A processing method and device for predicting a chronic disease hierarchical coping strategy

Abstract

This invention relates to a method and apparatus for predicting tiered coping strategies for chronic diseases. The method includes: constructing a tiered chronic disease prediction model for predicting the three-stage progression of N types of chronic diseases based on individual information; constructing a disease strategy atlas and a strategy knowledge base based on prior knowledge of first-stage blocking, second-stage intervention, and third-stage control strategies for various chronic diseases; constructing a first dataset through data collection; training the tiered chronic disease prediction model based on the first dataset; after model training, inputting an individual information set X from any user into the tiered chronic disease prediction model to obtain a corresponding prediction vector Y; and performing strategy identifier retrieval based on the prediction vector Y, the disease strategy atlas, and the strategy knowledge base, and synthesizing a corresponding coping strategy report based on the retrieval results to provide feedback to the current user. This invention enables parallel evaluation of multiple diseases and multiple stages.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a processing method and apparatus for predicting tiered coping strategies for chronic diseases. Background Technology

[0002] The progression of chronic diseases follows a gradual process: accumulation of risk factors → subclinical lesions → clinical disease → complications. The rate of disease progression is regulated by multiple factors, including environmental, behavioral, and psychological factors. By implementing primary prevention to block risk, secondary prevention for early intervention, and tertiary prevention to control symptoms, combined with life-cycle management, disease progression can be effectively slowed, and disability and mortality rates reduced. For individuals, regular screening, a healthy lifestyle, and standardized treatment are key to halting the progression of chronic diseases. However, current chronic disease health management technologies still face challenges in implementing these concepts, making it difficult to achieve truly individualized, precise, and full-cycle intervention.

[0003] First, existing technologies lack systematic assessment tools for coexisting multiple diseases. Existing assessment models often focus on a single disease (e.g., predicting only the 10-year risk of coronary heart disease), failing to simultaneously cover multiple disease states. However, individuals often suffer from multiple chronic diseases simultaneously, such as hypertension, diabetes, and osteoarthritis, and each disease is at different prevention stages (e.g., hypertension has reached stage 2 while diabetes remains at stage 1). In such cases, existing solutions cannot conduct simultaneous, staged predictions, leading to intervention strategies that are inadequate in their scope.

[0004] Secondly, the dimensions of health information are limited. Most existing technological solutions only collect basic demographic indicators and a few physical examination data, failing to adopt features with important predictive value such as lifestyle habits (eating frequency, sleep quality, mental stress) and face-to-face observation information (tongue appearance, complexion, gait). This results in a large amount of health information being idle and unable to be effectively utilized.

[0005] Third, there is a disconnect between predicted outcomes and intervention strategies. Current technologies typically stop at outputting disease risk scores, failing to automatically translate risk stratification into actionable primary, secondary, and tertiary prevention strategies. Strategy development still relies on manual review of guidelines or expert experience, which is inefficient and lacks personalization, making it impossible to dynamically match the most appropriate blocking, intervention, or control measures based on an individual's current specific disease stage. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by providing a method, apparatus, electronic device, and computer-readable storage medium for predicting tiered coping strategies for chronic diseases. This invention constructs a tiered chronic disease prediction model for predicting the three-stage progression of N types of chronic diseases based on individual information. It constructs a disease strategy map and strategy knowledge base based on prior knowledge of first-stage blocking, second-stage intervention, and third-stage control strategies for various chronic diseases. A first dataset is constructed through data collection, and the tiered chronic disease prediction model is trained based on this dataset. After model training, an individual information set X input by any user is fed into the tiered chronic disease prediction model for prediction, resulting in a corresponding prediction vector Y. Based on the prediction vector Y, the disease strategy map, and the strategy knowledge base, strategy identifiers are retrieved, and a corresponding coping strategy report is synthesized based on the retrieval results and fed back to the current user. This invention can enrich the feature dimensions of health information, improve the utilization rate of health information, enable parallel assessment of multiple diseases and multiple stages, and improve the personalization and processing efficiency of strategy analysis.

[0007] To achieve the above objectives, a first aspect of the present invention provides a method for predicting tiered coping strategies for chronic diseases, the method comprising: A chronic disease hierarchical prediction model is constructed to predict the third-order progression state of N types of chronic diseases based on individual information. The total number of chronic diseases N is a preset positive integer, and the N types of chronic diseases consist of N preset chronic disease categories. The chronic disease hierarchical prediction model is used to predict the third-order progression state of the N types of chronic diseases based on the individual information set X input to the model and outputs the corresponding prediction vector Y. The prediction vector Y includes N disease category vectors y. i The disease vector y i Each of the chronic diseases corresponds one-to-one, 1 ≤ index i ≤ N; the disease vector y i Includes 4 predicted probabilities y i,j , 1 ≤ index j ≤ 4; 4 predicted probabilities y i,j Each stage corresponds to one of four tiers: health stage, first-level stage, second-level stage, and third-level stage. Based on prior knowledge of first-order blocking, second-order intervention, and third-order control strategies for various chronic diseases, a disease strategy map and strategy knowledge base are constructed. The first dataset was constructed through data collection; The chronic disease classification prediction model is trained based on the first dataset; After the model training is completed, the individual information set X input by any user is sent to the chronic disease hierarchical prediction model to obtain the corresponding prediction vector Y; and based on the prediction vector Y, the disease strategy map and the strategy knowledge base, strategy identifier retrieval is performed, and based on the retrieval results, a strategy report is synthesized to obtain the corresponding response strategy report and fed back to the current user.

