Chronic disease information management system based on behavioral interaction model

Abstract

The application belongs to the technical field of intelligent medical treatment, and discloses a chronic disease information management system based on a behavior interaction model; the system comprises the following steps: acquiring background information, internal motivation and cognitive evaluation; based on the background information, a health education content library is constructed, and health education content is pushed in combination with the cognitive evaluation; background information and internal motivation are fused to generate a multi-dimensional treatment scheme matrix; patient willingness information is acquired, the acceptance degree of different treatment schemes is quantitatively evaluated, and the best treatment scheme is screened out; real-time interaction data are acquired and analyzed, and emotional support strategies are dynamically formulated; patient health data are collected, and the interaction effect is evaluated; according to the interaction effect, the health education content, the best treatment scheme and the emotional support strategies are intelligently optimized in sequence; the application takes behavior interaction as the core, realizes precise management of the whole life cycle of chronic disease patients, and thus significantly improves the health management effect of the patients and reduces the risk of complications.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent medical treatment, more particularly, to a chronic disease information management system based on a behavior interaction model. BACKGROUND

[0002] With the acceleration of population aging and the change of lifestyle, chronic diseases have gradually become the main component of global disease burden, bringing great challenges to the medical and health system. Among them, chronic diseases such as systemic lupus erythematosus and muscle reduction have become the focus and difficulty of medical management due to their complexity, long-term nature and multi-system involvement. The traditional chronic disease management mode mainly relies on hospital outpatient follow-up, which has obvious limitations: short doctor-patient contact time, long follow-up interval, poor patient compliance, formal health education, and inability to achieve continuous monitoring. This fragmented and passive medical service mode cannot meet the long-term, continuous and personalized health management needs of chronic disease patients.

[0003] In recent years, the rapid development of intelligent medical treatment provides a new way to solve the above problems. The wide application of information technology such as mobile Internet, big data and cloud computing makes it possible to break through the time and space limitations of medical services. For example, the patent with publication number CN111933278A discloses a chronic disease management system based on artificial intelligence, which includes: entering the basic information of patients, various examination results and medical advice; evaluating the chronic disease condition of patients according to their basic information and corresponding examination results; developing a patient treatment plan according to the evaluation results; developing a patient re-examination plan according to the evaluation results; managing patient re-examination data, and starting a short message warning module for warning when a patient is found to have not completed the re-examination plan within the preset time. This invention can realize the full monitoring and recording of the chronic disease treatment process, improve the efficiency of chronic disease management, and provide protection for on-demand treatment of chronic diseases.

[0004] However, the above technology, although it realizes chronic disease information management, still stays at the level of information collection and passive monitoring, ignores the active participation and interactive feedback mechanism of patients in the whole health behavior change process, lacks effective guidance for patient behavior change, and leads to difficulty in continuous patient health behavior, low treatment compliance, and ultimately affects the effect of chronic disease control, increases the risk of complications, and is difficult to fundamentally solve the core problem of long-term management of chronic diseases.

[0005] In view of this, the present application provides a chronic disease information management system based on a behavior interaction model to solve the above problems. SUMMARY

[0006] In order to overcome the above-mentioned defects of the prior art, in order to achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0007] The chronic disease information management system based on the behavioral interaction model comprises:

[0008] A data acquisition module is configured to acquire patient characteristic data, the patient characteristic data comprising background information, intrinsic motivation and cognitive evaluation;

[0009] An information pushing module is configured to construct a health education content library based on the background information, and to push health education content by using a hierarchical pushing strategy in combination with the cognitive evaluation;

[0010] A professional support module is configured to fuse the background information and the intrinsic motivation, to generate a multi-dimensional treatment scheme matrix by using a multi-criteria decision analysis method;

[0011] A decision control module is configured to acquire patient willingness information, to quantitatively evaluate the acceptance degree of different treatment schemes in the multi-dimensional treatment scheme matrix, and to screen out a best treatment scheme from the multi-dimensional treatment scheme matrix according to the evaluation result;

[0012] An emotional support module is configured to acquire real-time interaction data, to analyze the real-time interaction data by using an emotional analysis technology, to dynamically formulate an emotional support strategy, and to provide emotional support through a pre-established multi-level emotional support network;

[0013] An interaction evaluation module is configured to collect patient health data, and to comprehensively evaluate the interaction effect by using a deep learning technology;

[0014] A feedback optimization module is configured to intelligently optimize the health education content, the best treatment scheme and the emotional support strategy in sequence through a closed-loop feedback mechanism according to the interaction effect.

[0015] Further, the background information comprises basic conditions, disease history and living habits; the disease history comprises past medical history and chronic disease condition; and the chronic disease condition comprises disease name, disease time and severity;

[0016] The method for constructing the health education content library comprises:

[0017] The content resource library is constructed, and the health education content corresponding to different chronic diseases is included in the content resource library; each health education content in the content resource library corresponds to a group of content data, the content data includes a chronic disease name, a cognitive difficulty and a health field, according to the chronic disease name in the background information, the corresponding health education content is screened out from all the health education content in the content resource library, and is marked as suitable content; it is judged whether the health education content is pushed to the patient; if yes, the historical content is obtained, the clustering algorithm is used to cluster the historical content and the suitable content, a content cluster is obtained, a is an integer greater than 1; the health education content library is constructed according to the suitable content corresponding to the historical content in the content cluster; if not, the historical patient is obtained, and the clustering algorithm is used to cluster the historical patient and the current patient according to the patient characteristic data of the historical patient and the current patient, b patient clusters are obtained, b is an integer greater than 1; the health education content library is constructed according to the suitable content corresponding to the historical patient in the patient cluster corresponding to the current patient; wherein the historical content is the health education content pushed to the patient at the historical moment, the historical patient is the patient who has received chronic disease management at the historical moment, and the current patient is the patient who is currently undergoing chronic disease management;

[0018] The method for pushing the health education content comprises:

[0019] The health education content corresponding to the same health field in the health education content library is taken as a group of content sets; the cognitive difficulty corresponding to each health education content in each group of content sets is compared with the cognitive evaluation respectively, the health education content with the same cognitive difficulty and cognitive evaluation is marked as a push content, and the health education content is pushed to the current patient according to the push content.

[0020] Further, the method for generating the multi-dimensional treatment scheme matrix comprises:

[0021] The treatment scheme corresponding to the historical patient in the patient cluster corresponding to the current patient is obtained and is marked as a candidate treatment scheme; the data in the background information which is not a numerical value is marked as text data, and each text data is converted into a corresponding text vector by using a pre-trained word embedding model; different internal motivations are set with different digital labels and are marked as motivation labels; the text data in the background information is replaced with the corresponding text vector, and the background information and the motivation label after replacement are used as analysis data; different treatment schemes are set with different digital labels and are marked as scheme labels; the scheme labels of the candidate treatment schemes are marked as candidate labels, each candidate label is used as a group of evaluation data with the analysis data, and the evaluation data and the candidate label correspond one by one; the evaluation data is input into the trained index evaluation model to predict the corresponding decision index; the decision index includes treatment effect and implementation difficulty, and the index evaluation model is a random forest model;

[0022] A preset weight set is provided, and the weight set includes a weight set corresponding to each data in the decision index; each data in the decision index is multiplied by a corresponding weight coefficient in the weight set, and then added in sequence to obtain a decision value; a preset decision threshold is provided, and the decision value of each candidate treatment scheme is compared with the decision threshold; the candidate treatment scheme whose decision value is greater than or equal to the decision threshold is marked as an alternative scheme, and the candidate treatment scheme whose decision value is less than the decision threshold is not marked; and a multi-dimensional treatment scheme matrix is generated according to all the alternative schemes.

