Elderly comprehensive health assessment data processing method and system based on digital twinborn

Abstract

The invention discloses a method and a system for processing comprehensive health assessment data of old people based on digital twinning, and relates to the technical field of data processing, the method comprises the following steps: fusing multi-source health data of a target old people group, establishing a health knowledge graph, and storing the health knowledge graph; constructing a digital twinborn health model based on the health knowledge graph, performing a simulation test in the constructed digital twinborn health model, and identifying health risks existing in the health data; calculating the health feature similarity between the target elderly group and the reference group by using a similarity algorithm; and analyzing health problem risks existing in the target elderly group by using the health feature similarity, and making a comprehensive health assessment report based on an analysis result. According to the method, the health knowledge graph is established by fusing the multi-source health data, and the digital twinborn health model is constructed, so that the quality and comparability of the data are improved, and a solid foundation is provided for subsequent simulation tests and risk identification.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for processing data for comprehensive health assessment of the elderly based on digital twins. Background Art

[0002] With the acceleration of the aging process of the global population, the health issues of the elderly have received more and more attention from all walks of life. As a key tool for preventing and treating elderly diseases, comprehensive health assessment of the elderly plays an extremely important role in improving the quality of life of the elderly and prolonging their lifespan. In recent years, with the continuous development of digital twin technology, its application in the field of medical health has become increasingly extensive. Digital twin technology provides new ideas and methods for elderly health assessment by building virtual models to achieve real-time monitoring, analysis and prediction of physical objects. Through digital twin technology, a virtual health model of the elderly can be built to monitor multi-source information such as their physical condition, mental state and participation in social activities in real time, and make more accurate and comprehensive health evaluations based on this. This method can not only improve the accuracy and efficiency of the assessment, but also provide more personalized health management services for the elderly.

[0003] Although digital twin technology has shown great potential in comprehensive health assessment of the elderly, it still faces some challenges in its actual application. One of the most significant problems is the diversity of data sources. The health information of the elderly is usually scattered across multiple different channels, including but not limited to hospital medical records, family doctor records, data collected by personal wearable devices, and self-reported living habits. These data are not only in different formats (such as structured and unstructured text, images or videos), but also often lack unified standards to ensure their quality and consistency, which greatly increases the complexity of data integration and hinders a comprehensive insight into personal health status. This not only reduces the accuracy of health predictions and risk alerts, but also limits the ability to formulate personalized interventions based on data analysis.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In response to the problems in the related technology, the present invention proposes a data processing method and system for comprehensive health assessment of the elderly based on digital twins to overcome the above-mentioned technical problems existing in the existing related technology.

[0006] To this end, the specific technical solution adopted by the present invention is as follows:

[0007] According to one aspect of the present invention, a method for processing data of comprehensive health assessment for the elderly based on digital twins is provided, and the method comprises the following steps:

[0008] S1. Integrate multi-source health data of the target elderly group, establish a health knowledge graph and store the health knowledge graph;

[0009] S2. Build a digital twin health model based on the health knowledge graph, and conduct simulation experiments in the constructed digital twin health model to identify health risks in health data;

[0010] S3. Collect multi-source health data of the reference elderly group, establish a reference elderly health feature database, and use a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group;

[0011] S4. Use the similarity of health characteristics to analyze the risks of health problems in the target elderly group and develop a comprehensive health assessment report based on the analysis results.

[0012] Furthermore, integrating the multi-source health data of the target elderly group, establishing a health knowledge graph and storing the health knowledge graph includes the following steps:

[0013] S11. Collect multi-source health data of the target elderly group and pre-process the multi-source health data;

[0014] S12. Use multi-source data fusion technology to fuse the pre-processed multi-source health data, and build a health knowledge graph based on the fused health data;

[0015] S13. Import the constructed health knowledge graph into the graph database and optimize the index of the graph database.