[0008] Preferably, the individual information set X includes basic information, measurement information, allergy information, surgical information, female reproductive health information, medical history information, lifestyle information, and observation information; The basic information includes age, birth weight, and gender type; The measurement information includes height, weight, BMI, waist circumference, hip circumference, neck circumference, calf circumference, grip strength, blood pressure, heart rate, blood oxygen saturation, body temperature, body fat percentage, skeletal muscle mass, and body water percentage. The allergy information includes four types of allergy markers: no allergy marker, drug allergy marker, food allergy marker, and other allergy markers. Each of the four markers includes two states: yes and no. When the no allergy marker is yes, the other three markers are no. When the no allergy marker is no, some or all of the other three markers are yes. The surgical information includes a surgical marker and the interval between the most recent surgeries. The surgical marker has two states: yes and no. When the surgical marker is no, the most recent surgery time is empty. When the surgical marker is yes, the interval between the most recent surgery time is the interval between the most recent surgery time and the current time. The female reproductive health information includes a menstrual regularity marker, menstrual cycle marker, pregnancy marker, childbirth marker, and menopause marker; the menstrual regularity marker, the pregnancy marker, the childbirth marker, the childbirth marker, and the menopause marker all have two states: yes and no; when the menstrual regularity marker is no, the menstrual cycle is empty; when the gender type is male, all markers in the female reproductive health information are no. The medical history information includes a history of metabolic diseases, cardiovascular and cerebrovascular diseases, respiratory diseases, digestive diseases, urinary diseases, endocrine diseases, reproductive diseases, musculoskeletal diseases, nervous diseases, immune system diseases, circulatory system diseases, and mental and psychological diseases. Each type of medical history includes one or more corresponding disease information. Each disease information consists of a corresponding disease name, disease marker, frequency type of onset, and disease control status. The disease marker includes two states: yes and no. The frequency type of onset includes asymptomatic, occasional, frequent, and persistent. The disease control status includes meeting the target, partially meeting the target, not meeting the target, and not being monitored. The lifestyle information includes smoking habits, drinking habits, air conditioning usage habits, exercise habits, sleep quality, mental stress, drinking habits, dietary habits, three meals a day, monthly average food intake, and daily average food intake. The smoking habit information includes a smoking marker and smoking duration; the smoking marker includes two states: yes and no; when the smoking marker is no, the smoking duration is empty. The drinking habit information includes a drinking flag and the age of drinking; the drinking flag includes two states: yes and no; when the drinking flag is no, the age of drinking is empty. The air conditioner usage habit information includes air conditioner usage markers, air conditioner usage time period types, and air conditioner usage years; the air conditioner usage markers include two states: yes and no; the air conditioner usage time period types include year-round, summer, and winter; The exercise habit information includes exercise habit markers, exercise cycle, exercise type, and duration of a single exercise session; the exercise habit markers include two states: yes and no; when the exercise habit marker is no, the exercise cycle, the exercise type, and the duration of a single exercise session are all empty; The sleep quality information includes sleep quality type and daily sleep duration; the sleep quality type includes very good, good, average, poor, and very poor. The mental stress information includes a mental stress marker and a stressor type; the mental stress marker includes two states: yes and no; the stressor type includes work, family, interpersonal relationships, social environment, and other; when the mental stress marker is no, the stressor type is empty; The drinking water habit information includes water source type, drinking water temperature, and daily water consumption; the water source type includes urban tap water, bottled water, bottled water, purified water, rural tap water, and others; The dietary habit information includes dining behavior type, average daily number of takeout meals, average weekly number of takeout meals, takeout meal type, vegetarian preference marker, low-fat preference marker, and low-sugar preference marker; the dining behavior type includes cooking at home, ordering takeout, dining at restaurants, and dining in the cafeteria; the takeout meal type includes Chinese food, Western food, light meals and salads, barbecue, desserts and beverages, and late-night snacks; the vegetarian preference marker, the low-fat preference marker, and the low-sugar preference marker each include two states: yes and no; The information on meal habits includes breakfast habit type, lunch habit type, dinner habit type, and late-night snack habit type; each habit type includes three categories: eat, eat occasionally, and never eat. The monthly average food information includes the monthly average intake type of multiple specific food categories; the monthly average intake type of each specific food category includes almost no food, 1-2 times per week, 3-4 times per week, 5-6 times per week, and daily; the multiple specific food categories include grains, vegetables, fruits, meat and poultry, aquatic products, eggs, dairy products, soy products, fried foods, pickled foods, desserts, and beverages. The daily food information includes staple food type, dish type, meal portion type, and dish portion type; the staple food type includes rice and noodles; the dish type includes vegetables as the main component, meat and eggs as the main component, and a balanced mix of vegetables and meat; the meal portion type includes small, moderate, and excessive; the dish portion type includes small, moderate, and excessive. The observation information includes observations of mental appearance, body shape and posture, gait and movement, communication and interaction, complexion and skin color, eye condition, lip condition, skin condition, hair condition, nail condition, breathing condition, voice condition, temperature sensitivity, sweat condition, tongue condition, tongue coating condition, stool condition, and urination condition. The mental appearance observation consists of multiple selectable categories of mental appearance states; the body shape and posture observation consists of multiple selectable categories of body shape features; the gait and movement observation consists of multiple selectable categories of gait features; the communication and interaction observation consists of multiple selectable categories of communication ability features; the complexion and skin color observation consists of multiple selectable categories of complexion and skin color states; the eye condition observation consists of multiple selectable categories of eye features; and the lip condition observation consists of multiple selectable categories of lip features. The system comprises: skin condition observation (consisting of multiple selectable skin texture features); hair condition observation (consisting of multiple selectable hair features); nail condition observation (consisting of multiple selectable nail color and texture features); respiratory condition observation (consisting of multiple selectable respiratory function features); voice condition observation (consisting of multiple selectable volume and tone quality features); cold / heat condition observation (consisting of multiple selectable cold / heat manifestations); sweat condition observation (consisting of multiple selectable sweating features); tongue condition observation (consisting of multiple selectable tongue color and shape features); tongue coating condition observation (consisting of multiple selectable tongue coating color, texture, and dryness / wetness features); stool condition observation (consisting of multiple selectable stool shape, color, and frequency features); and urination condition observation (consisting of multiple selectable urination shape, color, and frequency features). The disease strategy map includes a first node set and a first edge set; The first node set includes multiple first nodes; the node attributes of the first nodes include node identifier, node type, and node characteristics; the node types include disease nodes, blocking strategy nodes, intervention strategy nodes, and control strategy nodes; when the node type is a disease node, the node characteristic is the disease name; when the node type is a blocking strategy node, intervention strategy node, or control strategy node, the node characteristic is the corresponding first-stage blocking strategy identifier, second-stage intervention strategy identifier, or third-stage control strategy identifier; each first node whose node type is a disease node corresponds to a type of chronic disease; each first node whose node type is a blocking strategy node, intervention strategy node, or control strategy node corresponds to a first-order blocking strategy, second-order intervention strategy, or third-order control strategy for a type of chronic disease. The first edge set includes multiple first edges; the edge attributes of the first edge include edge identifier, parent node identifier, child node identifier, and edge association relationship; the node type of the first node corresponding to the parent node identifier of each first edge is a disease node, the node type of the first node corresponding to the child node identifier is a blocking strategy node, an intervention strategy node, or a control strategy node, and the edge association relationship is a one-stage association, a two-stage association, or a three-stage association corresponding to the child node identifier; The strategy knowledge base includes multiple first strategy records; each first strategy record corresponds to a first-order blocking strategy, a second-order intervention strategy, or a third-order control strategy for a class of chronic diseases; the first strategy record includes a first strategy identifier and a first strategy description; the first strategy description is used to explain the specific strategy execution details of the current strategy; The first dataset includes multiple first data records; the first data record includes the individual information set X and the label vector Y. * The label vector Y * Includes N disease vectors The disease vector Each of the chronic diseases corresponds one-to-one; each disease vector Includes 4 label probabilities ; 4 label probabilities Each of the four hierarchical types corresponds one-to-one; each disease vector The four label probabilities There is only one 1 and the other 3 are 0.

[0009] Preferably, the model input terminal of the chronic disease classification prediction model is used to receive the individual information set X, and the model output terminal is used to output the corresponding prediction vector Y; The chronic disease classification prediction model includes an embedding coding layer, a first linear activation layer, a second linear activation layer, a third linear activation layer, N i-th prediction networks, and an output layer; the N i-th prediction networks correspond one-to-one with the N types of chronic diseases. The input of the embedded coding layer is connected to the input of the model, and its output is connected to the input of the first linear activation layer; the output of the first linear activation layer is connected to the input of the second linear activation layer; the output of the second linear activation layer is connected to the input of the third linear activation layer; the output of the third linear activation layer is connected to the inputs of N i-th prediction networks respectively; the outputs of the N i-th prediction networks are connected to the N inputs corresponding to the output layer respectively; the output of the output layer is connected to the output of the model. The embedding coding layer performs embedding coding on each data item of the individual information set X according to a preset embedding coding rule, and sequentially concatenates all codes and coding vectors to obtain the corresponding embedding vector E, which is then sent to the first linear activation layer. The embedding coding rule specifies that: z-core normalization coding is used for numerical data; one-hot coding is used for binary data; one-hot encoding vectors are used for multi-class data (binary or higher classification but only single selection is allowed); multi-hot encoding vectors are used for multi-class data (binary or higher classification and multiple selection is allowed); and word embedding coding is used for text data. The embedding vector E includes multiple codes e. k 1≤indexk≤L E L E The total number of codes in the embedding vector E; the vector shape of the embedding vector E is 1×L E ; The first linear activation layer is used to perform feature encoding on the embedding vector E to obtain the corresponding feature vector H1, which is then sent to the second linear activation layer. The encoding method of the feature vector H1 is as follows: ; ReLU() is the ReLU activation function; W1 and b1 are the weight matrix and bias vector of the first linear activation layer; the shape of the weight matrix W1 is L. E ×D1, the shape of the bias vector b1 is 1×D1, the shape of the feature vector H1 is 1×D1, and D1 is the preset first feature dimension; The second linear activation layer is used to encode the feature vector H1 to obtain the corresponding feature vector H2, which is then sent to the third linear activation layer. The encoding method of the feature vector H2 is as follows: ; W2 and b2 are the weight matrix and bias vector of the second linear activation layer; the weight matrix W2 has a shape of D1×D2, the bias vector b2 has a shape of 1×D2, and the feature vector H2 has a shape of 1×D2, where D2 is a preset second feature dimension. The third linear activation layer is used to encode the feature vector H2 to obtain the corresponding feature vector H3, which is then sent to the N i-th prediction networks. The encoding method of the feature vector H3 is as follows: ; W3 and b3 are the weight matrix and bias vector of the third linear activation layer; the weight matrix W3 has a shape of D2×D3, the bias vector b3 has a shape of 1×D3, and the feature vector H3 has a shape of 1×D3, where D3 is a preset third feature dimension. Each of the i-th prediction networks is used to perform feature vector transformation on the feature vector H3 in the i-th classification feature space to obtain the corresponding feature vector S. i Send to the output layer; Wherein, each of the aforementioned feature vectors S i The vector transformation method is as follows: ; , Let the first and second weight matrices of the i-th prediction network be... , The first and second bias vectors of the i-th prediction network; the first weight matrix. The shape is D3×D4, and the first bias vector The shape is 1×D4, where D4 is the preset fourth feature dimension; the second weight matrix The shape is D4×D5, and the second bias vector The shape is 1×D5, where D5 is the preset fifth feature dimension, and D5=4; each of the feature vectors S i The shape is 1×4, consisting of 4 corresponding feature data s i,j composition; The output layer is used to process the various feature vectors S i Substitute into the Softmax function to calculate the corresponding disease vector y i ; and the obtained N disease vectors y i The corresponding prediction vector Y is then constructed and output.

[0010] Preferably, the construction of the first dataset through data collection specifically includes: Step 41: Based on the aforementioned N types of chronic diseases, recruit multiple volunteers to form a volunteer group; and recruit multiple medical experts to form an expert group; The volunteer group includes healthy volunteers and chronic disease volunteers; each chronic disease volunteer suffers from one or more of the N types of chronic diseases; the set of chronic disease types of all chronic disease volunteers includes the N types of chronic diseases; the expert group includes at least several specialists or general practitioners in the fields of Western medicine, traditional Chinese medicine, and integrated traditional Chinese and Western medicine. Step 42: Each volunteer in the volunteer group is designated as the current volunteer; with authorization from the current volunteer or their guardian, data is collected from the individual information set X of the current volunteer; the expert group then identifies whether the current volunteer suffers from a chronic disease through expert consultation, and if a chronic disease is confirmed, further identifies the progression of the chronic disease, and sets the corresponding label vector Y based on the identification results. * ; and comprised of the individual information set X and the label vector Y * Form the corresponding first data record; Step 43: The first dataset is composed of all the first data records obtained.