[0023] Further, the method for screening the best treatment scheme comprises:

[0024] Treatment characteristic data corresponding to each alternative scheme is obtained, and the treatment characteristic data includes a treatment target and a risk degree; a pre-trained medical NLP model is used to perform entity recognition on each alternative scheme in sequence, and a treatment form corresponding to each alternative scheme is identified; the treatment form and the treatment target of each alternative scheme are taken as a group of scheme characteristic data; the cosine similarity between each data in each group of scheme characteristic data and the corresponding data in the patient's willingness information is calculated in sequence, and is marked as an acceptance probability; each data in each group of risk degrees is subtracted by the corresponding data in the risk tolerance to obtain a data difference value; the patient's willingness information includes a treatment target preference, a treatment form preference and a risk tolerance;

[0025] The acceptance probability corresponding to each group of alternatives and the data difference value are used as a group of evaluation characteristic data; a plurality of fuzzy sets are constructed for each data in the evaluation characteristic data; each group of evaluation characteristic data is converted into the membership degree of each corresponding fuzzy set through fuzzification technology; fuzzy rules are defined; each group of technical feature data after fuzzification is matched with the fuzzy rules, and fuzzy reasoning is performed using a fuzzy reasoning method to obtain a fuzzy reasoning result corresponding to each alternative, the fuzzy reasoning result being the membership degree of each acceptance level, the acceptance level including high acceptance, medium acceptance, low acceptance, and unacceptance; an acceptance interval is set, the range of the acceptance interval being [0, c], c being an integer greater than 1; the acceptance interval is evenly divided into four level intervals, each level interval corresponding to one level in the acceptance level; the maximum value of each level interval is added to the corresponding minimum value and then divided by 2 to obtain the interval mean value of each level interval; each membership degree corresponding to each alternative is multiplied by the corresponding interval mean value to obtain an acceptance score; a preset proportion set is included in the proportion coefficient corresponding to each data in the patient's willingness information; each acceptance score of each alternative is multiplied by the corresponding proportion coefficient in the proportion set and then added sequentially to obtain an acceptance total value of each alternative; each membership degree corresponding to each alternative is added sequentially to obtain a total membership degree; the acceptance total value of each alternative is divided by the corresponding total membership degree to obtain the acceptance level of each alternative, and the alternative with the highest acceptance level is taken as the best treatment scheme.

[0026] Further, the method for dynamically formulating an emotional support strategy comprises:

[0027] A pre-trained emotion recognition model is used to analyze real-time interaction data and identify the emotional response of the patient; a strategy set is set, the strategy set including a telephone interaction set and an on-site interaction set; one interaction method is randomly selected from the telephone interaction set and the on-site interaction set to construct a group of interaction combinations, a total of d groups of interaction combinations, the d groups of interaction combinations being different from each other, a sequentially increasing digital label is set for the d groups of interaction combinations and marked as a set label, the range of the set label being [1, d]; a set label is randomly selected as an initial iteration center, and the number of iterations is set to 0;

[0028] An iteration process is defined, the iteration process being: m candidate solutions are generated within the range of the set label, and the emotional relief effect corresponding to each candidate solution is calculated, 1 < m < d, the candidate solution corresponding to the set label; the candidate solution with the maximum emotional relief effect is marked as a temporary optimal solution, and the iteration center is moved to the temporary optimal solution;

[0029] An iterative process is performed, and each time the iterative process is performed, the iteration number is increased by one; a preset iteration threshold is set, and when the iteration number is greater than or equal to the iteration threshold, the iterative process is stopped, the set label corresponding to the iteration center is marked as the best label, and the interaction combination corresponding to the best label is taken as the emotional support strategy.

[0030] Further, the method of generating m candidate solutions comprises:

[0031] A preset selection interval is set, the iteration center is taken as the center of the selection interval, and m candidate solutions in the selection interval are obtained;

[0032] The method of calculating the emotional relief effect corresponding to the candidate solution comprises:

[0033] Different digital labels are set for different emotional types and are marked as type labels; different digital labels are set for different emotional intensities and are marked as intensity labels; the set label, the type label and the intensity label corresponding to the emotional response are sequentially input into the trained emotional analysis model, and the corresponding relief intensity is predicted; wherein the emotional analysis model is a random forest model, and the relief intensity is the emotional intensity of the patient after receiving the telephone interaction and the on-site interaction in the interaction combination; the emotional intensity is reduced by the relief intensity to obtain the emotional relief effect.

[0034] Further, the method of providing emotional support comprises:

[0035] The multi-level emotional support network comprises an automatic response layer, an artificial assistance layer and an artificial intervention layer; each type of emotion is converted into a corresponding type vector by using a pre-trained word embedding model; all type vectors are clustered by using a clustering algorithm to obtain three type clusters; the three type clusters correspond to positive emotions, negative emotions and neutral emotions respectively; the emotional type of the patient is analyzed, if the emotional type belongs to positive emotions, the automatic response layer is used to provide emotional support, if the emotional type is neutral emotions, the artificial assistance layer is used to provide emotional support, and if the emotional type is negative emotions, the artificial intervention layer is used to provide emotional support.

[0036] Further, the method of comprehensively evaluating the interaction effect comprises:

[0037] The patient health data and the historical health data are sequentially input into the trained effect evaluation model to predict the corresponding interaction effect; wherein the patient health data is the physiological index of the patient after receiving the chronic disease management, the historical health data is the physiological index of the patient before receiving the chronic disease management, the effect evaluation model is a deep learning model, and the interaction effect is the improvement effect of the chronic disease of the patient after receiving the chronic disease management;

[0038] The training process of the effect evaluation model comprises:

[0039] Pre-collect n groups of different health data, each group of health data including patient health data and historical health data, set a corresponding interaction effect for each group of health data, n is an integer greater than 1, convert the health data and the corresponding interaction effect into a corresponding group of feature vectors; each group of feature vectors is used as the input of the effect evaluation model, the effect evaluation model takes a group of predicted interaction effects corresponding to each group of health data as the output, and the actual interaction effect corresponding to each group of health data as the prediction target, the actual interaction effect is the interaction effect corresponding to the health data set in advance; the sum of the prediction errors of all health data is minimized as the training target; train the effect evaluation model until the sum of the prediction errors converges to stop training.

[0040] Further, a preset effect threshold is compared with the interaction effect; if the interaction effect is greater than or equal to the effect threshold, the health education content, the best treatment scheme and the emotional support strategy are not intelligently optimized; if the interaction effect is less than the effect threshold, the health education content, the best treatment scheme and the emotional support strategy are intelligently optimized in turn;

[0041] The method for intelligently optimizing the health education content includes:

[0042] The effect threshold is subtracted from the interaction effect to obtain an effect difference; different cognitive evaluations are set with different digital tags and marked as cognitive tags; the effect difference and the cognitive tags are input into the trained cognitive optimization model to predict the corresponding optimization tags, and the corresponding optimization evaluation is obtained according to the optimization tags, the optimization tags are the digital tags corresponding to the optimization evaluation, and the optimization evaluation is the optimized cognitive evaluation; the cognitive optimization model is a deep neural network model, and the training process of the cognitive optimization model is consistent with the training process of the effect evaluation model; the cognitive difficulty of each health education content in each group of content sets is compared with the optimization evaluation respectively, the health education content with the same cognitive difficulty and optimization evaluation is marked as optimization content, and the health education content is pushed to the current patient according to the optimization content.