[0016] Furthermore, collecting multi-source health data of the target elderly group and preprocessing the multi-source health data include the following steps:

[0017] S111. Collect multi-source health data of the target elderly group and perform data cleaning on the multi-source health data, where the multi-source health data includes text data, image data, and video data;

[0018] S112. Analyze the cleaned text data using natural language processing technology, extract text information from the text data, and convert the text information into a structured data format;

[0019] S113, using image recognition technology and optical character recognition technology to process the cleaned image data, extract image information from the image data, and convert the image information into a structured data format;

[0020] S114. Combining video analysis technology, speech recognition technology and image recognition technology, the cleaned video data is analyzed to identify video information in the video data, and the video information is converted into a structured data format.

[0021] Furthermore, the pre-processed multi-source health data is fused using multi-source data fusion technology, and a health knowledge graph is constructed based on the fused health data, including the following steps:

[0022] S121. Use multi-source data fusion technology to fuse the pre-processed multi-source health data and establish the correlation between the health data;

[0023] S122. Combining the correlation between health data and time series information, mining health information of the target elderly group through the rule engine;

[0024] S123. Analyze the mined health information features, and import the analyzed concepts, relationships, and attributes into the model layer to construct a health knowledge graph.

[0025] Furthermore, the mined health information features are analyzed, and the analyzed concepts, relationships, and attributes are imported into the model layer. The construction of the health knowledge graph includes the following steps:

[0026] S1231, analyzing data source characteristics, content characteristics, and form characteristics of health information, and defining conceptual information of the model layer based on the analysis results;

[0027] S1232. Based on the conceptual information of the model layer, define entity types and entity attributes, and determine the relationship information between entities;

[0028] S1233, importing the defined pattern layer information into the knowledge graph construction tool, and performing entity recognition, relationship extraction and attribute filling processing;

[0029] S1234. Combine the results of entity recognition, relationship extraction and attribute filling to construct a health knowledge graph.

[0030] Furthermore, a digital twin health model is constructed based on the health knowledge graph, and simulation experiments are conducted in the constructed digital twin health model to identify health risks in health data, including the following steps:

[0031] S21. Based on the health knowledge graph, extract health indicators and health parameters and establish a health data set;

[0032] S22, import the health data set into the pre-set modeling tool and build the initial digital twin health model in combination with the machine learning algorithm;

[0033] S23, applying a grid simplification algorithm to optimize the initial digital twin health model to obtain a digital twin health model;

[0034] S24. Conduct simulation experiments on the digital twin health model to simulate the health change process of the target elderly group;

[0035] S25. Analyze the results of the simulation test and identify the health risks existing in the health data.

[0036] Furthermore, the mesh simplification algorithm is applied to optimize the initial digital twin health model, and the digital twin health model is obtained, which includes the following steps:

[0037] S231, extracting geometric data from the initial digital twin health model, establishing a topological relationship between vertices, and calculating an error metric value of each vertex, wherein the geometric data includes information of vertices, edges, and faces;

[0038] S232, sorting the vertices according to their error metric values, and selecting the vertex with the smallest error metric value for edge folding processing;

[0039] S233, evaluating whether the processed digital twin health model meets the preset conditions. If not, repeat step S232; otherwise, proceed to the next step;

[0040] S234, calculating the optimization operator of each vertex, and updating the error metric value of the vertex;

[0041] S235. Compare the updated error metric value with the preset threshold. If the error metric value exceeds the preset threshold, repeat step S234; otherwise, continue to update the vertex position until the preset number of iterations is reached to obtain the digital twin health model.

[0042] Furthermore, the calculation formula of the error metric is:

[0043]

[0044] In the formula, E(v i ) represents the vertex v i The error measure of

[0045] f represents face;

[0046] F i Represents the vertex v i The set of all adjacent faces;

[0047] v i represents the i-th vertex;

[0048] p f The normal vector representing the plane equation of the face f;

[0049] T represents matrix transpose;

[0050] Qf represents the quadratic error matrix of surface f.

