening on the user's real-time encephalopathy physiological monitoring data after implementation and the user's real-time encephalopathy abnormal physiological monitoring data to generate personalized encephalopathy nursing change data; The database construction module is used to conduct brain disease nursing feedback evaluation on personalized brain disease nursing change data to generate brain disease nursing feedback evaluation data; and to construct a database for the brain disease nursing feedback evaluation data to generate a brain disease nursing knowledge base.

Images