[0043] Further, the method for intelligently optimizing the best treatment scheme includes:

[0044] A contribution set is preset, and the contribution set includes the contribution coefficients corresponding to the acceptance degree and the decision value; the acceptance degree and the decision value of each alternative scheme are multiplied by the corresponding contribution coefficients respectively and added in turn to obtain the scheme excellence degree, and the alternative scheme with the highest scheme excellence degree is taken as the best treatment scheme;

[0045] The method for intelligently optimizing the emotional support strategy includes:

[0046] The set label is input into the trained threshold prediction model in sequence with the type label and the intensity label corresponding to the emotional response, and the corresponding termination threshold is predicted; the threshold prediction model is a deep neural network model, and the training process of the threshold prediction model is consistent with the training process of the effect evaluation model; in the iteration process, the emotional relief effect corresponding to the current iteration center is subtracted from the emotional relief effect corresponding to the last iteration center to obtain an effect change; the effect change is compared with a preset change threshold; if the effect change is greater than or equal to the change threshold, a change instruction is generated; if the effect change is less than the change threshold, no change instruction is generated; the number of times of continuously generating the change instruction is counted and marked as the instruction number; when the instruction number is greater than or equal to the termination threshold, the iteration process is stopped; the set label corresponding to the iteration center is marked as the best label, and the interactive combination corresponding to the best label is taken as the emotional support strategy.

[0047] The technical effects and advantages of the chronic disease information management system based on the behavior interaction model of the present application are as follows:

[0048] By comprehensively obtaining the background information, intrinsic motivation and cognitive evaluation of the patient, a health education content library and a multi-dimensional treatment scheme matrix that meet the personalized needs of the patient are constructed, and the precise guidance and support for the behavior change process of the patient are realized; by using the method combining quantitative evaluation and fuzzy reasoning, the acceptance degree of different treatment schemes is comprehensively analyzed, the best treatment scheme that best meets the preferences and needs of the patient can be screened out, and the treatment compliance is improved; based on the emotional analysis and adaptive optimization of real-time interaction data, the targeted emotional support strategy is dynamically formulated, and the psychological health management of the patient is effectively enhanced; by using the deep learning technology to construct the effect evaluation model, the health improvement effect of the patient after receiving the chronic disease management can be accurately evaluated; by using the closed-loop feedback mechanism, the health education content, the treatment scheme and the emotional support strategy are sequentially optimized, the performance of the system is continuously improved, and the overall effect of the chronic disease management is enhanced; taking the behavior interaction as the core, the precise management of the chronic disease patient throughout the whole life cycle is realized, so that the health management effect of the patient is significantly improved, the risk of complications is reduced, and strong support is provided for the long-term control of chronic diseases. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 It is a schematic diagram of the chronic disease information management system based on the behavior interaction model of the present application in embodiment 1.

[0050] Figure 2 It is a flowchart of the chronic disease information management system based on the behavior interaction model of the present application in embodiment 1. DETAILED DESCRIPTION

[0051] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative effort are within the protection scope of the present application.

[0052] Embodiment one

[0053] Please refer to Figure 1 and Figure 2 , the chronic disease information management system based on the behavior interaction model comprises a data acquisition module, an information pushing module, a professional support module, a decision control module, an emotional support module, an interaction evaluation module and a feedback optimization module; each module is connected through wired and / or wireless mode to realize data transmission between the modules.

[0054] The data acquisition module is used for acquiring patient characteristic data, and the patient characteristic data comprises background information, intrinsic motivation and cognitive evaluation.

[0055] The background information comprises basic conditions, disease history and living habits; the basic conditions are, for example, age, gender, education level, occupation, marital status and the like of the patient; the disease history comprises past medical history and chronic disease condition; the past medical history is the disease that the patient has ever suffered from, and the past medical history comprises disease name, disease time, severity and the like, and the chronic disease condition is the chronic disease that the patient currently has, and the chronic disease condition comprises chronic disease name, diagnosis time, severity and the like; the living habits are, for example, eating habits, exercise habits, smoking and drinking conditions, work and rest rules and the like; the background information is acquired through an electronic medical record system in the hospital or a questionnaire filled by the patient about personal conditions, living habits and the like;

[0056] The intrinsic motivation is the internal driving force of the patient to spontaneously perform health management without external rewards or pressures, for example, high, medium and low; for example, Ms. Li hopes to reduce the disease burden and actively participates in the treatment by health management because she is worried about the aggravation of systemic lupus erythematosus, so the intrinsic motivation of Ms. Li is high; in contrast, Mr. Wang feels that the treatment has no obvious effect, and he lacks interest in health management and is unwilling to change his living habits, so the intrinsic motivation of Mr. Wang is low; the intrinsic motivation is acquired through a motivation evaluation scale filled by the patient;

[0057] Cognitive evaluation refers to the degree of cognition of patients on different health fields (such as health status, disease treatment, health behavior, etc.), for example, high, medium, low, etc.; cognitive evaluation directly affects whether patients think that health behavior change is feasible and effective, thereby determining whether to take action; cognitive evaluation includes health status cognitive degree (i.e. the degree of understanding of patients on their own health status, for example, the degree of cognition on the severity of the current chronic disease and the development of the disease), disease treatment cognitive degree (i.e. the degree of understanding of patients on the treatment method and drug used), health behavior cognitive degree (i.e. the degree of understanding of patients on the influence of health behavior change on health, such as changes in exercise, diet, and sleep), etc.; for example, Ms. Li understands the relationship between systemic lupus erythematosus and sarcopenia, and understands that exercise can effectively improve sarcopenia, which encourages her to exercise according to the doctor's advice, so Ms. Li's cognitive evaluation is high; in contrast, Mr. Wang has unclear understanding of sarcopenia and thinks that exercise will aggravate systemic lupus erythematosus, and hesitates to change exercise behavior, so Mr. Wang's cognitive evaluation is low; cognitive evaluation is obtained through relevant questionnaires filled out by patients or learning effect (such as the correct rate of answering questions on an online learning platform) in the process of health education.

[0058] The information pushing module is configured to construct a health education content library based on the background information, and push health education content by using a hierarchical pushing strategy in combination with cognitive evaluation.

[0059] The method for constructing the health education content library comprises:

[0060] The content resource library comprises health education content corresponding to different chronic diseases, and the health education content in the content resource library is obtained by technical personnel in the field through international guidelines (such as the American Rheumatism Association Guidelines, the Clinical Practice Guidelines of the International Society of Gerontology and Geriatric Medicine, etc.), evidence-based databases (such as UpToDate, PubMed, Cochrane Library, etc.), popular science materials (such as health manuals produced by hospitals, contents of authoritative medical platforms, etc.), etc.; each health education content in the content resource library corresponds to a group of content data, and the content data comprises a chronic disease name, a cognitive difficulty, and a health field, which are set by technical personnel in the field according to the health education content; the cognitive difficulty refers to the degree of difficulty for patients to understand and absorb the health education content, for example, high, medium, low, etc.; according to the chronic disease name in the background information, the corresponding health education content is screened out from all the health education content in the content resource library, and is marked as suitable content;

[0061] determining whether the health education content is pushed to the patient; if yes, obtaining historical content, clustering the historical content and suitable content by using a clustering algorithm (such as Gaussian mixture model, DBSCAN, hierarchical clustering, etc.) to obtain a content cluster, and a is an integer greater than 1; constructing a health education content library according to the suitable content in the content cluster corresponding to the historical content; if no, obtaining historical patients, and clustering the historical patients and the current patient by using a clustering algorithm according to the patient characteristic data of the historical patients and the current patient to obtain b patient clusters, and b is an integer greater than 1; constructing a health education content library according to the suitable content corresponding to the historical patients in the patient cluster corresponding to the current patient; wherein the historical content is the health education content pushed to the patient at the historical moment, the historical patient is the patient who has received chronic disease management at the historical moment, and the current patient is the patient who is currently undergoing chronic disease management; the historical content and the patient characteristic data of the historical patient are obtained through the patient information management in the hospital.