[0051] Furthermore, collecting multi-source health data of the reference elderly group, establishing a reference elderly health feature database, and using a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group include the following steps:

[0052] S31, collecting reference multi-source health data of the reference elderly group, and preprocessing the reference multi-source health data;

[0053] S32, extracting reference health information features from the preprocessed reference multi-source health data, and establishing an elderly health feature library based on the reference health information features;

[0054] S33, converting the health information characteristics of the target elderly group into a feature vector that matches a reference health information feature library;

[0055] S34. Use a similarity algorithm to calculate the similarity of health characteristics between the target elderly group and the reference elderly group based on the converted feature vectors.

[0056] According to another aspect of the present invention, there is also provided a digital twin-based elderly comprehensive health assessment data processing system, which includes a graph construction and storage module, a model construction and risk identification module, a feature library establishment and similarity calculation module, and a risk analysis and report formulation module;

[0057] The graph construction and storage module is used to integrate the multi-source health data of the target elderly group, build a health knowledge graph and store the health knowledge graph;

[0058] The model building and risk identification module is used to build a digital twin health model based on the health knowledge graph, and conduct simulation experiments in the constructed digital twin health model to identify health risks in health data;

[0059] The feature library establishment and similarity calculation module is used to collect multi-source health data of the reference elderly group, establish a reference elderly health feature library, and use a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group;

[0060] The risk analysis and report preparation module is used to analyze the risks of health problems in the target elderly group by using the similarity of health characteristics, and to prepare a comprehensive health assessment report based on the analysis results.

[0061] The beneficial effects of the present invention are:

[0062] 1. The present invention establishes a health knowledge graph and constructs a digital twin health model by integrating multi-source health data, which not only improves the quality and comparability of the data, but also provides a solid foundation for subsequent simulation experiments and risk identification.

[0063] 2. Through simulation experiments, the present invention can simulate the changes in health status under different health intervention measures, so as to identify potential health risks in advance; at the same time, through health feature comparison and similarity calculation, risks that may exist in the target elderly group and are similar to known health problems in the reference group are discovered, providing the possibility for timely warning and intervention.

[0064] 3. The present invention conducts an in-depth analysis of the health risks of the target elderly group through the similarity of health characteristics and simulation test results, and develops a comprehensive health assessment report, which provides comprehensive health information guidance for health managers, medical institutions and individual patients, helps to optimize the health management process and improve the pertinence and effectiveness of health management. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0066] Figure 1 is a flow chart of a method for processing data of comprehensive health assessment for the elderly based on digital twins according to an embodiment of the present invention;

[0067] Figure 2 It is a principle block diagram of a digital twin-based elderly comprehensive health assessment data processing system according to an embodiment of the present invention.

[0068] In the figure:

[0069] 1. Graph construction and storage module; 2. Model construction and risk identification module; 3. Feature library establishment and similarity calculation module; 4. Risk analysis and report preparation module. DETAILED DESCRIPTION

[0070] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0071] According to an embodiment of the present invention, a method and system for processing comprehensive health assessment data for the elderly based on digital twins are provided.

[0072] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the digital twin-based data processing method for comprehensive health assessment of the elderly according to an embodiment of the present invention, the method includes the following steps:

[0073] S1. Integrate the multi-source health data of the target elderly group, establish a health knowledge graph and store the health knowledge graph.

[0074] Specifically, integrating multi-source health data of the target elderly group, establishing a health knowledge graph and storing the health knowledge graph includes the following steps:

[0075] S11. Collect multi-source health data of the target elderly group and pre-process the multi-source health data.

[0076] Specifically, collecting multi-source health data of the target elderly group and preprocessing the multi-source health data include the following steps:

[0077] S111. Collect multi-source health data of the target elderly group and perform data cleaning on the multi-source health data, where the multi-source health data includes text data, image data, and video data.