[0011] Preferably, training the chronic disease classification prediction model based on the first dataset specifically includes: Step 51: Based on a preset first segmentation ratio, the first dataset is randomly divided into two sub-datasets, denoted as the first training set and the first evaluation set. The first training set and the first evaluation set are each composed of multiple first data records; the total number of records in the first training set and the first evaluation set is denoted as N. tr N av The ratio of the total number of records in the first training set to the total number of records in the first evaluation set is N. tr :N av Satisfying the first segmentation ratio; the label vector Y of each of the first data records in the first training set. * Record as the corresponding The label vector Y * Corresponding label probability Record as the corresponding 1 ≤ index u ≤ N tr ; Step 52: Input the individual information set X of each of the first data records in the first training set into the chronic disease hierarchical prediction model for processing, and record the prediction vector Y obtained in this processing as the corresponding Y. u And the prediction probabilities y of the prediction vector Y.i,j Let y be the corresponding u,i,j And the prediction vector Y obtained this time u Its corresponding label vector Form the corresponding first prediction-label pair; Step 53, obtain N tr The first prediction-label is substituted into the preset first model loss function L1 to calculate the corresponding first loss value; Specifically, the loss function L1 of the first model is: ; Step 54: Identify whether the first loss value meets the preset first loss value range; if it does, proceed to step 55; if it does not, perform a round of modulation on the model parameters of the first prediction model based on the preset first model optimizer in the direction of minimizing the first model loss function L1, and return to step 52 when the current round of modulation ends. The first model optimizer includes the Adam optimizer and the SGD optimizer. Step 55: Input the individual information set X of each of the first data records in the first evaluation set into the chronic disease classification prediction model for processing, and then use the prediction vector Y obtained from this processing and its corresponding label vector Y * Form the corresponding second prediction-label pair; and based on the obtained N av The second prediction-label pair is evaluated for accuracy, precision, recall, and F1 score to obtain the corresponding first accuracy, first precision, first recall, and first F1 score; Step 56: Identify whether the first accuracy, first precision, first recall, and first F1 score each satisfy their respective first accuracy range, first precision range, first recall range, and first F1 score range; if not, return to step 51; if yes, stop training and confirm that the model training is complete.

[0012] Preferably, the step of retrieving strategy identifiers based on the prediction vector Y, the disease strategy atlas, and the strategy knowledge base, and synthesizing a corresponding response strategy report based on the retrieval results to provide feedback to the current user, specifically includes: The prediction vector Y is composed of each of the disease vectors y. iAs the current vector; and when the tier type corresponding to the current vector with the highest probability is not the healthy stage, the chronic disease corresponding to the current vector is taken as the corresponding first disease, the current tier type is taken as the corresponding first stage, the edge association relationship corresponding to the first stage is taken as the corresponding first relationship, and the first node in the disease strategy graph whose node type is a disease node and whose node characteristics match the first disease is taken as the current node, and each first node whose edge association relationship with the current node matches the first relationship is taken as the corresponding retrieval node, and the node characteristics of each retrieval node are taken as the corresponding first identifier, and all the first identifiers corresponding to the current vector form the corresponding first identifier set; and all the obtained first identifier sets are merged and deduplicated to obtain the corresponding second identifier set; Each of the first identifiers in the second identifier set is taken as the current identifier; the first strategy description of the first strategy record in the strategy knowledge base that matches the current identifier is extracted as the corresponding second strategy description; and all the obtained second strategy descriptions are used to form a corresponding response strategy set. Each of the first diseases and its corresponding first stage constitutes a corresponding disease stage information; and all the obtained disease stage information constitutes a corresponding disease course information set; The corresponding response strategy report, composed of the disease course information set and the response strategy set, is then fed back to the current user.

[0013] A second aspect of the present invention provides an apparatus for implementing the processing method for predicting tiered coping strategies for chronic diseases as described in the first aspect above. The apparatus includes: a model building module, a knowledge graph building module, a data acquisition module, a model training module, and a model application module. The model building module is used to construct a chronic disease hierarchical prediction model for predicting the third-order progression state of N types of chronic diseases based on individual information; the total number of chronic diseases N is a preset positive integer, and the N types of chronic diseases consist of N preset chronic disease categories; the chronic disease hierarchical prediction model is used to predict the third-order progression state of the N types of chronic diseases based on the individual information set X input to the model and output the corresponding prediction vector Y; the prediction vector Y includes N disease category vectors y. i The disease vector y i Each of the chronic diseases corresponds one-to-one, 1 ≤ index i ≤ N; the disease vector y i Includes 4 predicted probabilities y i,j , 1 ≤ index j ≤ 4; 4 predicted probabilities y i,jEach stage corresponds to one of four tiers: health stage, first-level stage, second-level stage, and third-level stage. The knowledge graph construction module constructs a disease strategy graph and strategy knowledge base based on the prior knowledge of first-order blocking, second-order intervention and third-order control strategies for various chronic diseases. The data acquisition module is used to construct a first dataset through data acquisition; The model training module trains the chronic disease classification prediction model based on the first dataset; The model application module is used to input the individual information set X input by any user into the chronic disease hierarchical prediction model after the model training is completed to predict the corresponding prediction vector Y; and to perform strategy identification retrieval based on the prediction vector Y, the disease strategy map and the strategy knowledge base, and synthesize a corresponding response strategy report based on the retrieval results to provide feedback to the current user.

[0014] A third aspect of the present invention provides an electronic device, including: a memory, a processor, and a transceiver; The processor is used to couple with the memory, read and execute instructions in the memory to implement the steps of the method described in the first aspect above; The transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.

[0015] A fourth aspect of the present invention provides a computer-readable storage medium storing computer instructions that, when executed by a computer, cause the computer to perform the instructions described in the first aspect.

[0016] This invention provides a method, apparatus, electronic device, and computer-readable storage medium for predicting tiered coping strategies for chronic diseases. As described above, this invention constructs a tiered chronic disease prediction model to predict the three-stage progression of N types of chronic diseases based on individual information. It constructs a disease strategy map and strategy knowledge base based on prior knowledge of first-stage blocking, second-stage intervention, and third-stage control strategies for various chronic diseases. A first dataset is constructed through data collection, and the tiered chronic disease prediction model is trained based on this dataset. After model training, an individual information set X input by any user is fed into the tiered chronic disease prediction model for prediction, resulting in a corresponding prediction vector Y. Based on the prediction vector Y, the disease strategy map, and the strategy knowledge base, a strategy identifier is retrieved, and a corresponding coping strategy report is synthesized based on the retrieval results and fed back to the current user. This invention enriches the feature dimensions of health information, improves the utilization rate of health information, realizes parallel assessment of multiple diseases and multiple stages, and improves the personalization and processing efficiency of strategy analysis. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a processing method for predicting tiered coping strategies for chronic diseases, provided in Embodiment 1 of the present invention. Figure 2 This is a model structure diagram of the chronic disease classification prediction model provided in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the node connection relationship of a disease strategy atlas provided in Embodiment 1 of the present invention; Figure 4 This is a module structure diagram of a processing device for predicting tiered coping strategies for chronic diseases, provided in Embodiment 2 of the present invention. Figure 5 This is a schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0019] Embodiment 1 of the present invention provides a method for predicting tiered coping strategies for chronic diseases, such as... Figure 1 The schematic diagram shows a method for predicting tiered coping strategies for chronic diseases provided in Embodiment 1 of the present invention. The method mainly includes the following steps: Step 1: Construct a chronic disease tiered prediction model to predict the three-stage progression of N types of chronic diseases based on individual information.

[0020] Here, the total number N of chronic diseases in this embodiment of the invention is a preset positive integer. The N types of chronic diseases consist of N preset types of chronic diseases. The chronic diseases mentioned in this embodiment of the invention include coronary heart disease, stroke, arteriosclerosis obliterans, thromboangiitis obliterans, deep vein thrombosis, varicose veins, diabetic complications, COPD, asthma, depression, anxiety, bipolar disorder, schizophrenia, obsessive-compulsive disorder, etc.

[0021] The chronic disease progression prediction model of this invention is used to predict the third-order progression state of N types of chronic diseases based on the individual information set X input to the model and output the corresponding prediction vector Y.

[0022] The prediction vector Y in this embodiment of the invention includes N disease vectors y i Disease vector y i There is a one-to-one correspondence between chronic diseases, where 1 ≤ index i ≤ N; the disease vector y i Includes 4 predicted probabilities y i,j 1 ≤ index j ≤ 4; 4 predicted probabilities y i,j It corresponds one-to-one with the four tiered types, which include the health stage, the first-level stage, the second-level stage, and the third-level stage.

[0023] The individual information set X in this embodiment of the invention includes basic information, measurement information, allergy information, surgical information, female reproductive health information, medical history information, lifestyle information, and observation information.

[0024] 1) Basic Information: Basic information includes age, birth weight, and gender type.