[0062] The method for pushing the health education content comprises:

[0063] In the health education content library, the health education content corresponding to the same health field is taken as a set of content collection, and the content collection corresponds to the health field one by one; the cognitive difficulty corresponding to each health education content in each set of content collection is compared with the cognitive evaluation respectively, and the health education content with the same cognitive difficulty and cognitive evaluation is marked as a push content, and the push content is pushed to the current patient for health education content; for example, for systemic lupus erythematosus patients, the health education content about sun protection, drug management, immune regulation, regular monitoring, fatigue management, etc. is pushed; for patients with muscle loss, the health education content about protein supplement, progressive resistance training, functional activity, nutritional balance, etc. is pushed.

[0064] The professional support module is used for fusing background information and intrinsic motivation, and generating a multi-dimensional treatment scheme matrix by using a multi-criteria decision analysis method.

[0065] The method for generating a multi-dimensional treatment scheme matrix comprises:

[0066] The treatment plan corresponding to the historical patient in the patient cluster corresponding to the current patient is obtained and marked as a candidate treatment plan, which is obtained through a patient information management system in the hospital; data in the background information that is not numerical is marked as text data, and a pre-trained word embedding model (such as a FastText model, an ELMo model, or a GloVe model) is used to convert each text data into a corresponding text vector. The text data may include gender, education level, and chronic disease name; different internal motivations are set with different digital labels and marked as motivation labels; the text data in the background information is replaced with the corresponding text vector, and the replaced background information and the motivation labels are used as analysis data; different treatment plans are set with different digital labels and marked as plan labels; the plan labels of the candidate treatment plans are marked as candidate labels, and each candidate label is used as a group of evaluation data with the analysis data, which corresponds to the candidate label one by one; the evaluation data is input into the trained index evaluation model to predict the corresponding decision index; the decision index includes treatment effect and implementation ease;

[0067] The treatment effect is the expected effect of the treatment plan on the improvement of the patient's chronic disease state, including clinical effects such as symptom relief, chronic disease index improvement, and complication prevention. The higher the treatment effect, the more effective the treatment plan is in controlling or improving the patient's chronic disease state in a medical sense. The implementation ease is the ease of implementing the treatment plan in the patient's daily life, which involves the objective convenience of the patient in implementing the treatment plan, including factors such as medication frequency, operation complexity, economic burden, and time cost. The higher the implementation ease, the easier it is to implement the treatment plan in the patient's daily life, reducing execution obstacles such as simplified medication plan, reduced side effects, reduced economic burden, and reduced hospital visit frequency. The index evaluation model is a random forest model, which is a prior art and the specific training process is not described in detail here.

[0068] A set of preset weights is provided, including the weight corresponding to each data in the decision index; each data in the decision index is multiplied by the corresponding weight coefficient in the weight set, and then added sequentially to obtain a decision value; a decision threshold is preset, and the decision value of each candidate treatment plan is compared with the decision threshold. The candidate treatment plan with a decision value greater than or equal to the decision threshold is marked as a selected plan, and the candidate treatment plan with a decision value less than the decision threshold is not marked. A multi-dimensional treatment plan matrix is generated based on all selected plans. The weight set and the decision threshold are both pre-set by a person skilled in the art according to actual conditions.

[0069] A decision control module is configured to obtain patient willingness information, quantitatively evaluate the acceptance degree of different treatment plans in the multi-dimensional treatment plan matrix, and select the best treatment plan from the multi-dimensional treatment plan matrix based on the evaluation result.

[0070] The method for screening the optimal treatment plan comprises:

[0071] The patient willingness information comprises treatment target preference, treatment form preference and risk tolerance; the treatment target preference is the final effect pursued by the patient for the treatment plan, or the subjective emphasis on the treatment result, reflecting the individual difference of the patient in the treatment expectation, such as prolonging life, relieving symptoms, reducing side effects, etc.; the treatment form preference is the preference of the patient for the execution mode of the treatment, such as the treatment approach (such as oral medicine or injection medicine), the treatment environment (such as home treatment or hospitalization treatment), etc.; the risk tolerance is the degree of acceptance of the patient for the risks (such as side effects, complications, etc.) that may occur in the treatment process, including severity and probability; the patient willingness information is obtained through the questionnaire filled out by the patient, the doctor-patient interview, etc.

[0072] The treatment characteristic data corresponding to each alternative plan is obtained, the treatment characteristic data comprising treatment target and risk degree, which is obtained through the electronic medical record system in the hospital; the treatment target is the expected effect or the action focus that can be achieved by the treatment plan in medicine, corresponding to the treatment target preference in the patient willingness information; the risk degree is the severity and probability of the risk that may be caused by the treatment plan, corresponding to the risk tolerance in the patient willingness information; the pre-trained medical NLP model (such as Med-BERT model, ClinicalBERT model, etc.) is used to perform entity recognition on each alternative plan in turn, and the treatment form corresponding to each alternative plan is identified, corresponding to the treatment form preference in the patient willingness information; the treatment form and the treatment target of each alternative plan are taken as a group of plan characteristic data, and the plan characteristic data corresponds to the alternative plan one by one; the cosine similarity between each data in each group of plan characteristic data and the corresponding data in the corresponding patient willingness information is calculated in turn, and is marked as acceptance probability; the calculation method of the cosine similarity is the prior art, and the specific calculation process will not be described in detail here; each data in each group of risk degree is subtracted from the corresponding data in the risk tolerance to obtain the data difference value;

[0073] The acceptance probability corresponding to each group of alternatives and the data difference value are taken as a group of evaluation characteristic data, and the evaluation characteristic data correspond one-to-one to the alternatives; a plurality of fuzzy sets are respectively constructed for each data in the evaluation characteristic data; for example, the fuzzy set corresponding to the acceptance probability is high acceptance probability, medium acceptance probability, and low acceptance probability, and the fuzzy set corresponding to the data difference value is positive difference value (i.e., the data difference value is greater than 0), zero difference value (i.e., the data difference value is equal to 0), and negative difference value (i.e., the data difference value is less than 0); each group of evaluation characteristic data is respectively converted into the membership degree of each corresponding fuzzy set by a fuzzification technique; the fuzzification technique is a process of converting an accurate numerical value into the membership degree corresponding to the fuzzy set, and the fuzzification technique is, for example, a triangular membership function, a trapezoidal membership function, etc.; for example, if the numerical value of the acceptance probability is high, the membership degree of high acceptance probability is 0.9, the membership degree of medium acceptance probability is 0.3, and the membership degree of low acceptance probability is 0;