[0078] S112. Use natural language processing technology to parse the cleaned text data, extract text information from the text data, and convert the text information into a structured data format.

[0079] It should be noted that, according to the experimental requirements, choose appropriate natural language processing tools (such as NLTK, SpaCy), perform word segmentation and part-of-speech tagging on the text, identify and extract named entities; according to the experimental objectives, write rules or use machine learning models to extract key information, and store this information in a structured format (such as tables, databases).

[0080] S113. Using image recognition technology and optical character recognition technology, the cleaned image data is processed to extract image information from the image data and convert the image information into a structured data format.

[0081] It should be noted that image recognition algorithms (such as convolutional neural networks (CNNs)) are used to extract feature information from images, such as edges, textures, shapes, etc.; identify target objects in images and determine their positions and sizes. OCR algorithms are used to identify text information in images and convert it into editable text formats. Key information, such as time, location, person, value, etc., is extracted from the identified image information and text information, and the extracted key information is organized in a predetermined structured data format, such as a table, database, etc.

[0082] For example, suppose an image of a health checkup report for an elderly person contains the following information:

[0083] Report title: Elderly Health Examination Report;

[0084] Report date: XX / XX / XXXX;

[0085] Patient information: name, age, gender;

[0086] Examination items: blood pressure, blood sugar, blood lipids, etc.;

[0087] Examination results: blood pressure "140 / 90mmHg", blood sugar "6.5mmol / L", blood lipids "high", etc.;

[0088] Based on the above information, the specific implementation is as follows:

[0089] Use image recognition algorithms to identify key areas in the image, such as title, date, patient information, examination items and examination results, etc. Perform OCR recognition on the identified text areas to extract text information, such as "name, age, gender", "140 / 90mmHg", etc. Organize the extracted key information in a table-like structured data format.

[0090] S114. Combining video analysis technology, speech recognition technology and image recognition technology, the cleaned video data is analyzed to identify video information in the video data, and the video information is converted into a structured data format.

[0091] It should be noted that video analysis technology includes using deep learning algorithms (such as YOLO, Faster R-CNN, etc.) to detect target objects in the video, such as people and objects; using video behavior recognition algorithms (such as 3D convolutional neural networks, recurrent neural networks, etc.) to analyze the behavior patterns of people in the video, such as walking, running, sitting, etc.; using scene understanding algorithms to analyze scene information in the video, such as indoors, outdoors, and public places.

[0092] Speech recognition technology includes extracting audio information from video data as input for speech recognition; using speech recognition algorithms (such as Google speech recognition, Baidu speech recognition, etc.) to convert audio information into text information; and preprocessing the converted text information, such as removing redundant information, punctuation marks, etc., to facilitate subsequent analysis.

[0093] Image recognition technology includes extracting key frames from videos regularly or on demand as input for image recognition; using image recognition algorithms (such as convolutional neural networks) to identify objects in key frames and mark their locations and categories; and further extracting attribute information of objects, such as color, shape, size, etc.

[0094] The information obtained from video analysis, speech recognition and image recognition is integrated to form a complete video information description; according to a predetermined structured data format (such as JSON, XML, etc.), the integrated information is converted into a structured data format.

[0095] S12. Use multi-source data fusion technology to fuse the pre-processed multi-source health data, and build a health knowledge graph based on the fused health data.

[0096] Specifically, using multi-source data fusion technology to fuse the pre-processed multi-source health data, and constructing a health knowledge graph based on the fused health data includes the following steps:

[0097] S121. Use multi-source data fusion technology to fuse the pre-processed multi-source health data and establish the correlation between the health data.

[0098] S122. Combining the correlation between health data and time series information, the health information of the target elderly group is mined through the rule engine.

[0099] S123. Analyze the mined health information features, and import the analyzed concepts, relationships, and attributes into the model layer to construct a health knowledge graph.

