An artificial intelligence-based chronic disease patient health management method and system

Abstract

The application provides a chronic disease patient health management method and system based on artificial intelligence, and comprises the following steps: acquiring health monitoring dimensions, periodically monitoring a target patient according to the health monitoring dimensions, and obtaining physiological data of each health monitoring dimension in each cycle node; performing correlation fusion analysis on the physiological data of each health monitoring dimension in the cycle node to obtain the health status of the target patient in the cycle node; and generating a dynamic health management strategy according to the health status of each cycle node. The target patient is periodically monitored, the health status of the target patient in the cycle node is accurately obtained, and then a dynamic health management strategy is generated according to the health status of each cycle node, which is beneficial to the all-round health monitoring of the target patient, effectively guarantees the accuracy of the obtained health status through correlation fusion analysis, is beneficial to providing reliable data support for accurately generating a dynamic health management strategy, and guarantees the effectiveness, accuracy and intelligence of the health management of the chronic disease patient.

Description

Technical Field

[0001] The present invention relates to the field of health management technology, and in particular to an artificial intelligence-based health management method and system for chronic disease patients. Background Art

[0002] Currently, with the development of modern society, chronic diseases have become a major challenge in the field of global public health. Chronic diseases such as cardiovascular disease, diabetes, and hypertension are usually characterized by long courses, complex causes, and difficulty in cure. They require long-term health management to control the progression of the disease and improve the quality of life of patients.

[0003] However, in traditional chronic disease management, the inability to accurately and continuously collect patient examination data leads to data dispersion, making data analysis inaccurate and chronic disease monitoring of target patients incomplete. Furthermore, it is also impossible to detect abnormalities in chronic diseases in a timely and effective manner, and thus it is impossible to effectively adjust chronic disease health management strategies in real time, resulting in poor chronic disease health management strategies for target patients.

[0004] Therefore, in order to overcome the above technical problems, the present invention provides a health management method and system for chronic disease patients based on artificial intelligence. Summary of the Invention

[0005] The present invention provides a health management method and system for chronic disease patients based on artificial intelligence, which is used to obtain health monitoring dimensions, thereby effectively performing periodic monitoring of target patients according to the health monitoring dimensions, obtaining physiological data of each health monitoring dimension in each cycle node, and accurately obtaining the health status of the target patient at the cycle node by performing correlation fusion analysis on the physiological data of each health monitoring dimension in the cycle node, and then generating a dynamic health management strategy based on the health status of each cycle node, which is conducive to all-round health monitoring of the target patient, and effectively ensuring the accuracy of the obtained health status through correlation fusion analysis, and thus facilitating the accurate generation of dynamic health management strategies to provide reliable data support, thereby ensuring the effectiveness, accuracy and intelligence of health management for chronic disease patients.

[0006] The present invention provides a health management method for chronic disease patients based on artificial intelligence, comprising:

[0007] Step 1: Obtain health monitoring dimensions, and perform periodic monitoring on the target patient according to the health monitoring dimensions to obtain physiological data of each health monitoring dimension at each cycle node;

[0008] Step 2: Perform correlation fusion analysis on the physiological data of each health monitoring dimension within the cycle node to obtain the health status of the target patient at the cycle node;

[0009] Step 3: Generate a dynamic health management strategy based on the health status of each periodic node.

[0010] Preferably, a method for health management of chronic disease patients based on artificial intelligence, in step 1, obtaining health monitoring dimensions, includes:

[0011] Collect the medical records of the target patients and identify the medical records to determine the case information of the target patients;

[0012] Input the case information into the preset recognition model for recognition to obtain the target keywords of the case information;

[0013] Determine the main health monitoring dimensions based on target keywords;

[0014] Input the target keyword into the preset health management map for matching, and determine the association path with the target keyword in the preset health management map;

[0015] Determine the relevant nodes of the target keyword based on the association path between the target keyword and the target keyword;

[0016] Read the node information of the relevant nodes and determine the secondary health monitoring dimension based on the node information;

[0017] The acquisition of health monitoring dimensions is completed based on the primary monitoring dimensions and secondary monitoring dimensions.

[0018] Preferably, an artificial intelligence-based health management method for chronic disease patients identifies medical records to determine case information of target patients, including:

[0019] Collecting a first image of the medical record, performing text scanning on the first image, and splitting the first image according to the scanning result to obtain a second image;

[0020] Obtain the information type for identifying the case information and obtain the identification point corresponding to the information type;

[0021] Positioning the marker point in the second image;

[0022] At the same time, a first annotation is performed in the second image according to the positioning result;

[0023] Obtaining position logic of the text in the second image, and determining text information associated with the first annotation result based on the position logic;

[0024] Performing a second annotation on the second image using text information associated with the first annotation result;

[0025] The first annotation result and the second annotation result are extracted from the second image to obtain case information of the target patient.

[0026] Preferably, a health management method for chronic disease patients based on artificial intelligence, in step 1, periodic monitoring of the target patient is performed according to the health monitoring dimension, and physiological data of each health monitoring dimension in each periodic monitoring node is obtained, including:

[0027] Preset monitoring period and set periodic monitoring nodes according to the preset monitoring period;

[0028] Read the health monitoring dimensions and start the monitoring equipment according to the health monitoring dimensions;

[0029] Collect sub-physiological data corresponding to the health monitoring dimension based on the monitoring device, obtain the dimension identifier of the health monitoring dimension, and construct a data storage node according to the dimension identifier of the health monitoring dimension;

[0030] The sub-physiological data is stored in the corresponding data storage node according to the dimension identifier, and the sub-physiological data in the data storage node is divided according to the periodic monitoring node, and the sub-physiological data segments of several periodic monitoring nodes are determined in the data storage node;

[0031] The sub-physiological data segments of several periodic monitoring nodes are the physiological data of each health monitoring dimension in each periodic monitoring node.