[0074] A fuzzy rule is defined, and the fuzzy rule is defined according to expert knowledge or related literature; for example, if more than half of the acceptance probabilities in a group of evaluation characteristic data are high acceptance probabilities, and more than half of the data difference values are negative difference values, it is inferred that the membership degree of the acceptance degree of the corresponding alternative belongs to high acceptance degree; if more than half of the acceptance probabilities in a group of evaluation characteristic data are low acceptance probabilities, and more than half of the data difference values are positive difference values, it is inferred that the membership degree of the acceptance degree of the corresponding alternative belongs to unacceptability; each group of technical feature data after fuzzification is matched with the fuzzy rule, and a fuzzy reasoning method (such as a Mamdani fuzzy reasoning model, a Sugeno fuzzy reasoning model, etc.) is used for fuzzy reasoning to obtain the fuzzy reasoning result corresponding to each alternative, and the fuzzy reasoning result is the membership degree of each acceptance degree level, and the acceptance degree level includes high acceptance degree, medium acceptance degree, low acceptance degree, and unacceptability; for example, the membership degree of high acceptance degree is 0.6, the membership degree of medium acceptance degree is 0.8, the membership degree of low acceptance degree is 0.1, and the membership degree of unacceptability is 0;

[0075] An acceptance interval is set, the range of the acceptance interval is [0, c], and c is an integer greater than 1, and in the embodiment, c is preferably 100; the acceptance interval is evenly divided into four grade intervals, and each grade interval corresponds to a grade in the acceptance degree grade; the maximum value of each grade interval is added to the corresponding minimum value and then divided by 2 to obtain the interval mean of each grade interval; each membership corresponding to each candidate solution is multiplied by the corresponding interval mean to obtain an acceptance score; a preset proportion set is provided, the proportion set includes a proportion coefficient corresponding to each data in the patient's will information, and the proportion set is set in advance by a person skilled in the art according to the actual situation of the patient; each acceptance score of each candidate solution is multiplied by the corresponding proportion coefficient in the proportion set and then added in sequence to obtain an acceptance total value of each candidate solution; each membership corresponding to each candidate solution is added in sequence to obtain a total membership; the acceptance total value of each candidate solution is divided by the corresponding total membership to obtain an acceptance degree of each candidate solution, and the candidate solution with the highest acceptance degree is taken as the best treatment solution.

[0076] The emotional support module is configured to obtain real-time interaction data, analyze the real-time interaction data using an emotional analysis technique, dynamically formulate an emotional support strategy, and provide emotional support through a pre-established multi-level emotional support network.

[0077] The real-time interaction data is interactive content in the form of text generated in real time when the patient communicates with the chronic disease management system during the process of receiving chronic disease management, such as information input by the patient, such as “Will taking this medicine have side effects?” and “What should I do if the blood sugar does not decrease?” The real-time interaction data is obtained through a doctor-patient communication platform in the hospital.

[0078] The method for dynamically formulating the emotional support strategy includes:

[0079] The real-time interaction data is analyzed using a pre-trained emotional recognition model to identify the emotional response of the patient; the emotional response refers to the emotional experience of the patient in the face of the disease and the treatment process, which directly affects the treatment compliance of the patient and the effect of behavior change; the emotional response includes an emotional type and an emotional intensity, the emotional type is, for example, anxiety, worry, depression, anger, positivity, and the emotional intensity is, for example, mild, moderate, and severe; the emotional recognition model is fine-tuned based on a pre-trained language model (such as a ClinicalBERT model, a BioBERT model, a RoBERTa model, etc.) to be applied in a specific emotional analysis task; the emotional recognition model is a prior art, and the specific training process is not described in detail here;

[0080] A strategy set is set, the strategy set including a telephone interaction set and an on-site interaction set; the telephone interaction refers to communication through a telephone, has strong convenience and flexibility, and the telephone interaction set includes emotional counseling (i.e. a medical staff performs emotional counseling on a patient with emotional fluctuation through a telephone to help the patient relieve negative emotions), health education (i.e. the medical staff provides personalized health education content for the patient), psychological counseling (i.e. the medical staff provides psychological counseling for the patient with serious emotional problems through a telephone to reduce the psychological pressure of the patient) and the like; the on-site interaction refers to face-to-face communication between the patient and the medical staff, has high emotional support effect, and the on-site interaction set includes face-to-face consultation (i.e. the medical staff communicates with the patient face to face to answer the patient's questions), support group meeting (i.e. the medical staff organizes the patient to participate in a patient group meeting to share treatment experience with other patients, encourages and consoles each other), social support activity (i.e. the medical staff organizes leisure activities to enhance the social support of the patient, helps the patient release emotional pressure, such as group exercise, meditation course and the like) and the like; the strategy set is obtained by a person skilled in the art by referring to clinical practice, standard guidelines, user research and the like;

[0081] One interaction method is randomly selected from the telephone interaction set and the on-site interaction set, a set of interaction combinations is constructed, d sets of interaction combinations are constructed in total, the d sets of interaction combinations are all different, a sequentially increasing digital label is set for the d sets of interaction combinations, and the label is marked as a set label, the range of the set label is [1, d]; one set label is randomly selected as an initial iteration center, and the iteration number is set to 0;

[0082] An iteration process is defined, the iteration process being: m candidate solutions are generated within the range of the set label, and the emotional relief effect corresponding to each candidate solution is calculated, 1

[0083] The iteration process is executed, and the iteration number is increased by one each time the iteration process is executed; an iteration threshold is preset, and the iteration threshold is preset by a person skilled in the art according to actual conditions; when the iteration number is greater than or equal to the iteration threshold, the iteration process is stopped, the set label corresponding to the iteration center is marked as an optimal label, and the interaction combination corresponding to the optimal label is taken as an emotional support strategy.

[0084] The method for generating m candidate solutions includes:

[0085] The preset selection interval is set by a person skilled in the art according to actual conditions; the iteration center is taken as the center of the selection interval, and m candidate solutions in the selection interval are obtained; for example, the selection interval is [-2, 2], and the set label corresponding to the iteration center is 5, so 5 is taken as the center of the selection interval, that is, the set labels included in the selection interval are 3, 4, 5, 6 and 7.

[0086] The method for calculating the emotional relief effect corresponding to the candidate solution comprises:

[0087] Different digital labels are set for different emotional types and are marked as type labels; different digital labels are set for different emotional intensities and are marked as intensity labels; the set label, the type label and the intensity label corresponding to the emotional response are sequentially input into the trained emotional analysis model to predict the corresponding relief intensity; wherein the emotional analysis model is a random forest model, and the relief intensity is the emotional intensity of the patient after receiving the telephone interaction and the on-site interaction in the interactive combination; the emotional intensity is subtracted by the relief intensity to obtain the emotional relief effect.

[0088] The method for providing emotional support comprises:

[0089] The multi-level emotional support network comprises an automatic response layer, an artificial assistance layer and an artificial intervention layer; each emotional type is converted into a corresponding type vector by using a pre-trained word embedding model; three type clusters are obtained by clustering all type vectors by using a clustering algorithm; the three type clusters correspond to positive emotions (such as positive, confident, etc.), negative emotions (such as anxiety, worry, etc.) and neutral emotions (such as calm, confusion, etc.) respectively; the emotional type of the patient is analyzed, if the emotional type belongs to positive emotions, the automatic response layer is used to provide emotional support, if the emotional type is neutral emotions, the artificial assistance layer is used to provide emotional support, and if the emotional type is negative emotions, the artificial intervention layer is used to provide emotional support;

[0090] The automatic response layer automatically replies to the real-time interaction data of the patient by using a robot (such as a chat robot, a virtual health assistant, etc.), a preset dialogue template, natural language generation (NLG) technology and other methods; the purpose is to consolidate and continue the positive emotional state of the patient, so there is no need for artificial complex intervention; the artificial assistance layer automatically replies to the real-time interaction data of the patient by combining the preliminary response of the robot and manual review; that is, the medical staff judges whether the reply of the robot needs to be modified and then sent to the patient; at this time, the patient does not show strong emotions, but may hide needs or potential risks, which need to be moderately concerned; the artificial intervention layer replies to the real-time interaction data of the patient in real time by the medical staff; at this time, the patient has obvious emotional distress or psychological pressure, which needs timely and in-depth artificial intervention.