[0100] Specifically, the mined health information features are analyzed, and the analyzed concepts, relationships, and attributes are imported into the model layer. The construction of the health knowledge graph includes the following steps:

[0101] S1231, analyzing data source characteristics, content characteristics, and form characteristics of health information, and defining conceptual information of the model layer based on the analysis results;

[0102] S1232. Based on the conceptual information of the model layer, define entity types and entity attributes, and determine the relationship information between entities;

[0103] S1233, importing the defined pattern layer information into the knowledge graph construction tool, and performing entity recognition, relationship extraction and attribute filling processing;

[0104] S1234. Combine the results of entity recognition, relationship extraction and attribute filling to construct a health knowledge graph.

[0105] S13. Import the constructed health knowledge graph into the graph database and optimize the index of the graph database.

[0106] It should be noted that graph databases include Neo4j, JanusGraph, etc.

[0107] S2. Build a digital twin health model based on the health knowledge graph, and conduct simulation experiments in the constructed digital twin health model to identify health risks in health data.

[0108] Specifically, a digital twin health model is constructed based on the health knowledge graph, and simulation experiments are conducted in the constructed digital twin health model to identify health risks in health data, including the following steps:

[0109] S21. Based on the health knowledge graph, extract health indicators and health parameters and establish a health data set.

[0110] It should be noted that health indicators include body mass index, blood pressure, blood sugar, blood lipids, lung function, liver function, kidney function, bone health and other examination indicators.

[0111] Health parameters include weight parameters, blood pressure parameters, blood sugar parameters, blood lipid parameters, etc.

[0112] S22. Import the health data set into the pre-set modeling tool and build the initial digital twin health model in combination with the machine learning algorithm.

[0113] S23. Apply a grid simplification algorithm to optimize the initial digital twin health model to obtain a digital twin health model.

[0114] Specifically, the mesh simplification algorithm is applied to optimize the initial digital twin health model, and obtaining the digital twin health model includes the following steps:

[0115] S231. Extract geometric data from the initial digital twin health model, establish topological relationships between vertices, and calculate the error metric value of each vertex. The geometric data includes information on vertices, edges, and faces.

[0116] Specifically, the calculation formula of the error metric value is:

[0117]

[0118] In the formula, E(v i ) represents the vertex v i The error measure of

[0119] f represents face;

[0120] F i Represents the vertex v i The set of all adjacent faces;

[0121] v i represents the i-th vertex;

[0122] p f The normal vector representing the plane equation of the face f;

[0123] T represents matrix transpose;

[0124] Q f represents the quadratic error matrix of face f.

[0125] S232. Sort the vertices according to their error metric values, and select the vertex with the smallest error metric value for edge folding processing.

[0126] S233. Evaluate whether the processed digital twin health model meets the preset conditions. If not, repeat step S232; otherwise, proceed to the next step.

[0127] S234, calculating the optimization operator of each vertex and updating the error metric value of the vertex.

[0128] S235. Compare the updated error metric value with the preset threshold. If the error metric value exceeds the preset threshold, repeat step S234; otherwise, continue to update the vertex position until the preset number of iterations is reached to obtain the digital twin health model.

[0129] S24. Conduct simulation experiments on the digital twin health model to simulate the health change process of the target elderly group.

[0130] It should be noted that the simulation includes the health changes of the target elderly group over a period of time in the future, including fluctuations in physiological indicators, the occurrence and development of diseases, etc.

[0131] Set different life scenarios and intervention measures (such as eating habits, exercise methods, drug treatments, etc.) to observe their impact on health status; set a reasonable trial period (such as 1 year, 5 years or 10 years) according to research needs; import the health data set into the digital twin health model, and initialize the parameters to reflect the initial health status of the target elderly group; adjust the relevant parameters in the model (such as dietary intake, exercise volume, drug dosage, etc.) according to the set life scenarios and intervention measures; run the model to simulate the health changes of the target elderly group in the future, and automatically calculate key information such as fluctuations in physiological indicators and the probability of disease occurrence; record the health status data output by the model in real time, including the trend of changes in physiological indicators, the time of disease occurrence and development.