[0032] Preferably, a health management method for chronic disease patients based on artificial intelligence, in step 3, generates a dynamic health management strategy according to the health status of each cycle node, including:

[0033] Sort the health status of each periodic node in chronological order to obtain a health status sequence;

[0034] Read the health status of each periodic node, obtain the benchmark health interval, and match the health status sequence with the benchmark health interval respectively;

[0035] When there is a first target health state in the health state sequence that belongs to the benchmark health interval, adding a first label to the first target health state;

[0036] When there is a second target health state in the health state sequence that does not belong to the benchmark health interval, a second label is added to the second target health state;

[0037] Retrieving a first target health management policy based on a first tag from a preset policy management library;

[0038] At the same time, the physiological data corresponding to each health monitoring dimension of the target patient under the second tag is read, and the abnormal health monitoring dimension of the target patient under the second tag is determined according to the standard data interval of the physiological data corresponding to each health monitoring dimension;

[0039] Call the second target health management strategy corresponding to the abnormal health monitoring dimension in the preset policy management library;

[0040] Based on the first target health management strategy and the second target health management strategy, the health state sequence is dynamically adjusted to generate a dynamic health management strategy for each cycle node.

[0041] Preferably, in step 3 of the chronic disease patient health management method based on artificial intelligence, after generating the dynamic health management strategy according to the health state of each cycle node, the method comprises:

[0042] Construct a first visual window and a second visual window;

[0043] The health state of each cycle node is displayed in the first visual window, and the dynamic health management strategy is displayed in the second visual window.

[0044] Preferably, in step 2 of the chronic disease patient health management method based on artificial intelligence, the physiological data of each health monitoring dimension in the cycle node is associated and fused to obtain the health state of the target patient in the cycle node, comprising:

[0045] Obtain the physiological data of each health monitoring dimension in each cycle node, and determine the normalization standard of the physiological data based on the physiological properties between each health monitoring dimension;

[0046] Based on the normalization standard, the physiological data is uniformly formatted, and the physiological significance of each health monitoring dimension is extracted, and the feature extraction limited index is generated based on the physiological significance;

[0047] Based on the feature extraction limited index, the physiological data of the corresponding health monitoring dimension is scanned to obtain the corresponding feature index and the corresponding feature index quantization interval of each health monitoring dimension, and the feature index and the feature index quantization interval are associated to obtain the single-dimensional feature of each health monitoring dimension;

[0048] Access the server to call the health management system, and determine the mutual relationship between different health monitoring dimensions based on the health management system;

[0049] Based on the mutual relationship, the single-dimensional features of different health monitoring dimensions are analyzed, and the cross-dimensional features are constructed based on the analysis results;

[0050] Based on the single-dimensional features and the cross-dimensional features, the physiological data of different health monitoring dimensions are sequentially changed by a single amount, and the physiological change amount of the remaining health monitoring dimensions is monitored in real time based on the single amount change result;

[0051] Determine the strongly correlated health monitoring dimensions based on the variable threshold values ​​of physiological changes, and determine the data fusion nodes between the strongly correlated health monitoring dimensions based on the medical knowledge system;

[0052] Based on the data fusion node, the physiological data of the strongly correlated health monitoring dimensions are correlated and fused, and the correlation fusion results are input into the pre-trained health status assessment model for analysis;

[0053] Based on the analysis results, the health status of the target patient at the cycle node is obtained.

[0054] Preferably, an artificial intelligence-based health management method for chronic disease patients inputs the correlation fusion results into a pre-trained health status assessment model for analysis, including:

[0055] Based on the medical knowledge system, a health assessment indicator system is obtained under different health monitoring dimensions. At the same time, historical physiological data is retrieved from the preset medical database and correlation analysis is performed on the historical physiological data based on the health assessment indicator system.

[0056] Determine the evaluation effect of the health assessment index system based on the difference between the correlation analysis results and the baseline results of historical physiological data, and construct a health assessment function corresponding to the health assessment index system based on the correlation analysis results when the preset conditions are met;

[0057] The health assessment function is deployed in the preset model framework, and the received association fusion results are analyzed based on the scenario deployment results to obtain the health status of the target patient at the cycle node.

[0058] Preferably, an artificial intelligence-based health management method for chronic disease patients obtains the health status of the target patient at a cycle node, including:

[0059] Determine the influence weights of different health monitoring dimensions on health status based on the health assessment function, and determine the contribution of different health monitoring dimensions to health status based on the influence weights;

[0060] Generate a health description for each health monitoring dimension based on the contribution, and associate the health description with the corresponding health monitoring dimension;

[0061] Generate a health status assessment report based on the correlation results.

[0062] The present invention provides an artificial intelligence-based health management system for chronic disease patients, comprising:

[0063] The data acquisition module is used to obtain health monitoring dimensions and perform periodic monitoring on the target patient according to the health monitoring dimensions to obtain the physiological data of each health monitoring dimension at each cycle node;

[0064] The health status determination module is used to perform correlation and fusion analysis on the physiological data of each health monitoring dimension within the cycle node to obtain the health status of the target patient at the cycle node;

[0065] The health management module is used to generate dynamic health management strategies based on the health status of each periodic node.