[0091] It should be understood that the dynamically formulated emotional support strategy is a plan for future interaction behavior, which is a delayed behavior, that is, the chronic disease management system does not respond immediately, and is used to ensure that appropriate support is provided in future interactions; and the emotional support provided through the pre-established multi-level emotional support network is a timely feedback to the current interaction, which is a real-time behavior, that is, the patient immediately responds and replies to the patient in real time when the patient communicates with the chronic disease management system.

[0092] An interaction evaluation module is configured to collect patient health data and comprehensively evaluate the interaction effect by using deep learning technology.

[0093] The method for comprehensively evaluating the interaction effect comprises:

[0094] The patient health data and the historical health data are sequentially input into the trained effect evaluation model to predict the corresponding interaction effect; wherein the patient health data is a physiological index of the patient after receiving the chronic disease management, such as blood pressure, blood sugar, weight, etc.; the historical health data is a physiological index of the patient before receiving the chronic disease management; the effect evaluation model is a deep learning model; and the interaction effect is an improvement effect of the chronic disease after the patient receives the chronic disease management.

[0095] The training process of the effect evaluation model comprises:

[0096] n sets of different health data are collected in advance, each set of health data includes patient health data and historical health data, the corresponding interaction effect is set for the n sets of health data, n is an integer greater than 1, and the health data and the corresponding interaction effect are converted into a corresponding set of feature vectors; the interaction effect corresponding to the health data is collected by the person skilled in the art in the process of historical chronic disease management, and the interaction effect corresponding to each set of health data is evaluated by analyzing each set of health data in sequence according to the actual situation, and the corresponding interaction effect is set for the n sets of different health data in sequence;

[0097] Each set of feature vectors is taken as the input of the effect evaluation model, the effect evaluation model takes a set of predicted interaction effects corresponding to each set of health data as the output, takes the actual interaction effect corresponding to each set of health data as the prediction target, and the actual interaction effect is the interaction effect corresponding to the health data set in advance; the sum of the prediction errors of all health data is minimized as the training target; wherein the calculation formula of the prediction error is η w =(θ w -ε w ) 2 , wherein η w is the prediction error, w is the group number of the feature vector corresponding to the health data, θ w is the predicted interaction effect corresponding to the wth set of health data, and ε wThe actual interaction effect corresponding to the wth group of health data; training the effect evaluation model until the sum of prediction errors converges to stop training.

[0098] The feedback optimization module is configured to intelligently optimize the health education content, the optimal treatment scheme, and the emotional support strategy in sequence through a closed-loop feedback mechanism according to the interaction effect.

[0099] The preset effect threshold is set by a person skilled in the art according to actual conditions; the interaction effect is compared with the effect threshold; if the interaction effect is greater than or equal to the effect threshold, the health education content, the optimal treatment scheme, and the emotional support strategy are not intelligently optimized; and if the interaction effect is less than the effect threshold, the health education content, the optimal treatment scheme, and the emotional support strategy are intelligently optimized in sequence.

[0100] The method for intelligently optimizing the health education content comprises:

[0101] The effect threshold is subtracted from the interaction effect to obtain an effect difference; different cognitive evaluations are set with different digital tags and marked as cognitive tags; the effect difference and the cognitive tags are input into the trained cognitive optimization model to predict corresponding optimization tags, and the corresponding optimization evaluation is obtained according to the optimization tags; the optimization tag is a digital tag corresponding to the optimization evaluation, and the optimization evaluation is the optimized cognitive evaluation; the cognitive optimization model is a deep neural network model, and the training process of the cognitive optimization model is consistent with the training process of the effect evaluation model; the cognitive difficulty corresponding to each health education content in each content set is compared with the optimization evaluation, the health education content with the same cognitive difficulty and optimization evaluation is marked as optimization content, and the health education content is pushed to the current patient according to the optimization content.

[0102] The method for intelligently optimizing the optimal treatment scheme comprises:

[0103] The preset contribution set includes the contribution coefficients corresponding to the acceptance degree and the decision value; the contribution set is set by a person skilled in the art according to actual conditions; the acceptance degree and the decision value of each alternative scheme are multiplied by the corresponding contribution coefficients and added in sequence to obtain the scheme excellence degree, and the alternative scheme with the highest scheme excellence degree is taken as the optimal treatment scheme.

[0104] The method for intelligently optimizing the emotional support strategy comprises:

[0105] The set label is sequentially input into the trained threshold prediction model corresponding to the type label and the intensity label of the emotional response, and the corresponding termination threshold is predicted; the threshold prediction model is a deep neural network model, and the training process of the threshold prediction model is consistent with the training process of the effect evaluation model; in the iteration process, the emotional relief effect corresponding to the current iteration center is subtracted from the emotional relief effect corresponding to the last iteration center to obtain an effect change; the effect change is compared with a preset change threshold, and the change threshold is preset by a person skilled in the art according to the actual situation; if the effect change is greater than or equal to the change threshold, a change instruction is generated; if the effect change is less than the change threshold, no change instruction is generated; the number of consecutive change instructions is counted and marked as the instruction number; when the instruction number is greater than or equal to the termination threshold, the iteration process is stopped; the set label corresponding to the iteration center is marked as the best label, and the interactive combination corresponding to the best label is used as the emotional support strategy.

[0106] The embodiment comprehensively obtains the background information, intrinsic motivation and cognitive evaluation of the patient, constructs a health education content library and a multi-dimensional treatment scheme matrix that meet the personalized needs of the patient, realizes precise guidance and support for the behavior change process of the patient, adopts a method combining quantitative evaluation and fuzzy reasoning to comprehensively analyze the acceptance degree of different treatment schemes, can screen out the best treatment scheme that best meets the preferences and needs of the patient, and improves treatment compliance; based on emotional analysis and adaptive optimization of real-time interaction data, a targeted emotional support strategy is dynamically formulated, which effectively enhances the psychological health management of the patient; an effect evaluation model is constructed by using deep learning technology, which can accurately evaluate the health improvement effect of the patient after receiving chronic disease management; through a closed-loop feedback mechanism, the health education content, the treatment scheme and the emotional support strategy are sequentially optimized, the performance of the system is continuously improved, and the overall effect of chronic disease management is enhanced; taking behavior interaction as the core, precise management of the whole life cycle of the chronic disease patient is realized, thereby significantly improving the health management effect of the patient and reducing the risk of complications, and providing strong support for long-term control of chronic diseases.

[0107] Embodiment Two

[0108] The application also provides an electronic device. The electronic device can include one or more processors and one or more memories. The memory has computer readable code stored therein, which, when executed by the one or more processors, can perform the chronic disease information management system based on the behavior interaction model as described above.

[0109] The method or system according to the embodiments of the present application can also be implemented by means of the architecture of the electronic device shown in the present application. The electronic device can include a bus, one or more CPUs, ROM, RAM, a communication port connected to a network, input / output, a hard disk, etc. The storage device in the electronic device, such as ROM or a hard disk, can store the chronic disease information management system based on the behavioral interaction model provided by the present application. Further, the electronic device can also include a user interface. Of course, the architecture shown in the present application is only exemplary, and when implementing different devices, one or more components of the electronic device shown in the present application can be omitted according to actual needs.

