Data visualization management system

Abstract

The invention relates to the technical field of data management, and provides a data visualization management system. The system comprises a data receiving end used for receiving patient data uploaded by patients in a hospital or patients outside the hospital; the data classification module is used for receiving the patient data and classifying the patient data to establish a patient data set; the processing module calls a trained auxiliary diagnosis and treatment model to output a diagnosis and treatment recommendation scheme according to the patient data set; and the display end displays the diagnosis and treatment recommendation scheme. According to the system, patient data is classified to establish a patient data set, a trained auxiliary diagnosis and treatment model is called based on the patient data set to output a diagnosis and treatment recommendation scheme, the diagnosis and treatment recommendation scheme is displayed, massive patient data is accurately recognized through a data visualization management system, doctors are helped to rapidly determine a diabetes treatment scheme, and the diagnosis and treatment efficiency is improved. The precise adjuvant therapy of diabetes mellitus is improved.

Description

Technical Field

[0001] The present application relates to the technical field of data management, and in particular to a data visualization management system. Background Art

[0002] The existing disease diagnosis and treatment method generally requires patients to go to the hospital for consultation, and the doctor will diagnose the patient based on the patient's admission situation. If necessary, the patient will undergo various physical examinations, and then the doctor will issue a disease diagnosis and treatment plan. Traditional in-hospital management is limited to peripheral links such as hospital registration, billing, and medical office. Even in some hospitals, computers are used to manage diabetes, and only the electronic medical record system is used to record the basic information of diabetic patients. There is no scientific and systematic management flow to achieve accurate management of different symptoms. Traditional out-of-hospital management relies on the patient's own management ability. However, most patients' blood sugar self-management is at a low level, blood sugar monitoring is not timely, and there is a lack of relevant guidance, which makes it difficult for patients with blood sugar at home to control blood sugar, which in turn causes various acute and chronic complications.

[0003] Due to the large number of diabetic patients and the limited number of medical resources such as hospitals and doctors, the current management of diabetic patients is chaotic. Doctors cannot accurately identify patients' symptoms and conduct targeted diagnosis and treatment based on the data uploaded by patients inside and outside the hospital, which affects the treatment effect of diabetes. Summary of the invention

[0004] In view of this, an embodiment of the present application provides a data visualization management system, which improves data management capabilities, accurately identifies massive amounts of patient data, helps doctors quickly determine diabetes treatment plans, and improves auxiliary treatments for diabetes.

[0005] The present application embodiment provides a data visualization management system, including:

[0006] The data receiving end receives patient data uploaded by in-hospital or out-of-hospital patients;

[0007] A data classification module receives patient data and classifies the patient data to establish a patient data set;

[0008] The processing module calls the trained auxiliary diagnosis and treatment model according to the patient data set to output the diagnosis and treatment recommendation plan;

[0009] The display terminal shows the recommended diagnosis and treatment plan.

[0010] The data visualization management system provided in the embodiment of the present application includes a data receiving end, a data classification module, a processing module and a display end. The data visualization management system receives patient data uploaded by the patient, classifies and processes the patient data through the data classification module, thereby establishing a patient data set; then, the processing module calls the trained auxiliary diagnosis and treatment model based on the patient data set to output a diagnosis and treatment recommendation; finally, the diagnosis and treatment recommendation is displayed through the display end. The system establishes a patient data set by classifying the patient data, calls the trained auxiliary diagnosis and treatment model based on the patient data set to output a diagnosis and treatment recommendation, and displays the diagnosis and treatment recommendation through the display end, so as to help doctors quickly determine the diabetes treatment plan and improve the auxiliary treatment of diabetes. Among them, the data receiving end can receive patient data of in-hospital patients and / or out-of-hospital patients, and these patient data include basic data of the patient, including but not limited to age, gender, occupation, height, weight, waist circumference and allergy history; health data, including but not limited to family history of illness, chronic disease status and blood pressure level; behavioral data, including but not limited to smoking, drinking, diet, sleep and exercise. Clinical data, including but not limited to serum examination clinical indicator data, urine routine examination clinical indicator data and electrocardiogram clinical indicator data. Monitoring data, including but not limited to daily blood sugar changes, trends, insulin doses, and carbohydrate intake. Treatment data, including but not limited to disease classification, disease course, age, gender, dietary habits, blood sugar, blood pressure, complications, and medication use, etc.; also includes laboratory data and medication data. The data classification module receives massive amounts of patient data, performs data classification, and improves system computing efficiency.