[0066] Compared with the prior art, the present invention has the following beneficial effects:

[0067] By obtaining the health monitoring dimensions, we can effectively conduct periodic monitoring of the target patients according to the health monitoring dimensions, obtain the physiological data of each health monitoring dimension in each cycle node, and accurately obtain the health status of the target patients at the cycle nodes by performing correlation fusion analysis on the physiological data of each health monitoring dimension in the cycle nodes. Then, dynamic health management strategies are generated according to the health status of each cycle node, which is conducive to all-round health monitoring of the target patients. The accuracy of the obtained health status is effectively guaranteed through correlation fusion analysis, which is conducive to providing reliable data support for the accurate generation of dynamic health management strategies, and ensuring the effectiveness, accuracy and intelligence of health management for patients with chronic diseases.

[0068] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be achieved and obtained through the structures specifically pointed out in this application document.

[0069] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0071] Figure 1 This is a flow chart of a method for health management of chronic disease patients based on artificial intelligence in an embodiment of the present invention;

[0072] Figure 2 This is a flowchart of step 1 in a method for health management of chronic disease patients based on artificial intelligence in an embodiment of the present invention;

[0073] Figure 3 This is a structural diagram of an artificial intelligence-based health management system for chronic disease patients in an embodiment of the present invention. DETAILED DESCRIPTION

[0074] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0075] Example 1:

[0076] This embodiment provides a health management method for chronic disease patients based on artificial intelligence, such as Figure 1 Shown, including:

[0077] Step 1: Obtain health monitoring dimensions, and perform periodic monitoring on the target patient according to the health monitoring dimensions to obtain physiological data of each health monitoring dimension at each cycle node;

[0078] Step 2: Perform correlation fusion analysis on the physiological data of each health monitoring dimension within the cycle node to obtain the health status of the target patient at the cycle node;

[0079] Step 3: Generate a dynamic health management strategy based on the health status of each periodic node.

[0080] In this embodiment, the health monitoring dimension is used to characterize the type of health monitoring performed on the target patient, such as blood sugar, blood pressure, etc., wherein the target patient is a patient with a chronic disease.

[0081] In this embodiment, periodic monitoring refers to collecting data within a preset period interval according to the health monitoring dimension to obtain physiological data within each period node, wherein the preset period is set in advance, for example, one week.

[0082] In this embodiment, the association fusion analysis refers to determining the association relationship between each health monitoring dimension and performing a fusion analysis on the physiological data under different health monitoring dimensions through the association relationship between each health monitoring dimension, in order to ensure the accuracy of determining the health status.

[0083] In this embodiment, the dynamic health management strategy is used to generate guiding opinions on regulating health based on health status, including various opinions such as exercise, diet and rest. For example, increasing exercise can control blood pressure, and reducing carbohydrate intake in the diet can control blood sugar.

[0084] In this embodiment, a dynamic health management strategy is generated based on the health status of each periodic node. Each periodic node has a corresponding health management strategy. The health management strategy of each periodic node is obtained after adjusting the standard health management strategy when an abnormal health monitoring dimension occurs.

[0085] The working principle and beneficial effects of the above technical solution are: by obtaining health monitoring dimensions, the target patients can be effectively monitored periodically according to the health monitoring dimensions, and the physiological data of each health monitoring dimension in each cycle node can be obtained. By performing correlation fusion analysis on the physiological data of each health monitoring dimension in the cycle node, the health status of the target patients at the cycle node can be accurately obtained, and then a dynamic health management strategy can be generated according to the health status of each cycle node, which is conducive to all-round health monitoring of the target patients. The accuracy of the obtained health status can be effectively guaranteed through correlation fusion analysis, which is conducive to providing reliable data support for the accurate generation of dynamic health management strategies, and ensuring the effectiveness, accuracy and intelligence of health management for patients with chronic diseases.

[0086] Example 2:

[0087] Based on Example 1, this example provides a health management method for chronic disease patients based on artificial intelligence, such as Figure 2 As shown, in step 1, the health monitoring dimensions are obtained, including:

[0088] Step 101: Collect the medical records of the target patient and identify the medical records to determine the case information of the target patient;

[0089] Step 102: Input the case information into a preset recognition model for recognition to obtain target keywords of the case information;

[0090] Step 103: Determine the main health monitoring dimensions based on the target keywords;

[0091] Step 104: Input the target keyword into the preset health management map for matching, and determine the association path with the target keyword in the preset health management map;

[0092] Step 105: determining nodes related to the target keyword based on the association path between the target keyword and the target keyword;

[0093] Step 106: Read node information of relevant nodes and determine the secondary health monitoring dimension based on the node information;

[0094] Step 107: Complete the acquisition of health monitoring dimensions based on the primary monitoring dimensions and the secondary monitoring dimensions.

[0095] In this embodiment, the preset recognition model is a model set in advance and used to perform keyword recognition on case information, where the keywords include: blood pressure, blood sugar, blood lipids and other names.

[0096] In this embodiment, the preset health management graph is set in advance to represent different keywords and the association relationship between different keywords, wherein the association relationship is represented by an association path to realize the connection between different keywords.

[0097] In this embodiment, the main detection dimension is the recognition result of the keywords based on the case information input into the preset recognition model, that is, the detection index that needs to be focused on, and the secondary detection dimension is the keywords associated with the main detection dimension determined after the main detection dimension is positioned based on the preset management graph.