[0132] S25. Analyze the results of the simulation test and identify the health risks existing in the health data.

[0133] It should be noted that, according to the information in the health knowledge map, indicators for assessing health risks are determined, such as abnormal blood pressure rate, blood sugar fluctuation range, and disease occurrence probability. For each elderly individual, the scores of various risk assessment indicators are calculated based on the result data of the simulation test. The scoring criteria can be set according to the actual situation. For example, the higher the abnormal blood pressure rate, the higher the score; the larger the blood sugar fluctuation range, the higher the score. The scores of various risk assessment indicators are combined to obtain the total health risk score of each elderly individual; according to the total score, the elderly group is divided into different levels such as high risk, medium risk and low risk. For elderly individuals with high and medium risks, the result data of their simulation tests are further analyzed to identify the main factors leading to health risks, such as hypertension, hyperglycemia, obesity, etc.; according to different risk factors, the elderly group is classified and managed, and similar health management measures and intervention strategies can be formulated for elderly individuals with the same risk factors.

[0134] S3. Collect multi-source health data of the reference elderly group, establish a reference elderly health feature database, and use a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group.

[0135] Specifically, collecting multi-source health data of the reference elderly group, establishing a reference elderly health feature database, and using a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group include the following steps:

[0136] S31, collecting reference multi-source health data of the reference elderly group, and preprocessing the reference multi-source health data;

[0137] S32, extracting reference health information features from the preprocessed reference multi-source health data, and establishing an elderly health feature library based on the reference health information features;

[0138] S33, converting the health information characteristics of the target elderly group into a feature vector that matches a reference health information feature library;

[0139] S34. Use a similarity algorithm to calculate the similarity of health characteristics between the target elderly group and the reference elderly group based on the converted feature vectors.

[0140] It should be noted that the similarity algorithm includes cosine similarity, Euclidean distance or Mahalanobis distance and other algorithms.

[0141] S4. Use the similarity of health characteristics to analyze the risks of health problems in the target elderly group and develop a comprehensive health assessment report based on the analysis results.

[0142] It should be noted that, based on the similarity calculation results, the reference group most similar to the target group is selected for health feature comparison. Assuming that the similarity between reference group B and target group A is the highest (0.85), analyze the risks of major health problems in reference group B, such as the prevalence of hypertension, hyperglycemia, and cardiovascular disease. Based on the similarity between the target group and the reference group, infer the possible health risks of the target group. For example, if 60% of the members of reference group B are at risk of hyperglycemia, it is inferred that target group A also has a higher risk of hyperglycemia. The health risks are then converted into specific risk probabilities, for example: hyperglycemia risk probability: 65%; cardiovascular disease risk probability: 50%.

[0143] Summarize the target group’s health characteristic data, similarities, and inferred health risk probabilities to form a complete health assessment report.

[0144] like Figure 2 As shown, according to another embodiment of the present invention, a digital twin-based elderly comprehensive health assessment data processing system is also provided, the system comprising a graph construction and storage module 1, a model construction and risk identification module 2, a feature library establishment and similarity calculation module 3 and a risk analysis and report formulation module 4;

[0145] Graph construction and storage module 1, used to integrate multi-source health data of the target elderly group, build a health knowledge graph and store the health knowledge graph;

[0146] Model building and risk identification module 2, used to build a digital twin health model based on the health knowledge graph, and conduct simulation experiments in the constructed digital twin health model to identify health risks in health data;

[0147] The feature library establishment and similarity calculation module 3 is used to collect multi-source health data of the reference elderly group, establish a reference elderly health feature library, and use a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group;

[0148] The risk analysis and report preparation module 4 is used to analyze the health problem risks of the target elderly group by using the similarity of health characteristics, and to prepare a comprehensive health assessment report based on the analysis results.