[0025] 2) Measurement information: Measurement information includes height, weight, BMI, waist circumference, hip circumference, neck circumference, calf circumference, grip strength, blood pressure, heart rate, blood oxygen saturation, body temperature, body fat percentage, skeletal muscle mass, and body water percentage.

[0026] 3) Allergy information: Allergy information includes four categories of allergy markers: no allergy marker, drug allergy marker, food allergy marker, and other allergy markers.

[0027] All four categories of labels include both yes and no states. When the "no allergy" label is "yes," the other three categories of labels are "no"; when the "no allergy" label is "no," some or all of the other three categories of labels are "yes."

[0028] 4) Surgical information: Surgical information includes surgical markers and the time interval between recent surgeries.

[0029] The surgical marker includes two states: yes and no. When the surgical marker is no, the most recent surgical time is empty. When the surgical marker is yes, the most recent surgical interval is the interval between the most recent surgical time and the current time.

[0030] 5) Women's reproductive health information: Women’s reproductive health information includes markers for regularity of menstruation, menstrual cycle, planning for pregnancy, pregnancy, having given birth, and menopause.

[0031] The markers for regular menstruation, preparing for pregnancy, pregnancy, having given birth, and menopause all include a yes / no status. When the regularity marker is no, the menstrual cycle is empty. When the gender type is male, all markers for female reproductive health information are no.

[0032] 6) Medical history information: Medical history information includes a history of metabolic diseases, cardiovascular and cerebrovascular diseases, respiratory diseases, digestive diseases, urinary diseases, endocrine diseases, reproductive diseases, musculoskeletal diseases, nervous diseases, immune system diseases, circulatory system diseases, and mental and psychological diseases.

[0033] Each type of medical history includes information about one or more corresponding diseases.

[0034] Each disease information consists of a corresponding set of disease name, disease marker, incidence frequency type, and disease control status. Among them, the disease marker includes two statuses: yes and no; the incidence frequency type includes asymptomatic, occasional, frequent, and persistent; and the disease control status includes meeting the target, partially meeting the target, not meeting the target, and not being monitored.

[0035] 7) Lifestyle information: Lifestyle information includes information on smoking habits, drinking habits, air conditioning usage habits, exercise habits, sleep quality, mental stress, drinking habits, eating habits, three meals a day, monthly average food intake, and daily average food intake.

[0036] Smoking habit information includes smoking markers and smoking history; smoking markers include two states: yes and no; when the smoking marker is no, the smoking history is empty.

[0037] Drinking habit information includes drinking markers and age of drinking; drinking markers include two states: yes and no; when the drinking marker is no, the age of drinking is empty.

[0038] Air conditioner usage information includes air conditioner usage status, air conditioner usage time period type, and air conditioner usage years. Air conditioner usage status includes two states: yes and no; air conditioner usage time period type includes year-round, summer, and winter.

[0039] Exercise habit information includes exercise habit tags, exercise cycle, exercise type, and duration of a single exercise session. Exercise habit tags have two states: yes and no. When the exercise habit tag is no, the exercise cycle, exercise type, and duration of a single exercise session are all empty.

[0040] Sleep quality information includes sleep quality type and daily sleep duration. Sleep quality types include very good, good, average, poor, and very poor.

[0041] The information on mental stress includes a mental stress label and a stressor type. The mental stress label includes two states: yes and no; the stressor type includes work, family, interpersonal relationships, social environment, and other; when the mental stress label is no, the stressor type is empty.

[0042] Information on drinking habits includes water source type, drinking water temperature, and daily water consumption. Water source types include municipal tap water, bottled water, bottled water, purified water, rural tap water, and others.

[0043] Dietary habit information includes dining behavior type, average daily number of takeout meals, average weekly number of takeout meals, takeout meal type, vegetarian preference marker, low-fat preference marker, and low-sugar preference marker. Among them, dining behavior type includes cooking at home, ordering takeout, dining at restaurants, and dining in the cafeteria; takeout meal type includes Chinese food, Western food, light meals and salads, barbecue, desserts and beverages, and late-night snacks; vegetarian preference marker, low-fat preference marker, and low-sugar preference marker all include two states: yes and no.

[0044] The information on meal habits includes breakfast habit types, lunch habit types, dinner habit types, and late-night snack habit types. Each habit type includes three categories: eating, occasionally eating, and never eating.

[0045] Monthly food information includes the average monthly intake patterns for several specific food groups. The average monthly intake patterns for each specific food group include: almost never eaten, 1-2 times per week, 3-4 times per week, 5-6 times per week, and daily. These specific food groups include grains, vegetables, fruits, meat and poultry, seafood, eggs, dairy products, soy products, fried foods, pickled foods, desserts, and beverages.

[0046] Daily food information includes staple food type, dish type, meal portion type, and dish portion type; staple food type includes rice and noodles; dish type includes vegetables as the main component, meat and eggs as the main component, and a balanced mix of vegetables and meat; meal portion type includes small amount, moderate amount, and excessive amount; dish portion type includes small amount, moderate amount, and excessive amount.

[0047] The observation information includes observations of mental appearance, body shape and posture, gait and movement, communication and interaction, complexion and skin color, eye condition, lip condition, skin condition, hair condition, nail condition, breathing condition, voice condition, cold and heat condition, sweat condition, tongue condition, tongue coating condition, stool condition, and urination condition.

[0048] The observation of mental state consists of multiple categories of mental state states that can be selected from multiple options; Body shape and posture observation consists of multiple types of body posture features that can be selected from multiple options; Gait observation consists of multiple types of gait features that can be selected from multiple options; Communication observation consists of multiple categories of communication ability characteristics that can be selected from multiple options; The facial and skin tone observation consists of multiple categories of facial and skin tone conditions that can be selected in multiple ways; Eye condition observation consists of multiple types of eye features that can be selected from multiple options; The lip state observation consists of multiple lip features that can be selected from multiple categories; Skin condition observation consists of multiple selectable skin texture features; Hair condition observation consists of multiple types of hair features that can be selected from multiple options; Nail condition observation consists of multiple selectable nail color and texture characteristics; Respiratory status observation consists of multiple types of selectable respiratory function features; Sound state observation consists of multiple selectable volume and sound quality features; The observation of cold and heat states consists of multiple types of cold and heat characteristic states that can be selected in multiple ways; The sweat state observation consists of multiple selectable sweating characteristics; The observation of tongue condition consists of multiple types of tongue color and morphological features that can be selected from multiple options; The observation of tongue coating condition consists of multiple selectable characteristics of tongue coating color, texture, and dryness / wetness; The stool state observation consists of multiple selectable stool morphology, color, and frequency characteristics; the urine state observation consists of multiple selectable urine morphology, color, and frequency characteristics.

[0049] like Figure 2 As shown in the model structure diagram of the chronic disease tiered prediction model provided in Embodiment 1 of the present invention, the model input end of the chronic disease tiered prediction model is used to receive individual information set X, and the model output end is used to output the corresponding prediction vector Y.

[0050] like Figure 2As shown, the model components of the chronic disease classification prediction model include: an embedding coding layer, a first linear activation layer, a second linear activation layer, a third linear activation layer, N i-th prediction networks, and an output layer; the N i-th prediction networks correspond one-to-one with the N types of chronic diseases.

[0051] like Figure 2 As shown, the connection relationships of the model components in the chronic disease classification prediction model are as follows: the input of the embedded coding layer is connected to the model input, and its output is connected to the input of the first linear activation layer; the output of the first linear activation layer is connected to the input of the second linear activation layer; the output of the second linear activation layer is connected to the input of the third linear activation layer; the output of the third linear activation layer is connected to the inputs of the N i-th prediction networks respectively; the outputs of the N i-th prediction networks are connected to the N inputs of the corresponding output layer respectively; and the output of the output layer is connected to the model output.

[0052] The functionalities of the model components in the chronic disease classification prediction model are shown below.

[0053] 1) Embedded coding layer: In this embodiment of the invention, the embedding coding layer performs embedding coding on each data item of the individual information set X according to a preset embedding coding rule, and sequentially concatenates all codes and coding vectors to obtain the corresponding embedding vector E, which is then sent to the first linear activation layer.

[0054] Here, the embedding encoding rules of this embodiment of the invention stipulate that: numerical data is encoded using z-core normalization encoding, binary data is encoded using one-hot encoding, multi-class data with more than two categories but only single selection is encoded using one-hot encoding vector encoding, multi-class data with more than two categories and multiple selection is encoded using multi-hot encoding vector encoding, and text data is encoded using word embedding encoding.

[0055] The embedding vector E in this embodiment of the invention includes multiple encodings e k 1≤indexk≤L E L E This represents the total number of codes in the embedding vector E; the shape of the embedding vector E is 1×L. E .

[0056] 2) First linear activation layer: In this embodiment of the invention, the first linear activation layer is used to perform feature encoding on the embedded vector E to obtain the corresponding feature vector H1, which is then sent to the second linear activation layer.

[0057] Here, the encoding method of feature vector H1 in this embodiment of the invention is as follows: ; Where ReLU() is the ReLU activation function; W1 and b1 are the weight matrix and bias vector of the first linear activation layer; the shape of the weight matrix W1 is L. E ×D1, the bias vector b1 has a shape of 1×D1, the feature vector H1 has a shape of 1×D1, and D1 is the preset first feature dimension.