[0110] Embodiment Three

[0111] One embodiment of the present application discloses a computer readable storage medium. The computer readable storage medium stores computer readable instructions. When the computer readable instructions are run by a processor, the chronic disease information management system based on the behavioral interaction model according to the embodiments of the present application described with reference to the above figures can be executed. The storage medium includes but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and cache memory, etc. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0112] In addition, according to the embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, the present application provides a non-transitory machine readable storage medium storing machine readable instructions executable by a processor to perform instructions corresponding to the method steps provided by the present application, for example: the chronic disease information management system based on the behavioral interaction model. When this computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0113] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent replacements to some technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0114] It should be noted that, in the present document, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises... a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0115] In the description of the application, it should be understood that the terms "first", "second", etc. are used only to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0116] In the description of the application, unless otherwise specified, "a plurality" means two or more.

[0117] In the description of the application, "several" means one or more, and "a large number" means two or more.

[0118] In the description of the present specification, the description referring to the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0119] For the formula of the present specification, the value is calculated by de-dimensioning, the formula is obtained by software simulation of a large amount of data to obtain a formula of the most recent real situation, and the preset parameters and threshold values in the formula are set by a person skilled in the art according to the actual situation.

[0120] Although the embodiments of the present application have been shown and described, those skilled in the art can understand that various changes, modifications, replacements and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the claims and their equivalents.

Claims

1. A chronic disease information management system based on a behavioral interaction model, characterized in that, The application comprises the following steps: a data acquisition module is used to acquire patient characteristic data, including background information, intrinsic motivation and cognitive evaluation; an information push module is used to construct a health education content library based on the background information, and to push health education content by adopting a hierarchical push strategy combined with the cognitive evaluation; a professional support module is used to fuse the background information and the intrinsic motivation, adopt a multi-criteria decision analysis method, and generate a multi-dimensional treatment scheme matrix; a decision control module is used to acquire patient willingness information, quantitatively evaluate the acceptance degree of different treatment schemes in the multi-dimensional treatment scheme matrix, and select the best treatment scheme from the multi-dimensional treatment scheme matrix according to the evaluation result; wherein the method for selecting the best treatment scheme comprises the following steps: acquiring treatment characteristic data corresponding to each alternative scheme, the treatment characteristic data including a treatment target and a risk degree; adopting a pre-trained medical NLP model to sequentially perform entity recognition on each alternative scheme to identify the treatment form corresponding to each alternative scheme; taking the treatment form and the treatment target of each alternative scheme as a group of scheme characteristic data; sequentially calculating the cosine similarity between each data in each group of scheme characteristic data and the corresponding data in the corresponding patient willingness information, and marking the acceptance probability; sequentially subtracting the corresponding data in the risk tolerance from each data in each group of risk degree to obtain a data difference; the patient willingness information includes a treatment target preference, a treatment form preference and a risk tolerance; an emotional support module is used to acquire real-time interaction data, analyze the real-time interaction data by adopting an emotional analysis technology, dynamically develop an emotional support strategy, and provide emotional support through a pre-established multi-level emotional support network; wherein the method for dynamically developing the emotional support strategy comprises the following steps: adopting a pre-trained emotion recognition model to analyze the real-time interaction data and identify the emotional response of the patient; setting a strategy set, the strategy set including a telephone interaction set and an on-site interaction set; defining an iteration process; executing the iteration process; predefining an iteration threshold, and stopping the execution of the iteration process when the iteration number is greater than or equal to the iteration threshold, marking the set label corresponding to the iteration center as the best label, and taking the interaction combination corresponding to the best label as the emotional support strategy; the method for providing emotional support comprises the following steps: the multi-level emotional support network includes an automatic response layer, an artificial assistance layer and an artificial intervention layer; each type of emotion is converted into a corresponding type vector by adopting a pre-trained word embedding model; all type vectors are clustered by adopting a clustering algorithm to obtain three type clusters; the three type clusters correspond to positive emotion, negative emotion and neutral emotion respectively; the emotional type of the patient is analyzed, if the emotional type belongs to positive emotion, the automatic response layer is adopted to provide emotional support, if the emotional type is neutral emotion, the artificial assistance layer is adopted to provide emotional support, and if the emotional type is negative emotion, the artificial intervention layer is adopted to provide emotional support; an interaction evaluation module is used to collect patient health data and comprehensively evaluate the interaction effect by utilizing deep learning technology. The feedback optimization module is configured to intelligently optimize the health education content, the optimal treatment scheme and the emotional support strategy in sequence through a closed-loop feedback mechanism according to the interactive effect.

2. The behavioral interaction model based chronic disease information management system according to claim 1, wherein, The background information includes basic information, disease history and living habits; the disease history includes past medical history and chronic disease conditions; The chronic disease conditions include disease name, disease duration and severity; The method for constructing the health education content library comprises: The content resource library comprises health education content corresponding to different chronic diseases; each health education content in the content resource library corresponds to a group of content data, the content data comprising a chronic disease name, a cognitive difficulty and a health field; the corresponding health education content is selected from all the health education content in the content resource library according to the chronic disease name in the background information, and is marked as suitable content; it is determined whether the health education content has been pushed to the patient; if yes, historical content is obtained, a clustering algorithm is used to cluster the historical content and the suitable content, a content cluster a is obtained, a is an integer greater than 1; the health education content library is constructed according to the suitable content of the historical patients in the corresponding content cluster of the historical content; if no, historical patients are obtained, and a clustering algorithm is used to cluster the historical patients and the current patient according to the patient feature data of the historical patients and the current patient, a patient cluster b is obtained, b is an integer greater than 1; the health education content library is constructed according to the suitable content of the historical patients in the corresponding patient cluster of the current patient; wherein the historical content is the health education content pushed to the patient at a historical moment, the historical patient is a patient who has received chronic disease management at a historical moment, and the current patient is a patient who is currently undergoing chronic disease management; The method for pushing the health education content comprises: The health education content corresponding to the same health field in the health education content library is taken as a group of content sets; the cognitive difficulty of each health education content in each group of content sets is compared with the cognitive evaluation respectively, the health education content with the same cognitive difficulty and cognitive evaluation is marked as a push content, and the push content is pushed to the current patient for health education content.

3. The behavioral interaction model based chronic disease information management system according to claim 2, wherein, The method for generating the multi-dimensional treatment scheme matrix comprises: The treatment scheme corresponding to the historical patient in the corresponding patient cluster of the current patient is obtained and marked as a candidate treatment scheme; the data in the background information which is not a numerical value is marked as text data, and a pre-trained word embedding model is used to convert each text data into a corresponding text vector; different internal motivations are set with different digital labels and marked as motivation labels; the text data in the background information is replaced with the corresponding text vector, and the replaced background information and the motivation label are used as analysis data; different treatment schemes are set with different digital labels and marked as scheme labels; the scheme labels of the candidate treatment schemes are marked as candidate labels, each candidate label is taken as a group of evaluation data with the analysis data, and the evaluation data and the candidate labels correspond to each other; the evaluation data is input into the trained index evaluation model to predict the corresponding decision index; the decision index comprises a treatment effect and an implementation ease, and the index evaluation model is a random forest model; A preset weight set includes a weight set corresponding to each data in the decision index; each data in the decision index is multiplied by the corresponding weight coefficient in the weight set, and the results are sequentially added to obtain a decision value; a preset decision threshold is compared with the decision value of each candidate treatment scheme; the candidate treatment scheme with a decision value greater than or equal to the decision threshold is marked as a selected scheme, and the candidate treatment scheme with a decision value less than the decision threshold is not marked; and a multi-dimensional treatment scheme matrix is generated based on all the selected schemes.