[0098] The above technical solution has the beneficial effect that the acquisition of the monitoring dimension can be effectively realized by determining the main monitoring dimension and the secondary monitoring dimension, thereby guaranteeing the comprehensiveness of the monitoring management.

[0099] Embodiment 3

[0100] Based on the embodiment 2, the embodiment provides a chronic disease patient health management method based on artificial intelligence, which identifies the medical record sheet to determine the case information of the target patient, including:

[0101] The first image of the medical record sheet is collected, and the first image is subjected to character scanning, and the first image is split according to the scanning result to obtain a second image;

[0102] The information type for recognizing the case information is acquired, and the identification point corresponding to the information type is acquired;

[0103] The identification point is positioned in the second image;

[0104] Meanwhile, the first annotation is performed in the second image according to the positioning result;

[0105] The position logic of the character in the second image is acquired, and the character information associated with the first annotation result is determined according to the position logic;

[0106] The character information associated with the first annotation result is subjected to second annotation in the second image;

[0107] The first annotation result and the second annotation result are extracted in the second image to obtain the case information of the target patient.

[0108] In this embodiment, the first image is the initial image corresponding to the medical record sheet.

[0109] In this embodiment, the first image is split to split the character region and the non-character region in the first image, wherein the obtained second image is the character region in the first image.

[0110] In this embodiment, the information type is pre-set and represents the type of information used to identify it, such as user personal information.

[0111] In this embodiment, the identification point refers to a specific symbol corresponding to the information type.

[0112] In this embodiment, the first annotation refers to marking and displaying specific information located in the second image.

[0113] In this embodiment, the position logic refers to information such as the specific position and sequence of the characters in the second image.

[0114] In this embodiment, the second annotation refers to marking and displaying text information associated with the first annotation result.

[0115] The beneficial effect of the above technical solution is to ensure the comprehensiveness, accuracy and reliability of the case information extraction of the target patient.

[0116] Example 4:

[0117] Based on Example 1, this embodiment provides an artificial intelligence-based health management method for chronic disease patients. In step 1, the target patient is periodically monitored according to the health monitoring dimension, and physiological data of each health monitoring dimension in each periodic monitoring node is obtained, including:

[0118] Preset monitoring period and set periodic monitoring nodes according to the preset monitoring period;

[0119] Read the health monitoring dimensions and start the monitoring equipment according to the health monitoring dimensions;

[0120] Collect sub-physiological data corresponding to the health monitoring dimension based on the monitoring device, obtain the dimension identifier of the health monitoring dimension, and construct a data storage node according to the dimension identifier of the health monitoring dimension;

[0121] The sub-physiological data is stored in the corresponding data storage node according to the dimension identifier, and the sub-physiological data in the data storage node is divided according to the periodic monitoring node, and the sub-physiological data segments of several periodic monitoring nodes are determined in the data storage node;

[0122] The sub-physiological data segments of several periodic monitoring nodes are the physiological data of each health monitoring dimension in each periodic monitoring node.

[0123] In this embodiment, the periodic monitoring node refers to a specific monitoring time point determined according to a preset monitoring period.

[0124] In this embodiment, sub-physiological data refers to the results obtained by collecting physiological data under the health monitoring dimension through monitoring equipment.

[0125] In this embodiment, the dimension identifier refers to a marking symbol for distinguishing different health monitoring dimensions.

[0126] In this embodiment, the data storage nodes are constructed according to the dimension identifiers of the health monitoring dimensions, and the sub-physiological data of different health monitoring dimensions correspond to different data storage nodes.

[0127] The beneficial effects of the above technical solution are: by determining the periodic monitoring nodes, the physiological data can be effectively divided based on time transformation, and the physiological data of each health dimension can be effectively collected and managed.

[0128] Example 5:

[0129] Based on Example 1, this embodiment provides an artificial intelligence-based health management method for chronic disease patients. In step 3, a dynamic health management strategy is generated based on the health status of each cycle node, including:

[0130] Sort the health status of each periodic node in chronological order to obtain a health status sequence;

[0131] Read the health status of each periodic node, obtain the benchmark health interval, and match the health status sequence with the benchmark health interval respectively;

[0132] When there is a first target health state in the health state sequence that belongs to the benchmark health interval, adding a first label to the first target health state;

[0133] When there is a second target health state in the health state sequence that does not belong to the benchmark health interval, a second label is added to the second target health state;

[0134] Retrieving a first target health management policy based on a first tag from a preset policy management library;

[0135] At the same time, the physiological data corresponding to each health monitoring dimension of the target patient under the second tag is read, and the abnormal health monitoring dimension of the target patient under the second tag is determined according to the standard data interval of the physiological data corresponding to each health monitoring dimension;

[0136] Retrieve the second target health management strategy corresponding to the abnormal health monitoring dimension from the preset strategy management library;

[0137] Based on the first target health management strategy and the second target health management strategy, dynamic adjustment is performed according to the health status sequence to generate a dynamic health management strategy for each periodic node.

[0138] In this embodiment, the health status sequence is based on the result of sorting the health status in chronological order.

[0139] In this embodiment, the reference health interval can be set in advance as a standard for measuring whether the health state sequence is healthy.

[0140] In this embodiment, the first target health state refers to a health state in the health state sequence that belongs to the reference health interval, wherein the first label is an identifier representing that the health state belongs to the reference health interval, and is used to distinguish health states that do not belong to the reference health interval, wherein the health state that does not belong to the reference health interval in the health state sequence is the second target health state, and the second target health state is added with a second label, and the second label is an identifier representing that the health state does not belong to the reference health interval, and is used to distinguish health states that belong to the reference health interval.