[0058] 3) Second linear activation layer: In this embodiment of the invention, the second linear activation layer is used to encode the feature vector H1 to obtain the corresponding feature vector H2, which is then sent to the third linear activation layer.

[0059] Here, the encoding method of feature vector H2 in this embodiment of the invention is as follows: ; Where W2 and b2 are the weight matrix and bias vector of the second linear activation layer; the shape of the weight matrix W2 is D1×D2, the shape of the bias vector b2 is 1×D2, the shape of the feature vector H2 is 1×D2, and D2 is the preset second feature dimension.

[0060] 4) Third linear activation layer: In this embodiment of the invention, the third linear activation layer is used to encode the feature vector H2 to obtain the corresponding feature vector H3, which is then sent to the N i-th prediction networks.

[0061] Here, the encoding method of feature vector H3 in this embodiment of the invention is as follows: ; Where W3 and b3 are the weight matrix and bias vector of the third linear activation layer; the weight matrix W3 has a shape of D2×D3, the bias vector b3 has a shape of 1×D3, the feature vector H3 has a shape of 1×D3, and D3 is the preset third feature dimension.

[0062] 5) Each i-th prediction network: Each i-th prediction network in this embodiment of the invention is used to perform feature vector transformation on the feature vector H3 in the i-th classification feature space to obtain the corresponding feature vector S. i Send to the output layer.

[0063] Here, the various feature vectors S in the embodiments of the present invention i The vector transformation method is as follows: ; in, , Let the first and second weight matrices of the i-th prediction network be... , Here are the first and second bias vectors of the i-th prediction network; the first weight matrix. The shape is D3×D4, and the first bias vector The shape is 1×D4, where D4 is the preset fourth feature dimension; the second weight matrix The shape is D4×D5, and the second bias vector The shape is 1×D5, where D5 is the preset fifth feature dimension, D5=4; each feature vector S i The shape is 1×4, consisting of 4 corresponding feature data s i,j composition.

[0064] 6) Output layer: The output layer is used to process the various feature vectors S i Substitute into the Softmax function to calculate the corresponding disease vector y i ; and from the obtained N disease vectors y i The corresponding prediction vector Y is constructed and output.

[0065] Step 2: Construct a disease strategy map and strategy knowledge base based on prior knowledge of first-order blocking, second-order intervention, and third-order control strategies for various chronic diseases.

[0066] Here, the disease strategy graph in this embodiment of the invention includes a first node set and a first edge set.

[0067] The first node set comprises multiple first nodes; the node attributes of a first node include node identifier, node type, and node characteristics. Node types include disease nodes, blocking strategy nodes, intervention strategy nodes, and control strategy nodes. When the node type is a disease node, the node characteristic is the disease name; when the node type is a blocking strategy node, intervention strategy node, or control strategy node, the node characteristic is the corresponding first-stage blocking strategy identifier, second-stage intervention strategy identifier, or third-stage control strategy identifier. Each first node of type disease corresponds to a type of chronic disease; each first node of type blocking strategy, intervention strategy, or control strategy corresponds to a type of first-order blocking strategy, second-order intervention strategy, or third-order control strategy for a type of chronic disease.

[0068] The first edge set includes multiple first edges; the edge attributes of the first edge include edge identifier, parent node identifier, child node identifier, and edge association relationship; the node type of the first node corresponding to the parent node identifier of each first edge is a disease node, the node type of the first node corresponding to the child node identifier is a blocking strategy node, an intervention strategy node, or a control strategy node, and the edge association relationship is a one-stage association, a two-stage association, or a three-stage association corresponding to the child node identifier.

[0069] Figure 3 This is a schematic diagram of the node connection relationship of a disease strategy atlas provided in Embodiment 1 of the present invention, which can be referred to. Figure 3To understand the node network of the disease strategy map.

[0070] Furthermore, the strategy knowledge base of this embodiment includes multiple first strategy records. Each first strategy record corresponds to a first-order blocking strategy, a second-order intervention strategy, or a third-order control strategy for a class of chronic diseases. The first strategy record includes a first strategy identifier and a first strategy description; wherein, the first strategy description is used to explain the specific strategy execution details of the current strategy.

[0071] Step 3: Construct the first dataset through data collection.

[0072] Here, the first dataset in this embodiment of the invention includes multiple first data records; the first data record includes an individual information set X and a label vector Y. * Label vector Y * Includes N disease vectors Disease vector Each disease corresponds one-to-one with a chronic disease; each disease vector Includes 4 label probabilities ; 4-label probability Each disease corresponds one-to-one with one of the four hierarchical types; each disease vector The probability of the 4 tags There is only one 1 and the other 3 are 0.

[0073] Step 3 specifically includes: Step 31: Recruit multiple volunteers to form a volunteer group based on N types of chronic diseases; and recruit multiple medical experts to form an expert group.

[0074] Here, in this embodiment of the invention, the volunteer group includes healthy volunteers and chronic disease volunteers; each chronic disease volunteer suffers from one or more of the N types of chronic diseases; the set of chronic disease types of all chronic disease volunteers includes the N types of chronic diseases.

[0075] The expert group in this embodiment of the invention includes at least several specialists or general practitioners in the fields of Western medicine, traditional Chinese medicine, and integrated traditional Chinese and Western medicine.

[0076] Step 32: Each volunteer in the volunteer group is designated as the current volunteer; with authorization from the current volunteer or their guardian, data is collected from the individual information set X of the current volunteer; an expert group then identifies whether the current volunteer has a chronic disease through expert consultation, and if a chronic disease is confirmed, further identifies the progression of the chronic disease, and sets a corresponding label vector Y based on the identification results. * ; and consists of individual information set X and label vector Y * This forms the corresponding first data record.

[0077] Step 33: The first dataset is composed of all the first data records obtained.

[0078] Step 4: Train a chronic disease classification prediction model based on the first dataset.

[0079] Specifically, it includes: Step 41: Based on the preset first segmentation ratio, the first dataset is randomly divided into two sub-datasets, denoted as the first training set and the first evaluation set.

[0080] Here, in this embodiment of the invention, both the first training set and the first evaluation set consist of multiple first data records; the total number of records in the first training set and the first evaluation set is denoted as N. tr N av The ratio of the total number of records in the first training set to the total number of records in the first evaluation set is N. tr :N av Satisfying the first segmentation ratio; the label vector Y of each first data record in the first training set. * Record as the corresponding Label vector Y * Corresponding label probability Record as the corresponding 1 ≤ index u ≤ N tr .

[0081] Step 42: Input the individual information set X of each first data record in the first training set into the chronic disease hierarchical prediction model for processing, and record the prediction vector Y obtained from this processing as the corresponding Y. u And the prediction probabilities y of the prediction vector Y i,j Let y be the corresponding u,i,j And the prediction vector Y obtained this time u Its corresponding label vector Form the corresponding first prediction-label pair.

[0082] Step 43, obtain N tr The first prediction label is substituted into the preset first model loss function L1 to calculate the corresponding first loss value.

[0083] Here, the first model loss function L1 in this embodiment of the invention is specifically: .

[0084] Step 44: Identify whether the first loss value meets the preset first loss value range; if it does, proceed to step 45; if it does not, modulate the model parameters of the first prediction model in one round based on the preset first model optimizer in the direction of minimizing the first model loss function L1, and return to step 42 when the first round of modulation ends.

[0085] Here, the first loss value range in this embodiment of the invention is a pre-set numerical range. The first model optimizer includes the Adam optimizer and the SGD optimizer.

[0086] Step 45: Input the individual information set X of each first data record in the first evaluation set into the chronic disease hierarchical prediction model for processing, and then use the prediction vector Y obtained from this processing and its corresponding label vector Yi to... * Form the corresponding second prediction-label pair; and based on the obtained N av The first precision, first accuracy, first recall, and first F1 score are obtained by evaluating the accuracy, precision, recall, and F1 score of each second prediction-label pair.

[0087] Step 46: Identify whether the first accuracy, first precision, first recall, and first F1 score each satisfy their respective first accuracy range, first precision range, first recall range, and first F1 score range; if not, return to step 41; if yes, stop training and confirm that the model training is complete.

[0088] Here, the first accuracy range, the first precision range, the first recall range, and the first F1 score range in this embodiment of the invention are four preset numerical ranges.

[0089] Step 5: After the model training is completed, the individual information set X input by any user is sent to the chronic disease hierarchical prediction model to obtain the corresponding prediction vector Y; and based on the prediction vector Y, the disease strategy map and the strategy knowledge base, the strategy identifier is retrieved, and the corresponding response strategy report is synthesized based on the retrieval results and fed back to the current user.

[0090] Specifically, step 51 involves sending the individual information set X input by any user into the chronic disease classification prediction model to obtain the corresponding prediction vector Y.

[0091] Step 52 involves retrieving strategy identifiers based on the prediction vector Y, the disease strategy map, and the strategy knowledge base, and synthesizing a corresponding response strategy report based on the retrieval results to provide feedback to the current user.