[0011] In one embodiment of the present application, the patient data includes treatment data, and the data classification module obtains a treatment plan feature identifier by parsing the treatment data, determines the patient category according to the treatment plan feature identifier, and classifies the patient data according to the patient category to obtain a patient data set.

[0012] In one embodiment of the present application, the patient data set includes a current blood glucose value, and the processing module includes:

[0013] The blood sugar risk level identification submodule determines the patient's blood sugar risk level based on the current blood sugar value;

[0014] The model building submodule builds and trains auxiliary diagnosis and treatment models based on patient data sets;

[0015] The model calling submodule calls the auxiliary diagnosis and treatment model according to the patient data and the patient's blood sugar risk level;

[0016] The recommended solution output submodule obtains the recommended treatment solution output by the auxiliary diagnosis and treatment model.

[0017] In one embodiment of the present application, the blood sugar risk level identification submodule includes:

[0018] The blood glucose risk level identification unit calls the trained blood glucose prediction model according to the current blood glucose value to output the patient's blood glucose risk level.

[0019] In one embodiment of the present application, the blood sugar risk level identification submodule further includes:

[0020] The blood glucose prediction model training unit uses the processed blood glucose data of the patient to train the blood glucose prediction model, wherein different blood glucose prediction models are obtained by training blood glucose data under different modes.

[0021] In one embodiment of the present application, the system further includes a blood glucose monitoring device, which visually displays the current blood glucose value.

[0022] In one embodiment of the present application, the system also includes a data acquisition terminal, which collects the patient's exercise data, diet data and heart rate data, and uploads the exercise data, diet data and heart rate data to a data receiving terminal.

[0023] In one embodiment of the present application, the characteristic fields of the patient data set include gender, age, height, weight, waist circumference, eating habits, exercise status, heart rate, course of disease, complications and drug dosage; the auxiliary diagnosis and treatment model includes a treatment method diagnosis and treatment model; the diagnosis and treatment recommendation plan includes a target treatment method; the treatment method diagnosis and treatment model is expressed as:

[0024] O s1 =f s1 (z1)

[0025] Among them, O s1 represents the output target treatment method, z1 represents the input feature field, and f s1 It is a machine learning classification model, which represents the treatment method diagnosis model.

[0026] In another embodiment of the present application, the characteristic fields of the patient data set include disease classification, disease course, age, gender, eating habits, blood sugar, blood pressure, complications and drug use; the auxiliary diagnosis and treatment model includes a treatment plan diagnosis and treatment model; the diagnosis and treatment recommendation plan includes a target treatment plan; the treatment plan diagnosis and treatment model is expressed as:

[0027] O s2 =f s2 (z2)

[0028] Among them, O s2 represents the output target treatment plan, z2 represents the input feature field, and f s2 It is a machine learning classification model, representing the treatment plan diagnosis model.

[0029] In another embodiment of the present application, the characteristic fields of the patient data set include disease classification, disease course, age, gender, eating habits, blood sugar, blood pressure, complications, past medical history, drug use and lipoprotein status; the auxiliary diagnosis and treatment model includes a risk level diagnosis and treatment model; the diagnosis and treatment recommendation plan includes the patient risk level; the risk level diagnosis and treatment model is expressed as:

[0030] O s3 =f s3 (z3)

[0031] Among them, O s3 represents the output patient risk level, z3 represents the input feature field, and f s3 It is a machine learning classification model, representing the risk level diagnosis and treatment model. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is a schematic diagram of a data visualization management system provided in an embodiment of the present application;