[0149] In summary, with the help of the above technical solution of the present invention, by integrating multi-source health data, a health knowledge graph is established, and a digital twin health model is constructed, which not only improves the quality and comparability of the data, but also provides a solid foundation for subsequent simulation experiments and risk identification. Through simulation experiments, it is possible to simulate the changes in health status under different health intervention measures, so as to identify potential health risks in advance; at the same time, through health feature comparison and similarity calculation, risks that may exist in the target elderly group and are similar to known health problems in the reference group are found, which provides the possibility for timely warning and intervention. Through the similarity of health features and simulation test results, the risks of health problems in the target elderly group are deeply analyzed, and a comprehensive health assessment report is formulated, which provides comprehensive health information guidance for health managers, medical institutions and individual patients, and helps to optimize the health management process and improve the pertinence and effectiveness of health management.

[0150] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A data processing method for comprehensive health assessment of the elderly based on digital twins, characterized in that: The method comprises the following steps: S1. Integrate multi-source health data of the target elderly group, establish a health knowledge graph and store the health knowledge graph; S2. Build a digital twin health model based on the health knowledge graph, and conduct simulation experiments in the constructed digital twin health model to identify health risks in health data; S3. Collect multi-source health data of the reference elderly group, establish a reference elderly health feature database, and use a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group; S4. Use the similarity of health characteristics to analyze the risks of health problems in the target elderly group and develop a comprehensive health assessment report based on the analysis results.

2. According to claim 1, a method for processing data of comprehensive health assessment for the elderly based on digital twins is characterized in that: The method of integrating multi-source health data of the target elderly group, establishing a health knowledge graph and storing the health knowledge graph includes the following steps: S11. Collect multi-source health data of the target elderly group and pre-process the multi-source health data; S12. Use multi-source data fusion technology to fuse the pre-processed multi-source health data, and build a health knowledge graph based on the fused health data; S13. Import the constructed health knowledge graph into the graph database and optimize the index of the graph database.

3. According to claim 2, a method for processing data of comprehensive health assessment for the elderly based on digital twins is characterized in that: The collecting of multi-source health data of the target elderly population and preprocessing of the multi-source health data comprises the following steps: S111, collecting multi-source health data of the target elderly group and performing data cleaning on the multi-source health data, wherein the multi-source health data includes text data, image data, and video data; S112. Analyze the cleaned text data using natural language processing technology, extract text information from the text data, and convert the text information into a structured data format; S113, using image recognition technology and optical character recognition technology to process the cleaned image data, extract image information from the image data, and convert the image information into a structured data format; S114. Combining video analysis technology, speech recognition technology and image recognition technology, the cleaned video data is analyzed to identify video information in the video data, and the video information is converted into a structured data format.

4. According to claim 2, a method for processing data of comprehensive health assessment for the elderly based on digital twins is characterized in that: The method of fusing the preprocessed multi-source health data using multi-source data fusion technology and constructing a health knowledge graph based on the fused health data includes the following steps: S121. Use multi-source data fusion technology to fuse the pre-processed multi-source health data and establish the correlation between the health data; S122. Combining the correlation between health data and time series information, mining health information of the target elderly group through the rule engine; S123. Analyze the mined health information features, and import the analyzed concepts, relationships, and attributes into the model layer to construct a health knowledge graph.

5. According to claim 4, a method for processing data of comprehensive health assessment for the elderly based on digital twins is characterized in that: The analysis of the mined health information features and importing the analyzed concepts, relationships and attributes into the model layer to construct a health knowledge graph includes the following steps: S1231, analyzing data source characteristics, content characteristics, and form characteristics of health information, and defining conceptual information of the model layer based on the analysis results; S1232. Based on the conceptual information of the model layer, define entity types and entity attributes, and determine the relationship information between entities; S1233, importing the defined pattern layer information into the knowledge graph construction tool, and performing entity recognition, relationship extraction and attribute filling processing; S1234. Combine the results of entity recognition, relationship extraction and attribute filling to construct a health knowledge graph.