[0092] Specifically, it includes: Step 521, convert the prediction vector Y to the disease vectors y of each disease. iAs the current vector; and when the tier type corresponding to the current vector with the highest probability is not the healthy stage, the chronic disease corresponding to the current vector is taken as the corresponding first disease, the current tier type is taken as the corresponding first stage, the edge association relationship corresponding to the first stage is taken as the corresponding first relationship, the first node in the disease strategy graph whose node type is a disease node and whose node characteristics match the first disease is taken as the current node, and the first node whose edge association relationship with the current node matches the first relationship is taken as the corresponding retrieval node, and the node characteristics of each retrieval node are taken as the corresponding first identifier, and all the first identifiers corresponding to the current vector form the corresponding first identifier set; and all the obtained first identifier sets are merged and deduplicated to obtain the corresponding second identifier set.

[0093] Step 522: Take each of the first identifiers in the second identifier set as the current identifier; extract the first strategy description of the first strategy record in the strategy knowledge base that matches the current identifier as the corresponding second strategy description; and form the corresponding response strategy set by all the obtained second strategy descriptions.

[0094] Step 523: Each first disease and its corresponding first stage form a corresponding disease stage information; and all the obtained disease stage information forms a corresponding disease course information set.

[0095] Step 524: A corresponding response strategy report, composed of the disease course information set and the response strategy set, is fed back to the current user.

[0096] Figure 4 This is a module structure diagram of a processing device for predicting tiered coping strategies for chronic diseases, provided in Embodiment 2 of the present invention. This device can be a terminal device or server implementing the aforementioned method embodiments, or it can be a device that enables the aforementioned terminal device or server to implement the aforementioned method embodiments. For example, the device can be a device or chip system of the aforementioned terminal device or server. Figure 4 As shown, the device includes: a model building module 201, a knowledge graph building module 202, a data acquisition module 203, a model training module 204, and a model application module 205.

[0097] The model building module 201 is used to build a chronic disease hierarchical prediction model for predicting the third-order progression state of N types of chronic diseases based on individual information; the total number of chronic diseases N is a preset positive integer, and the N types of chronic diseases consist of N preset chronic disease categories; the chronic disease hierarchical prediction model is used to predict the third-order progression state of N types of chronic diseases based on the individual information set X input to the model and output the corresponding prediction vector Y; the prediction vector Y includes N disease category vectors y i Disease vector y i There is a one-to-one correspondence between chronic diseases, where 1 ≤ index i ≤ N; the disease vector yi Includes 4 predicted probabilities y i,j 1 ≤ index j ≤ 4; 4 predicted probabilities y i,j It corresponds one-to-one with the four tiered types, which include the health stage, the first-level stage, the second-level stage, and the third-level stage.

[0098] The knowledge graph construction module 202 constructs disease strategy graphs and strategy knowledge bases based on prior knowledge of first-order blocking, second-order intervention, and third-order control strategies for various chronic diseases.

[0099] The data acquisition module 203 is used to construct the first dataset through data acquisition.

[0100] Model training module 204 trains a chronic disease classification prediction model based on the first dataset.

[0101] The model application module 205 is used to input any user-input individual information set X into the chronic disease hierarchical prediction model after the model training is completed to obtain the corresponding prediction vector Y; and to perform strategy identification retrieval based on the prediction vector Y, the disease strategy map and the strategy knowledge base, and synthesize the corresponding response strategy report based on the retrieval results to provide feedback to the current user.

[0102] The processing device for predicting tiered coping strategies for chronic diseases provided in this embodiment of the invention can execute the method steps in the above method embodiment. Its implementation principle and technical effect are similar, and will not be repeated here.

[0103] It should be noted that the division of the various modules in the above device is merely a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, these modules can be implemented entirely in software via processing elements; they can be fully implemented in hardware; or some modules can be implemented by processing elements calling software, while others are implemented in hardware. For example, the model building module can be a separate processing element, or it can be integrated into a chip in the above device. Alternatively, it can be stored as program code in the memory of the above device, and its functions can be called and executed by a processing element. The implementation of other modules is similar. Moreover, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed through integrated logic circuits in the hardware of the processor element or through software instructions.

[0104] For example, these modules can be one or more integrated circuits configured to implement the above methods, such as one or more Application Specific Integrated Circuits (ASICs), one or more Digital Signal Processors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs). As another example, when a module is implemented using processing element scheduler code, the processing element can be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. Furthermore, these modules can be integrated together as a System-on-a-Chip (SOC).

[0105] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the foregoing method embodiments are generated. The computer described above can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The aforementioned computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the aforementioned computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, Bluetooth, microwave, etc.) means. The aforementioned computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The aforementioned available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state disks (SSDs)).

[0106] Figure 5 This is a schematic diagram of an electronic device provided in Embodiment 3 of the present invention. This electronic device can be a terminal device or server implementing the methods of the aforementioned embodiments, or it can be a terminal device or server connected to the aforementioned terminal device or server implementing the methods of the aforementioned embodiments. Figure 5As shown, the electronic device may include: a processor 301 (e.g., CPU), a memory 302, and a transceiver 303; the transceiver 303 is coupled to the processor 301, and the processor 301 controls the transmission and reception operations of the transceiver 303. The memory 302 may store various instructions for performing various processing functions and implementing the processing steps described in the foregoing embodiments. Preferably, the electronic device involved in the embodiments of the present invention further includes: a power supply 304, a system bus 305, and a communication port 306. The system bus 305 is used to realize communication connections between components. The communication port 306 is used for communication between the electronic device and other peripherals.

[0107] exist Figure 5 The system bus 305 mentioned can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This system bus can be divided into address bus, data bus, control bus, etc. For ease of representation, it is represented by only one thick line in the figure, but this does not indicate that there is only one bus or one type of bus. The communication interface is used to enable communication between the database access device and other devices (e.g., clients, read-write libraries, and read-only libraries). Memory may include Random Access Memory (RAM) and may also include Non-Volatile Memory, such as at least one disk storage device.

[0108] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), graphics processing units (GPUs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0109] It should be noted that the embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when run on a computer, cause the computer to perform the methods and processes provided in the above embodiments.

[0110] This invention provides a method, apparatus, electronic device, and computer-readable storage medium for predicting tiered coping strategies for chronic diseases. As described above, this invention constructs a tiered chronic disease prediction model to predict the three-stage progression of N types of chronic diseases based on individual information. It constructs a disease strategy map and strategy knowledge base based on prior knowledge of first-stage blocking, second-stage intervention, and third-stage control strategies for various chronic diseases. A first dataset is constructed through data collection, and the tiered chronic disease prediction model is trained based on this dataset. After model training, an individual information set X input by any user is fed into the tiered chronic disease prediction model for prediction, resulting in a corresponding prediction vector Y. Based on the prediction vector Y, the disease strategy map, and the strategy knowledge base, a strategy identifier is retrieved, and a corresponding coping strategy report is synthesized based on the retrieval results and fed back to the current user. This invention enriches the feature dimensions of health information, improves the utilization rate of health information, realizes parallel assessment of multiple diseases and multiple stages, and improves the personalization and processing efficiency of strategy analysis.

[0111] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0112] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting tiered coping strategies for chronic diseases, characterized in that, The method includes: A chronic disease hierarchical prediction model is constructed to predict the third-order progression state of N types of chronic diseases based on individual information. The total number of chronic diseases N is a preset positive integer, and the N types of chronic diseases consist of N preset chronic disease categories. The chronic disease hierarchical prediction model is used to predict the third-order progression state of the N types of chronic diseases based on the individual information set X input to the model and outputs the corresponding prediction vector Y. The prediction vector Y includes N disease category vectors y. i The disease vector y i Each of the chronic diseases corresponds one-to-one, 1 ≤ index i ≤ N; the disease vector y i Includes 4 predicted probabilities y i,j , 1 ≤ index j ≤ 4; 4 predicted probabilities y i,j Each stage corresponds to one of four tiers: health stage, first-level stage, second-level stage, and third-level stage. Based on prior knowledge of first-order blocking, second-order intervention, and third-order control strategies for various chronic diseases, a disease strategy map and strategy knowledge base are constructed. The first dataset was constructed through data collection; The chronic disease classification prediction model is trained based on the first dataset; After the model training is completed, the individual information set X input by any user is sent to the chronic disease hierarchical prediction model to obtain the corresponding prediction vector Y; and based on the prediction vector Y, the disease strategy map and the strategy knowledge base, strategy identifier retrieval is performed, and based on the retrieval results, a strategy report is synthesized to obtain the corresponding response strategy report and fed back to the current user.