[0141] In this embodiment, the preset strategy management library is set in advance to save different health management strategies.

[0142] In this embodiment, the first target health management strategy is a health management strategy for the current cycle node when the health state satisfies the reference health interval, which is retrieved from the preset strategy management library. When the heart rate under the health state is 65 and the health data interval is 60-100, the first target health management strategy is a method for user health management under normal heart rate.

[0143] In this embodiment, the standard data interval is a data range allowed by different health monitoring dimensions under normal conditions, wherein the health monitoring dimension not in the standard data interval is an abnormal health monitoring dimension.

[0144] In this embodiment, the second target health management strategy is a health management strategy corresponding to an abnormal condition.

[0145] The beneficial effects of the above technical solutions are: effectively generating corresponding health management strategies under different health states, thereby effectively adjusting according to the health state sequence, and ensuring the effectiveness and accuracy of the generated dynamic health management strategies for each cycle node.

[0146] Embodiment 6:

[0147] Based on embodiment 1, the present embodiment provides a chronic disease patient health management method based on artificial intelligence, and after generating the dynamic health management strategy according to the health state of each cycle node in step 3, the method comprises:

[0148] A first visual window and a second visual window are constructed.

[0149] The health state of each cycle node is first displayed in the first visual window, and the dynamic health management strategy is second displayed in the second visual window.

[0150] The beneficial effects of the above technical solutions are: it is beneficial to intuitively and accurately see the current health status and the corresponding dynamic health management strategy.

[0151] Embodiment 7:

[0152] Based on the embodiment 1, the embodiment provides a chronic disease patient health management method based on artificial intelligence, in step 2, the physiological data of each health monitoring dimension in the cycle node is associated and fused to obtain the health status of the target patient in the cycle node, including:

[0153] Obtain the physiological data under each health monitoring dimension in each cycle node, and determine the normalization standard of the physiological data based on the physiological properties between each health monitoring dimension;

[0154] Based on the normalization standard, the physiological data is uniformly formatted, and the physiological significance of each health monitoring dimension is extracted, and the feature extraction limited index is generated based on the physiological significance;

[0155] Based on the feature extraction limited index, the physiological data of the corresponding health monitoring dimension is scanned for features, the corresponding feature index and the corresponding feature index quantization interval of each health monitoring dimension are obtained, and the feature index and the feature index quantization interval are associated to obtain the single dimension feature of each health monitoring dimension;

[0156] Access the server to call the health management system, and determine the mutual relationship between different health monitoring dimensions based on the health management system;

[0157] Based on the mutual relationship, the single dimension feature of different health monitoring dimensions is analyzed, and the cross dimension feature is constructed based on the analysis result;

[0158] Based on the single dimension feature and the cross dimension feature, the physiological data of different health monitoring dimensions is sequentially changed by a single amount, and the physiological change amount of the remaining health monitoring dimension is monitored in real time based on the single amount change result;

[0159] Based on the variable threshold value of the physiological change amount, the strongly associated health monitoring dimensions are determined, and the data fusion node between the strongly associated health monitoring dimensions is determined based on the medical knowledge system;

[0160] Based on the data fusion node, the physiological data of the strongly associated health monitoring dimensions is associated and fused, and the associated fusion result is input into the pre-trained health status evaluation model for analysis;

[0161] Based on the analysis result, the health status of the target patient in the cycle node is obtained.

[0162] In this embodiment, the physiological property refers to the association relationship or mutual limiting relationship between different health monitoring dimensions.

[0163] In this embodiment, the normalization standard refers to unifying the value range and format requirements of physiological data under different health monitoring dimensions.

[0164] In this embodiment, physiological significance refers to the physiological monitoring purpose and monitoring focus of each health monitoring dimension. For example, heart rate monitoring can determine the user's cardiopulmonary function, and different value intervals correspond to different cardiopulmonary functions.

[0165] In this embodiment, the feature extraction limiting index is generated based on physiological significance, that is, a reference basis for feature extraction of corresponding physiological data.

[0166] In this embodiment, the characteristic index refers to the corresponding specific characteristic information obtained after reading and extracting the characteristics of the physiological data according to the characteristic extraction limiting index.

[0167] In this embodiment, the characteristic indicator quantization interval refers to the value range corresponding to the characteristic indicator.

[0168] In this embodiment, the single-dimensional feature refers to the association result between each characteristic indicator and the corresponding characteristic indicator quantization interval, which is used to characterize the specific situation of each characteristic indicator, that is, the specific situation corresponding to each monitoring dimension.

[0169] In this embodiment, the health management system is obtained from the server and is used to represent the items or focuses that need to be managed when performing health management.

[0170] In this embodiment, cross-dimensional features refer to correlation features between different health monitoring dimensions.

[0171] In this embodiment, a single quantitative change refers to sequentially changing the values ​​of physiological data of different health monitoring dimensions, that is, when changing the values, only the value of one physiological data can be changed at a time.

[0172] In this embodiment, the physiological change refers to the specific changes in other health monitoring dimensions that occur based on a change in the value of one data.

[0173] In this embodiment, the variable threshold value refers to the specific degree of change in the physiological change amount of other health monitoring dimensions.

[0174] In this embodiment, a strongly associated health monitoring dimension refers to a health monitoring dimension in which the value of the physiological change exceeds a preset threshold value.

[0175] In this embodiment, the medical knowledge system is known in advance and is used to record the associations between different health monitoring dimensions.