6. According to claim 1, a method for processing data of comprehensive health assessment for the elderly based on digital twins is characterized in that: The digital twin health model is constructed based on the health knowledge graph, and a simulation experiment is performed in the constructed digital twin health model to identify health risks existing in health data, including the following steps: S21. Based on the health knowledge graph, extract health indicators and health parameters and establish a health data set; S22, import the health data set into the pre-set modeling tool and build the initial digital twin health model in combination with the machine learning algorithm; S23, applying a grid simplification algorithm to optimize the initial digital twin health model to obtain a digital twin health model; S24. Conduct simulation experiments on the digital twin health model to simulate the health change process of the target elderly group; S25. Analyze the results of the simulation test and identify the health risks existing in the health data.

7. The method for processing data of comprehensive health assessment for the elderly based on digital twins according to claim 6 is characterized in that: The method of applying a grid simplification algorithm to optimize the initial digital twin health model to obtain the digital twin health model includes the following steps: S231, extracting geometric data from the initial digital twin health model, establishing a topological relationship between vertices, and calculating an error metric value of each vertex, wherein the geometric data includes information of vertices, edges, and faces; S232, sorting the vertices according to their error metric values, and selecting the vertex with the smallest error metric value for edge folding processing; S233, evaluating whether the processed digital twin health model meets the preset conditions. If not, repeat step S232; otherwise, proceed to the next step; S234, calculating the optimization operator of each vertex, and updating the error metric value of the vertex; S235. Compare the updated error metric value with the preset threshold. If the error metric value exceeds the preset threshold, repeat step S234; otherwise, continue to update the vertex position until the preset number of iterations is reached to obtain the digital twin health model.

8. The method for processing data of comprehensive health assessment for the elderly based on digital twins according to claim 7 is characterized in that: The calculation formula of the error metric value is: In the formula, E(v i ) represents the vertex v i The error measure of f represents face; F i Represents the vertex v i The set of all adjacent faces; v i represents the i-th vertex; p f The normal vector representing the plane equation of the face f; T represents matrix transpose; Q f represents the quadratic error matrix of surface f.

9. The method for processing data of comprehensive health assessment for the elderly based on digital twins according to claim 1 is characterized in that: The collecting of multi-source health data of the reference elderly group, establishing a reference elderly health feature database, and using a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group includes the following steps: S31, collecting reference multi-source health data of the reference elderly group, and preprocessing the reference multi-source health data; S32, extracting reference health information features from the preprocessed reference multi-source health data, and establishing an elderly health feature library based on the reference health information features; S33, converting the health information characteristics of the target elderly group into a feature vector that matches a reference health information feature library; S34. Use a similarity algorithm to calculate the similarity of health characteristics between the target elderly group and the reference elderly group based on the converted feature vectors.

10. A digital twin-based elderly comprehensive health assessment data processing system, used to implement the digital twin-based elderly comprehensive health assessment data processing method according to any one of claims 1 to 9, characterized in that: The system includes a graph construction and storage module, a model construction and risk identification module, a feature library establishment and similarity calculation module, and a risk analysis and report formulation module; The graph construction and storage module is used to integrate multi-source health data of the target elderly group, establish a health knowledge graph and store the health knowledge graph; The model building and risk identification module is used to build a digital twin health model based on the health knowledge graph, and to conduct simulation experiments in the constructed digital twin health model to identify health risks existing in health data; The feature library establishment and similarity calculation module is used to collect multi-source health data of the reference elderly group, establish a reference elderly health feature library, and use a similarity algorithm to calculate the similarity of health features between the target elderly group and the reference group; The risk analysis and report preparation module is used to analyze the risks of health problems in the target elderly group by using the similarity of health characteristics, and to prepare a comprehensive health assessment report based on the analysis results.

Images