2. The processing method for predicting tiered coping strategies for chronic diseases according to claim 1, characterized in that, The individual information set X includes basic information, measurement information, allergy information, surgical information, women's reproductive health information, medical history information, lifestyle information, and observation information; The basic information includes age, birth weight, and gender type; The measurement information includes height, weight, BMI, waist circumference, hip circumference, neck circumference, calf circumference, grip strength, blood pressure, heart rate, blood oxygen saturation, body temperature, body fat percentage, skeletal muscle mass, and body water percentage. The allergy information includes four types of allergy markers: no allergy marker, drug allergy marker, food allergy marker, and other allergy markers. Each of the four markers includes two states: yes and no. When the no allergy marker is yes, the other three markers are no. When the no allergy marker is no, some or all of the other three markers are yes. The surgical information includes a surgical marker and the interval between the most recent surgeries. The surgical marker has two states: yes and no. When the surgical marker is no, the most recent surgery time is empty. When the surgical marker is yes, the interval between the most recent surgery time is the interval between the most recent surgery time and the current time. The female reproductive health information includes a menstrual regularity marker, menstrual cycle marker, pregnancy marker, childbirth marker, and menopause marker; the menstrual regularity marker, the pregnancy marker, the childbirth marker, the childbirth marker, and the menopause marker all have two states: yes and no; when the menstrual regularity marker is no, the menstrual cycle is empty; when the gender type is male, all markers in the female reproductive health information are no. The medical history information includes a history of metabolic diseases, cardiovascular and cerebrovascular diseases, respiratory diseases, digestive diseases, urinary diseases, endocrine diseases, reproductive diseases, musculoskeletal diseases, nervous diseases, immune system diseases, circulatory system diseases, and mental and psychological diseases. Each type of medical history includes one or more corresponding disease information. Each disease information consists of a corresponding disease name, disease marker, frequency type of onset, and disease control status. The disease marker includes two states: yes and no. The frequency type of onset includes asymptomatic, occasional, frequent, and persistent. The disease control status includes meeting the target, partially meeting the target, not meeting the target, and not being monitored. The lifestyle information includes smoking habits, drinking habits, air conditioning usage habits, exercise habits, sleep quality, mental stress, drinking habits, dietary habits, three meals a day, monthly average food intake, and daily average food intake. The smoking habit information includes a smoking marker and smoking duration; the smoking marker includes two states: yes and no; when the smoking marker is no, the smoking duration is empty. The drinking habit information includes a drinking flag and the age of drinking; the drinking flag includes two states: yes and no; when the drinking flag is no, the age of drinking is empty. The air conditioner usage habit information includes air conditioner usage markers, air conditioner usage time period types, and air conditioner usage years; the air conditioner usage markers include two states: yes and no; the air conditioner usage time period types include year-round, summer, and winter; The exercise habit information includes exercise habit markers, exercise cycle, exercise type, and duration of a single exercise session; the exercise habit markers include two states: yes and no; when the exercise habit marker is no, the exercise cycle, the exercise type, and the duration of a single exercise session are all empty; The sleep quality information includes sleep quality type and daily sleep duration; the sleep quality type includes very good, good, average, poor, and very poor. The mental stress information includes a mental stress marker and a stressor type; the mental stress marker includes two states: yes and no; the stressor type includes work, family, interpersonal relationships, social environment, and other; when the mental stress marker is no, the stressor type is empty; The drinking water habit information includes water source type, drinking water temperature, and daily water consumption; the water source type includes urban tap water, bottled water, bottled water, purified water, rural tap water, and others; The dietary habit information includes dining behavior type, average daily number of takeout meals, average weekly number of takeout meals, takeout meal type, vegetarian preference marker, low-fat preference marker, and low-sugar preference marker; the dining behavior type includes cooking at home, ordering takeout, dining at restaurants, and dining in the cafeteria; the takeout meal type includes Chinese food, Western food, light meals and salads, barbecue, desserts and beverages, and late-night snacks; the vegetarian preference marker, the low-fat preference marker, and the low-sugar preference marker each include two states: yes and no; The information on meal habits includes breakfast habit type, lunch habit type, dinner habit type, and late-night snack habit type; each habit type includes three categories: eat, eat occasionally, and never eat. The monthly average food information includes the monthly average intake type of multiple specific food categories; the monthly average intake type of each specific food category includes almost no food, 1-2 times per week, 3-4 times per week, 5-6 times per week, and daily; the multiple specific food categories include grains, vegetables, fruits, meat and poultry, aquatic products, eggs, dairy products, soy products, fried foods, pickled foods, desserts, and beverages. The daily food information includes staple food type, dish type, meal portion type, and dish portion type; the staple food type includes rice and noodles; the dish type includes vegetables as the main component, meat and eggs as the main component, and a balanced mix of vegetables and meat; the meal portion type includes small, moderate, and excessive; the dish portion type includes small, moderate, and excessive. The observation information includes observations of mental appearance, body shape and posture, gait and movement, communication and interaction, complexion and skin color, eye condition, lip condition, skin condition, hair condition, nail condition, breathing condition, voice condition, temperature sensitivity, sweat condition, tongue condition, tongue coating condition, stool condition, and urination condition. The mental appearance observation consists of multiple selectable categories of mental appearance states; the body shape and posture observation consists of multiple selectable categories of body shape features; the gait and movement observation consists of multiple selectable categories of gait features; the communication and interaction observation consists of multiple selectable categories of communication ability features; the complexion and skin color observation consists of multiple selectable categories of complexion and skin color states; the eye condition observation consists of multiple selectable categories of eye features; and the lip condition observation consists of multiple selectable categories of lip features. The system comprises: skin condition observation (consisting of multiple selectable skin texture features); hair condition observation (consisting of multiple selectable hair features); nail condition observation (consisting of multiple selectable nail color and texture features); respiratory condition observation (consisting of multiple selectable respiratory function features); voice condition observation (consisting of multiple selectable volume and tone quality features); cold / heat condition observation (consisting of multiple selectable cold / heat manifestations); sweat condition observation (consisting of multiple selectable sweating features); tongue condition observation (consisting of multiple selectable tongue color and shape features); tongue coating condition observation (consisting of multiple selectable tongue coating color, texture, and dryness / wetness features); stool condition observation (consisting of multiple selectable stool shape, color, and frequency features); and urination condition observation (consisting of multiple selectable urination shape, color, and frequency features). The disease strategy map includes a first node set and a first edge set; The first node set includes multiple first nodes; the node attributes of the first nodes include node identifier, node type, and node characteristics; the node types include disease nodes, blocking strategy nodes, intervention strategy nodes, and control strategy nodes; when the node type is a disease node, the node characteristic is the disease name; when the node type is a blocking strategy node, intervention strategy node, or control strategy node, the node characteristic is the corresponding first-stage blocking strategy identifier, second-stage intervention strategy identifier, or third-stage control strategy identifier; each first node whose node type is a disease node corresponds to a type of chronic disease; each first node whose node type is a blocking strategy node, intervention strategy node, or control strategy node corresponds to a first-order blocking strategy, second-order intervention strategy, or third-order control strategy for a type of chronic disease. The first edge set includes multiple first edges; the edge attributes of the first edge include edge identifier, parent node identifier, child node identifier, and edge association relationship; the node type of the first node corresponding to the parent node identifier of each first edge is a disease node, the node type of the first node corresponding to the child node identifier is a blocking strategy node, an intervention strategy node, or a control strategy node, and the edge association relationship is a one-stage association, a two-stage association, or a three-stage association corresponding to the child node identifier; The strategy knowledge base includes multiple first strategy records; each first strategy record corresponds to a first-order blocking strategy, a second-order intervention strategy, or a third-order control strategy for a class of chronic diseases; the first strategy record includes a first strategy identifier and a first strategy description; the first strategy description is used to explain the specific strategy execution details of the current strategy; The first dataset includes multiple first data records; the first data record includes the individual information set X and the label vector Y. * The label vector Y * Includes N disease vectors The disease vector Each of the chronic diseases corresponds one-to-one; each disease vector Includes 4 label probabilities ; 4 label probabilities Each of the four hierarchical types corresponds one-to-one; each disease vector The four label probabilities There is only one 1 and the other 3 are 0.