[0176] In this embodiment, the data fusion node refers to a specific data point for associating and fusing physiological data between strongly correlated health monitoring dimensions.

[0177] The beneficial effects of the above technical solution are: by processing and analyzing the physiological data under each health monitoring dimension in each cycle node, the data fusion nodes between different health monitoring dimensions can be effectively determined, the physiological data of the health monitoring dimensions can be associated and fused according to the data fusion nodes, and the associated fusion results can be analyzed to achieve accurate and effective determination of the health status of the target patient at the cycle node, thereby improving the accuracy and reliability of health status assessment.

[0178] Example 8:

[0179] Based on Example 7, this example provides an artificial intelligence-based health management method for chronic disease patients, which inputs the association fusion results into a pre-trained health status assessment model for analysis, including:

[0180] Based on the medical knowledge system, a health assessment indicator system is obtained under different health monitoring dimensions. At the same time, historical physiological data is retrieved from the preset medical database and correlation analysis is performed on the historical physiological data based on the health assessment indicator system.

[0181] Determine the evaluation effect of the health assessment index system based on the difference between the correlation analysis results and the baseline results of historical physiological data, and construct a health assessment function corresponding to the health assessment index system based on the correlation analysis results when the preset conditions are met;

[0182] The health assessment function is deployed in the preset model framework, and the received association fusion results are analyzed based on the scenario deployment results to obtain the health status of the target patient at the cycle node.

[0183] In this embodiment, the health assessment index system refers to a set of indicators for health assessment of corresponding health monitoring dimensions. It is based on the medical knowledge system and determines relevant health assessment indicators from different health monitoring dimensions. These indicators are the basis for comprehensive assessment of health status. They are interrelated and influence each other, and together reflect the health status of the human body.

[0184] In this embodiment, correlation analysis of historical physiological data based on the health assessment index system can be to determine the specific nature of the correlation between historical physiological data through the health assessment index system. For example, there is a correlation between blood pressure and liver and kidney function indicators, and the specific nature of the correlation (such as positive correlation, negative correlation, etc.) is determined through data analysis.

[0185] In this embodiment, the baseline results of the historical physiological data may be standard values ​​obtained based on clinical studies of normal conditions, and are set in advance.

[0186] In this embodiment, the preset condition is set in advance and is used to evaluate whether to construct a health assessment indicator system based on the correlation results. For example, when the assessment effect is equal to or greater than 80%, the preset condition is met.

[0187] In this embodiment, the health assessment function may be constructed by traversing different health assessment indicators and performing weighted summation.

[0188] In this embodiment, the preset model framework is set in advance, such as a health management system framework. After deployment, when the associated fusion result is received (the result of fusing the newly collected multiple health data), it is substituted into the health assessment function for calculation to obtain the health status of the target patient at the cycle node.

[0189] The beneficial effects of the above technical solution are: by constructing a health assessment indicator system from multiple health monitoring dimensions and conducting correlation analysis, various factors affecting health and their interrelationships can be considered more comprehensively; by conducting correlation analysis and constructing a health assessment function based on the patient's historical physiological data, personalized health assessment can be achieved, which is conducive to discovering the changing trends of health risk factors in advance by analyzing the correlation between health assessment indicators.

[0190] Example 9:

[0191] Based on Example 8, this example provides an artificial intelligence-based health management method for chronic disease patients, which obtains the health status of the target patient at a periodic node, including:

[0192] Determine the influence weights of different health monitoring dimensions on health status based on the health assessment function, and determine the contribution of different health monitoring dimensions to health status based on the influence weights;

[0193] Generate a health description for each health monitoring dimension based on the contribution, and associate the health description with the corresponding health monitoring dimension;

[0194] Generate a health status assessment report based on the correlation results.

[0195] In this embodiment, the contribution is obtained based on the product of the influence weight and the actual value (normalized) of the corresponding health monitoring dimension.

[0196] In this embodiment, a corresponding health description is generated based on the contribution of each health monitoring dimension. For health monitoring dimensions with high contribution, if their values ​​are within the normal range, the health description may emphasize the positive role of the dimension in maintaining health; if the values ​​deviate from the normal range, the health description will point out possible health risks.

[0197] In this embodiment, based on the correlation results, the health descriptions of all health monitoring dimensions are integrated to generate a comprehensive health status assessment report. This report can intuitively present the status of individuals or groups in each health monitoring dimension and the comprehensive impact of these conditions on the overall health status.

[0198] The beneficial effects of the above technical solution are: effectively ensuring the comprehensiveness and accuracy of the generated health status assessment report, effectively providing personalized health assessments for different individuals, and ensuring individuals' attention to health.

[0199] Example 10:

[0200] This embodiment provides a health management system for chronic disease patients based on artificial intelligence, such as Figure 3 Shown, including:

[0201] The data acquisition module is used to obtain health monitoring dimensions and perform periodic monitoring on the target patient according to the health monitoring dimensions to obtain the physiological data of each health monitoring dimension at each cycle node;

[0202] The health status determination module is used to perform correlation and fusion analysis on the physiological data of each health monitoring dimension within the cycle node to obtain the health status of the target patient at the cycle node;

[0203] The health management module is used to generate dynamic health management strategies based on the health status of each periodic node.