4. The behavioral interaction model based chronic disease information management system according to claim 3, wherein, The method for screening the best treatment scheme further includes: Each group of selected schemes and the data difference value are used as a group of evaluation characteristic data; a plurality of fuzzy sets are constructed for each data in the evaluation characteristic data; each group of evaluation characteristic data is converted into the membership of each fuzzy set by a fuzzy technology; a fuzzy rule is defined; each group of technical feature data after the fuzzy conversion is matched with the fuzzy rule, and a fuzzy inference is performed by using a fuzzy inference method to obtain a fuzzy inference result corresponding to each selected scheme, the fuzzy inference result being the membership of each acceptance level, the acceptance level including a high acceptance level, a medium acceptance level, a low acceptance level and an unacceptable level; an acceptance interval is set, the range of the acceptance interval being [0, c], c being an integer greater than 1; the acceptance interval is evenly divided into four level intervals, each level interval corresponding to one level in the acceptance level; the maximum value of each level interval is added to the minimum value of the corresponding level interval and then divided by 2 to obtain the interval mean of each level interval; each membership corresponding to each selected scheme is multiplied by the interval mean of the corresponding level interval to obtain an acceptance score; a preset proportion set includes a proportion coefficient corresponding to each data in the patient willingness information; each acceptance score of each selected scheme is multiplied by the proportion coefficient in the proportion set, and the results are sequentially added to obtain an acceptance total value of each selected scheme; each membership corresponding to each selected scheme is sequentially added to obtain a total membership; the acceptance total value of each selected scheme is divided by the total membership of the corresponding selected scheme to obtain the acceptance level of each selected scheme, and the selected scheme with the highest acceptance level is taken as the best treatment scheme.

5. The behavioral interaction model based chronic disease information management system as claimed in claim 1, wherein, The method for dynamically formulating the emotional support strategy further includes: One interaction method is randomly selected from each of the telephone interaction set and the on-site interaction set to construct one interaction combination, d interaction combinations are constructed, the d interaction combinations are all different, a sequentially increasing digital label is set for the d interaction combinations, and the label is marked as a set label, the range of the set label being [1, d]; one set label is randomly selected as an initial iteration center, and the iteration number is set to 0; An iteration process is defined, the iteration process being: m candidate solutions are generated within the range of the set label, and the emotional relief effect corresponding to each candidate solution is calculated, 1 < m < d, the candidate solution corresponding to the set label; the candidate solution with the maximum emotional relief effect is marked as a temporary optimal solution, and the iteration center is moved to the temporary optimal solution; The iteration process is executed, and the iteration number is increased by one each time the iteration process is executed.

6. The behavioral interaction model based chronic disease information management system as claimed in claim 5, wherein, The method for generating m candidate solutions comprises: a preset selection interval, taking the iteration center as the center of the selection interval, and obtaining m candidate solutions in the selection interval; The method for calculating the emotional relief effect corresponding to the candidate solution comprises: different emotional types are set with different digital labels and marked as type labels; different emotional intensities are set with different digital labels and marked as intensity labels; the set label, the type label and the intensity label corresponding to the emotional response are sequentially input into the trained emotional analysis model to predict the corresponding relief intensity; wherein the emotional analysis model is a random forest model, and the relief intensity is the emotional intensity after the patient receives the telephone interaction and the on-site interaction in the interactive combination; the emotional intensity is subtracted by the relief intensity to obtain the emotional relief effect.

7. The behavioral interaction model based chronic disease information management system as claimed in claim 1, wherein, The method for comprehensively evaluating the interaction effect comprises: the patient health data and the historical health data are sequentially input into the trained effect evaluation model to predict the corresponding interaction effect; wherein the patient health data is the physiological index of the patient after receiving the chronic disease management, the historical health data is the physiological index of the patient before receiving the chronic disease management, the effect evaluation model is a deep learning model, and the interaction effect is the improvement effect of the chronic disease after the patient receives the chronic disease management; The training process of the effect evaluation model comprises: n groups of different health data are collected in advance, each group of health data includes patient health data and historical health data, the n groups of health data are set with corresponding interaction effects, n is an integer greater than 1, the health data and the corresponding interaction effect are converted into a corresponding group of feature vectors; each group of feature vectors is taken as the input of the effect evaluation model, the effect evaluation model takes a group of predicted interaction effects corresponding to each group of health data as the output, and takes the actual interaction effect corresponding to each group of health data as the prediction target, the actual interaction effect is the interaction effect set in advance corresponding to the health data; the sum of the prediction errors of all health data is minimized as the training target; the effect evaluation model is trained until the sum of the prediction errors reaches convergence, and the training is stopped.

8. The behavioral interaction model based chronic disease information management system as claimed in claim 1, wherein, a preset effect threshold is compared with the interaction effect; if the interaction effect is greater than or equal to the effect threshold, the health education content, the best treatment scheme and the emotional support strategy are not intelligently optimized; if the interaction effect is less than the effect threshold, the health education content, the best treatment scheme and the emotional support strategy are intelligently optimized in sequence; The method for intelligently optimizing the health education content comprises: the effect threshold is subtracted by the interaction effect to obtain an effect difference; different cognitive evaluations are set with different digital labels and marked as cognitive labels; The effect difference and the cognitive label are input into the trained cognitive optimization model, a corresponding optimization label is predicted, and a corresponding optimization evaluation is obtained according to the optimization label; the optimization label is a digital label corresponding to the optimization evaluation, and the optimization evaluation is an optimized cognitive evaluation; the cognitive optimization model is a deep neural network model, and the training process of the cognitive optimization model is consistent with the training process of the effect evaluation model; the cognitive difficulty corresponding to each health education content in each set of content groups is compared with the optimization evaluation, and the health education content with the same cognitive difficulty and optimization evaluation is marked as an optimized content; and the optimized content is pushed to the current patient for health education content. 9.The chronic disease information management system based on a behavioral interaction model according to claim 4, wherein, The method for intelligently optimizing the best treatment plan comprises: A contribution set is preset, and the contribution set includes a contribution coefficient corresponding to the acceptance degree and the decision value; the acceptance degree and the decision value of each candidate plan are multiplied by the corresponding contribution coefficient, and then added in sequence to obtain a plan excellence degree; and the candidate plan with the highest plan excellence degree is taken as the best treatment plan. 10.The chronic disease information management system based on a behavioral interaction model according to claim 5, wherein, The method for intelligently optimizing the emotional support strategy comprises: The set label, the type label and the intensity label corresponding to the emotional response are input into the trained threshold prediction model in sequence to predict a corresponding termination threshold; the threshold prediction model is a deep neural network model, and the training process of the threshold prediction model is consistent with the training process of the effect evaluation model; in the iteration process, the emotional relief effect corresponding to the current iteration center is subtracted from the emotional relief effect corresponding to the last iteration center to obtain an effect change; the effect change is compared with a preset change threshold; if the effect change is greater than or equal to the change threshold, a change instruction is generated; if the effect change is less than the change threshold, no change instruction is generated; the number of continuously generated change instructions is counted and marked as an instruction number; when the instruction number is greater than or equal to the termination threshold, the iteration process is stopped; the set label corresponding to the iteration center is marked as a best label, and the interaction combination corresponding to the best label is taken as the emotional support strategy.

Images