3. The processing method for predicting tiered coping strategies for chronic diseases according to claim 1, characterized in that, The input end of the chronic disease classification prediction model is used to receive the individual information set X, and the output end is used to output the corresponding prediction vector Y. The chronic disease classification prediction model includes an embedding coding layer, a first linear activation layer, a second linear activation layer, a third linear activation layer, N i-th prediction networks, and an output layer; the N i-th prediction networks correspond one-to-one with the N types of chronic diseases. The input of the embedded coding layer is connected to the input of the model, and its output is connected to the input of the first linear activation layer; the output of the first linear activation layer is connected to the input of the second linear activation layer; the output of the second linear activation layer is connected to the input of the third linear activation layer; the output of the third linear activation layer is connected to the inputs of N i-th prediction networks respectively; the outputs of the N i-th prediction networks are connected to the N inputs corresponding to the output layer respectively; the output of the output layer is connected to the output of the model. The embedding coding layer performs embedding coding on each data item of the individual information set X according to a preset embedding coding rule, and sequentially concatenates all codes and coding vectors to obtain the corresponding embedding vector E, which is then sent to the first linear activation layer. The embedding coding rule specifies that: z-core normalization coding is used for numerical data; one-hot coding is used for binary data; one-hot encoding vectors are used for multi-class data (binary or higher classification but only single selection is allowed); multi-hot encoding vectors are used for multi-class data (binary or higher classification and multiple selection is allowed); and word embedding coding is used for text data. The embedding vector E includes multiple codes e. k 1≤indexk≤L E L E The total number of codes in the embedding vector E; the vector shape of the embedding vector E is 1×L E ; The first linear activation layer is used to perform feature encoding on the embedding vector E to obtain the corresponding feature vector H1, which is then sent to the second linear activation layer. The encoding method of the feature vector H1 is as follows: ； ReLU() is the ReLU activation function; W1 and b1 are the weight matrix and bias vector of the first linear activation layer; the shape of the weight matrix W1 is L. E ×D1, the shape of the bias vector b1 is 1×D1, the shape of the feature vector H1 is 1×D1, and D1 is the preset first feature dimension; The second linear activation layer is used to encode the feature vector H1 to obtain the corresponding feature vector H2, which is then sent to the third linear activation layer. The encoding method of the feature vector H2 is as follows: ； W2 and b2 are the weight matrix and bias vector of the second linear activation layer; the weight matrix W2 has a shape of D1×D2, the bias vector b2 has a shape of 1×D2, and the feature vector H2 has a shape of 1×D2, where D2 is a preset second feature dimension. The third linear activation layer is used to encode the feature vector H2 to obtain the corresponding feature vector H3, which is then sent to the N i-th prediction networks. The encoding method of the feature vector H3 is as follows: ； W3 and b3 are the weight matrix and bias vector of the third linear activation layer; the weight matrix W3 has a shape of D2×D3, the bias vector b3 has a shape of 1×D3, and the feature vector H3 has a shape of 1×D3, where D3 is a preset third feature dimension. Each of the i-th prediction networks is used to perform feature vector transformation on the feature vector H3 in the i-th classification feature space to obtain the corresponding feature vector S. i Send to the output layer; Wherein, each of the aforementioned feature vectors S i The vector transformation method is as follows: ； , Let the first and second weight matrices of the i-th prediction network be... , The first and second bias vectors of the i-th prediction network; the first weight matrix. The shape is D3×D4, and the first bias vector The shape is 1×D4, where D4 is the preset fourth feature dimension; the second weight matrix The shape is D4×D5, and the second bias vector The shape is 1×D5, where D5 is the preset fifth feature dimension, and D5=4; each of the feature vectors S i The shape is 1×4, consisting of 4 corresponding feature data s i,j composition; The output layer is used to process the various feature vectors S i Substitute into the Softmax function to calculate the corresponding disease vector y i ; and the obtained N disease vectors y i The corresponding prediction vector Y is then constructed and output.

4. The processing method for predicting tiered coping strategies for chronic diseases according to claim 2, characterized in that, The construction of the first dataset through data collection specifically includes: Step 41: Based on the aforementioned N types of chronic diseases, recruit multiple volunteers to form a volunteer group; and recruit multiple medical experts to form an expert group; The volunteer group includes healthy volunteers and chronic disease volunteers; each chronic disease volunteer suffers from one or more of the N types of chronic diseases; the set of chronic disease types of all chronic disease volunteers includes the N types of chronic diseases; the expert group includes at least several specialists or general practitioners in the fields of Western medicine, traditional Chinese medicine, and integrated traditional Chinese and Western medicine. Step 42: Each volunteer in the volunteer group is designated as the current volunteer; with authorization from the current volunteer or their guardian, data is collected from the individual information set X of the current volunteer; the expert group then identifies whether the current volunteer suffers from a chronic disease through expert consultation, and if a chronic disease is confirmed, further identifies the progression of the chronic disease, and sets the corresponding label vector Y based on the identification results. * ; and comprised of the individual information set X and the label vector Y * Form the corresponding first data record; Step 43: The first dataset is composed of all the first data records obtained.

5. The processing method for predicting tiered coping strategies for chronic diseases according to claim 2, characterized in that, The step of training the chronic disease classification prediction model based on the first dataset specifically includes: Step 51: Based on a preset first segmentation ratio, the first dataset is randomly divided into two sub-datasets, denoted as the first training set and the first evaluation set. The first training set and the first evaluation set are each composed of multiple first data records; the total number of records in the first training set and the first evaluation set is denoted as N. tr N av The ratio of the total number of records in the first training set to the total number of records in the first evaluation set is N. tr :N av Satisfying the first segmentation ratio; the label vector Y of each of the first data records in the first training set. * Record as the corresponding The label vector Y * Corresponding label probability Record as the corresponding 1 ≤ index u ≤ N tr ; Step 52: Input the individual information set X of each of the first data records in the first training set into the chronic disease hierarchical prediction model for processing, and record the prediction vector Y obtained in this processing as the corresponding Y. u And the prediction probabilities y of the prediction vector Y. i,j Let y be the corresponding u,i,j And the prediction vector Y obtained this time u Its corresponding label vector Form the corresponding first prediction-label pair; Step 53, obtain N tr The first prediction-label is substituted into the preset first model loss function L1 to calculate the corresponding first loss value; Specifically, the loss function L1 of the first model is: ； Step 54: Identify whether the first loss value meets the preset first loss value range; if it does, proceed to step 55; if it does not, perform a round of modulation on the model parameters of the first prediction model based on the preset first model optimizer in the direction of minimizing the first model loss function L1, and return to step 52 when the current round of modulation ends. The first model optimizer includes the Adam optimizer and the SGD optimizer. Step 55: Input the individual information set X of each of the first data records in the first evaluation set into the chronic disease classification prediction model for processing, and then use the prediction vector Y obtained from this processing and its corresponding label vector Y * Form the corresponding second prediction-label pair; and based on the obtained N av The second prediction-label pair is evaluated for accuracy, precision, recall, and F1 score to obtain the corresponding first accuracy, first precision, first recall, and first F1 score; Step 56: Identify whether the first accuracy, first precision, first recall, and first F1 score each satisfy their respective first accuracy range, first precision range, first recall range, and first F1 score range; if not, return to step 51; if yes, stop training and confirm that the model training is complete.

6. The processing method for predicting tiered coping strategies for chronic diseases according to claim 2, characterized in that, The process of retrieving strategy identifiers based on the prediction vector Y, the disease strategy atlas, and the strategy knowledge base, and synthesizing a corresponding response strategy report based on the retrieval results to provide feedback to the current user specifically includes: The prediction vector Y is composed of each of the disease vectors y. i As the current vector; and when the tier type corresponding to the current vector with the highest probability is not the healthy stage, the chronic disease corresponding to the current vector is taken as the corresponding first disease, the current tier type is taken as the corresponding first stage, the edge association relationship corresponding to the first stage is taken as the corresponding first relationship, and the first node in the disease strategy graph whose node type is a disease node and whose node characteristics match the first disease is taken as the current node, and each first node whose edge association relationship with the current node matches the first relationship is taken as the corresponding retrieval node, and the node characteristics of each retrieval node are taken as the corresponding first identifier, and all the first identifiers corresponding to the current vector form the corresponding first identifier set; and all the obtained first identifier sets are merged and deduplicated to obtain the corresponding second identifier set; Each of the first identifiers in the second identifier set is taken as the current identifier; the first strategy description of the first strategy record in the strategy knowledge base that matches the current identifier is extracted as the corresponding second strategy description; and all the obtained second strategy descriptions are used to form a corresponding response strategy set. Each of the first diseases and its corresponding first stage constitutes a corresponding disease stage information; and all the obtained disease stage information constitutes a corresponding disease course information set; The corresponding response strategy report, composed of the disease course information set and the response strategy set, is then fed back to the current user.

7. An apparatus for performing the processing method for predicting tiered coping strategies for chronic diseases as described in any one of claims 1-6, characterized in that, The device includes: a model building module, a knowledge graph building module, a data acquisition module, a model training module, and a model application module; The model building module is used to construct a chronic disease hierarchical prediction model for predicting the third-order progression state of N types of chronic diseases based on individual information; the total number of chronic diseases N is a preset positive integer, and the N types of chronic diseases consist of N preset chronic disease categories; the chronic disease hierarchical prediction model is used to predict the third-order progression state of the N types of chronic diseases based on the individual information set X input to the model and output the corresponding prediction vector Y; the prediction vector Y includes N disease category vectors y. i The disease vector y i Each of the chronic diseases corresponds one-to-one, 1 ≤ index i ≤ N; the disease vector y i Includes 4 predicted probabilities y i,j , 1 ≤ index j ≤ 4; 4 predicted probabilities y i,j Each stage corresponds to one of four tiers: health stage, first-level stage, second-level stage, and third-level stage. The knowledge graph construction module constructs a disease strategy graph and strategy knowledge base based on the prior knowledge of first-order blocking, second-order intervention and third-order control strategies for various chronic diseases. The data acquisition module is used to construct a first dataset through data acquisition; The model training module trains the chronic disease classification prediction model based on the first dataset; The model application module is used to input the individual information set X input by any user into the chronic disease hierarchical prediction model after the model training is completed to predict the corresponding prediction vector Y; and to perform strategy identification retrieval based on the prediction vector Y, the disease strategy map and the strategy knowledge base, and synthesize a corresponding response strategy report based on the retrieval results to provide feedback to the current user.

8. An electronic device, characterized in that, include: Memory, processor, and transceiver; The processor is configured to be coupled to the memory, read and execute instructions in the memory to implement the method according to any one of claims 1-6; The transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a computer, cause the computer to perform the method according to any one of claims 1-6.

Images