[0204] The working principle and beneficial effects of the above technical solution are: by obtaining health monitoring dimensions, the target patients can be effectively monitored periodically according to the health monitoring dimensions, and the physiological data of each health monitoring dimension in each cycle node can be obtained. By performing correlation fusion analysis on the physiological data of each health monitoring dimension in the cycle node, the health status of the target patients at the cycle node can be accurately obtained, and then a dynamic health management strategy can be generated according to the health status of each cycle node, which is conducive to all-round health monitoring of the target patients. The accuracy of the obtained health status can be effectively guaranteed through correlation fusion analysis, which is conducive to providing reliable data support for the accurate generation of dynamic health management strategies, and ensuring the effectiveness, accuracy and intelligence of health management for patients with chronic diseases.

[0205] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A health management method for chronic disease patients based on artificial intelligence, characterized in that: include: Step 1: Obtain health monitoring dimensions, and perform periodic monitoring on the target patient according to the health monitoring dimensions to obtain physiological data of each health monitoring dimension at each cycle node; Step 2: Perform correlation fusion analysis on the physiological data of each health monitoring dimension within the cycle node to obtain the health status of the target patient at the cycle node; Step 3: Generate a dynamic health management strategy based on the health status of each cycle node; In step 1, the health monitoring dimension is obtained, including: collecting the medical record of the target patient and identifying the medical record to determine the case information of the target patient; inputting the case information into a preset recognition model for identification to obtain the target keyword of the case information; determining the main health monitoring dimension according to the target keyword; inputting the target keyword into a preset health management map for matching to determine the association path with the target keyword in the preset health management map; determining the relevant node with the target keyword according to the association path between the target keyword and the target keyword; reading the node information of the relevant node and determining the secondary health monitoring dimension according to the node information; completing the acquisition of the health monitoring dimension according to the main monitoring dimension and the secondary monitoring dimension; In step 2, the physiological data of each health monitoring dimension in the cycle node are correlated and fused to obtain the health status of the target patient at the cycle node, including: obtaining the physiological data under each health monitoring dimension in each cycle node, and determining the normalization standard for the physiological data based on the physiological attributes between each health monitoring dimension; unifying the format of the physiological data based on the normalization standard, and at the same time, extracting the physiological significance of each health monitoring dimension, and generating feature extraction limiting indicators based on the physiological significance; performing feature scanning on the physiological data of the corresponding health monitoring dimension based on the feature extraction limiting indicators to obtain the corresponding feature indicators and the corresponding feature indicator quantization intervals of each health monitoring dimension, and associating the feature indicators with the feature indicator quantization intervals to obtain the single-dimensional features of each health monitoring dimension; accessing the server, retrieving the health management system, and determining the relationship between different health monitoring dimensions based on the health management system; parsing the single-dimensional features of different health monitoring dimensions based on the mutual relationship, and constructing cross-dimensional features based on the parsing results; performing single quantitative changes on the physiological data of different health monitoring dimensions based on the single-dimensional features and cross-dimensional features, and monitoring the remaining in real time based on the single quantitative change results. Physiological changes in health monitoring dimensions; determining strongly correlated health monitoring dimensions based on variable threshold values ​​of physiological changes, and determining data fusion nodes between strongly correlated health monitoring dimensions based on the medical knowledge system; correlating and fusing physiological data of strongly correlated health monitoring dimensions based on the data fusion nodes, and inputting the correlation fusion results into a pre-trained health status assessment model for analysis; obtaining the health status of the target patient at the cycle node based on the analysis results; inputting the correlation fusion results into the pre-trained health status assessment model for analysis, including: obtaining a health assessment indicator system under different health monitoring dimensions based on the medical knowledge system, and at the same time, retrieving historical physiological data from a preset medical database, and performing correlation analysis on the historical physiological data based on the health assessment indicator system; determining the evaluation effect of the health assessment indicator system based on the difference between the correlation analysis results and the baseline results of the historical physiological data, and constructing a health assessment function corresponding to the health assessment indicator system based on the correlation analysis results when the preset conditions are met; deploying the health assessment function in a preset model framework, and analyzing the received correlation fusion results based on the scenario deployment results to obtain the health status of the target patient at the cycle node.

2. The method for health management of chronic disease patients based on artificial intelligence according to claim 1, characterized in that: Identify the medical records to determine the target patient's case information, including: Acquire a first image of the medical record, perform text scanning on the first image, and split the first image according to the scanning result to obtain a second image; Obtain the information type for identifying the case information and obtain the identification point corresponding to the information type; Positioning the marker point in the second image; At the same time, a first annotation is performed in the second image according to the positioning result; Obtaining position logic of the text in the second image, and determining text information associated with the first annotation result based on the position logic; Performing a second annotation on the second image using text information associated with the first annotation result; The first annotation result and the second annotation result are extracted from the second image to obtain case information of the target patient.

3. The method for health management of chronic disease patients based on artificial intelligence according to claim 1, characterized in that: In step 1, the target patient is periodically monitored according to the health monitoring dimension to obtain physiological data of each health monitoring dimension in each periodic monitoring node, including: Preset monitoring period and set periodic monitoring nodes according to the preset monitoring period; Read the health monitoring dimensions and start the monitoring equipment according to the health monitoring dimensions; Collect sub-physiological data corresponding to the health monitoring dimension based on the monitoring device, obtain the dimension identifier of the health monitoring dimension, and construct a data storage node according to the dimension identifier of the health monitoring dimension; The sub-physiological data is stored in the corresponding data storage node according to the dimension identifier, and the sub-physiological data in the data storage node is divided according to the periodic monitoring node, and the sub-physiological data segments of several periodic monitoring nodes are determined in the data storage node; The sub-physiological data segments of several periodic monitoring nodes are the physiological data of each health monitoring dimension in each periodic monitoring node.

4. The method for health management of chronic disease patients based on artificial intelligence according to claim 1, characterized in that: In step 3, a dynamic health management strategy is generated based on the health status of each periodic node, including: Sort the health status of each periodic node in chronological order to obtain a health status sequence; Read the health status of each periodic node, obtain the benchmark health interval, and match the health status sequence with the benchmark health interval respectively; When there is a first target health state in the health state sequence that belongs to the benchmark health interval, adding a first label to the first target health state; When there is a second target health state in the health state sequence that does not belong to the benchmark health interval, a second label is added to the second target health state; Retrieving a first target health management policy based on a first tag from a preset policy management library; At the same time, the physiological data corresponding to each health monitoring dimension of the target patient under the second tag is read, and the abnormal health monitoring dimension of the target patient under the second tag is determined according to the standard data interval of the physiological data corresponding to each health monitoring dimension; Retrieve the second target health management strategy corresponding to the abnormal health monitoring dimension from the preset strategy management library; Based on the first target health management strategy and the second target health management strategy, dynamic adjustment is performed according to the health status sequence to generate a dynamic health management strategy for each periodic node.

5. The method for health management of chronic disease patients based on artificial intelligence according to claim 1, characterized in that: In step 3, after generating a dynamic health management strategy based on the health status of each periodic node, the following steps are included: Constructing a first visual window and a second visual window; The health status of each periodic node is first displayed in the first visual window, and at the same time, the dynamic health management strategy is secondly displayed in the second visual window.

6. The method for health management of chronic disease patients based on artificial intelligence according to claim 1, characterized in that: Get the health status of the target patient at the cycle node, including: Determine the influence weights of different health monitoring dimensions on health status based on the health assessment function, and determine the contribution of different health monitoring dimensions to health status based on the influence weights; Generate a health description for each health monitoring dimension based on the contribution, and associate the health description with the corresponding health monitoring dimension; Generate a health status assessment report based on the correlation results.

7. A health management system for chronic disease patients based on artificial intelligence, characterized in that: include: The data acquisition module is used to obtain health monitoring dimensions and perform periodic monitoring on the target patient according to the health monitoring dimensions to obtain the physiological data of each health monitoring dimension at each cycle node; The health status determination module is used to perform correlation and fusion analysis on the physiological data of each health monitoring dimension within the cycle node to obtain the health status of the target patient at the cycle node; Health management module, used to generate dynamic health management strategies based on the health status of each cycle node; In the data acquisition module, the health monitoring dimensions are obtained, including: collecting the medical records of the target patients and identifying the medical records to determine the case information of the target patients; inputting the case information into the preset recognition model for identification to obtain the target keywords of the case information; determining the main health monitoring dimensions according to the target keywords; inputting the target keywords into the preset health management map for matching to determine the association path with the target keywords in the preset health management map; determining the relevant nodes with the target keywords according to the association path between the target keywords and the target keywords; reading the node information of the relevant nodes and determining the secondary health monitoring dimensions according to the node information; completing the acquisition of the health monitoring dimensions according to the main monitoring dimensions and the secondary monitoring dimensions; In the health status determination module, the physiological data of each health monitoring dimension in the cycle node are correlated and fused and analyzed to obtain the health status of the target patient at the cycle node, including: obtaining the physiological data under each health monitoring dimension in each cycle node, and determining the normalization standard for the physiological data based on the physiological attributes between each health monitoring dimension; unifying the format of the physiological data based on the normalization standard, and at the same time, extracting the physiological significance of each health monitoring dimension, and generating feature extraction limiting indicators based on the physiological significance; performing feature scanning on the physiological data of the corresponding health monitoring dimension based on the feature extraction limiting indicators to obtain the corresponding feature indicators and the corresponding feature indicator quantization intervals of each health monitoring dimension, and associating the feature indicators with the feature indicator quantization intervals to obtain the single-dimensional features of each health monitoring dimension; accessing the server, retrieving the health management system, and determining the relationship between different health monitoring dimensions based on the health management system; parsing the single-dimensional features of different health monitoring dimensions based on the mutual relationship, and constructing cross-dimensional features based on the parsing results; performing single quantitative changes on the physiological data of different health monitoring dimensions based on single-dimensional features and cross-dimensional features, and monitoring in real time based on the single quantitative change results. Measure the physiological changes of the remaining health monitoring dimensions; determine the strongly correlated health monitoring dimensions based on the variable threshold values ​​of the physiological changes, and determine the data fusion nodes between the strongly correlated health monitoring dimensions based on the medical knowledge system; correlate and fuse the physiological data of the strongly correlated health monitoring dimensions based on the data fusion nodes, and input the correlated fusion results into the pre-trained health status assessment model for analysis; obtain the health status of the target patient at the cycle node based on the analysis results; input the correlated fusion results into the pre-trained health status assessment model for analysis, including: obtaining a health assessment indicator system under different health monitoring dimensions based on the medical knowledge system, and at the same time, retrieving historical physiological data from a preset medical database, and performing correlation analysis on the historical physiological data based on the health assessment indicator system; determine the evaluation effect of the health assessment indicator system based on the difference between the correlation analysis results and the baseline results of the historical physiological data, and when the preset conditions are met, construct a health assessment function corresponding to the health assessment indicator system based on the correlation analysis results; deploy the health assessment function in the preset model framework for a scenario, and analyze the received correlated fusion results based on the scenario deployment results to obtain the health status of the target patient at the cycle node.

Images