[0033] Figure 2 It is a schematic diagram of the interaction between the data visualization management system provided in the embodiment of the present application and the patient client and the doctor client;

[0034] Figure 3 It is a schematic diagram of a treatment method diagnosis and treatment model provided in an embodiment of the present application;

[0035] Figure 4 It is a schematic diagram of a treatment plan diagnosis and treatment model provided in an embodiment of the present application;

[0036] Figure 5 is a schematic diagram of a risk level diagnosis and treatment model provided in an embodiment of the present application;

[0037] Figure 6 is a working schematic diagram of the unified management platform for diabetes treatment provided by an embodiment of the present application;

[0038] Figure 7 It is a flowchart of obtaining a diagnosis and treatment recommendation plan through an intelligent algorithm provided in an embodiment of the present application;

[0039] Figure 8 It is a working diagram of the off-site unified management platform provided in the embodiment of the present application. DETAILED DESCRIPTION

[0040] In the following description, for the purpose of illustration rather than limitation, specific details such as specific system structures, technologies, etc. are proposed, so as to thoroughly understand the embodiments of the present application. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details. In other cases, the detailed description of well-known systems, devices, circuits and methods is omitted to prevent unnecessary details from hindering the description of the present application. In addition, in the description of the present application specification and the attached claims, the terms "first", "second", "third" etc. are only used to distinguish the description, and cannot be interpreted as indicating or suggesting relative importance.

[0041] The existing disease diagnosis and treatment method is generally that the patient goes to the hospital for consultation, and the doctor diagnoses and issues a disease diagnosis and treatment plan based on the patient's admission situation. However, due to the limited number of medical resources, the efficiency of obtaining disease diagnosis and treatment plans is currently low. In response to the problems existing in the prior art, the embodiment of the present application provides a data visualization management system that improves data management capabilities, accurately identifies massive patient data, helps doctors quickly determine diabetes treatment plans, and improves auxiliary treatments for diabetes.

[0042] like Figure 1 FIG. 1 is a schematic diagram of a data visualization management system provided by an embodiment of the present application. Figure 1 In the data visualization management system, the data receiving end, data classification module, processing module and display end are included. Among them, the data receiving end is mainly used to receive patient data uploaded by in-hospital patients or out-of-hospital patients; the data classification module is mainly used to receive patient data and classify and process the patient data to establish a patient data set; the processing module is mainly used to call the trained auxiliary diagnosis and treatment model according to the patient data set to output the diagnosis and treatment recommendation plan; the display end is mainly used to display the diagnosis and treatment recommendation plan. For the specific working principle of the data visualization management system, please refer to the following description.

[0043] like Figure 2 The figure shows a schematic diagram of the interaction between the data visualization management system provided by the embodiment of the present application and the patient client and the doctor client. In actual operation, the data visualization management system can be deployed on the server side, and the server side can establish communication connections with the patient client and the doctor client respectively through the network or the like. For example, the patient can connect to the server side through the patient client and upload the patient data to the data visualization management system; the doctor can obtain the diagnosis and treatment recommendation plan output by the data visualization management system through the doctor client.

[0044] The following is a detailed description of how the data visualization management system works:

[0045] The data visualization management system is a unified data management platform for both in-hospital and out-of-hospital patients. Both in-hospital and out-of-hospital patients can upload their own patient data to the data visualization management system through the patient client, and the data visualization management system receives the patient data uploaded by each patient through the data receiving terminal.

[0046] As an example, the patient data uploaded by in-hospital patients may include but are not limited to: the patient's personal data, treatment data and laboratory test data, etc. Among them, personal data may include but are not limited to: gender, age, height, weight, waist circumference, eating habits, exercise conditions, course of disease, complications and glycation data. Treatment data may include but are not limited to: blood sugar monitoring data, finger blood data, high dose, basal rate and drug data. Laboratory test data mainly includes physical test data obtained by in-hospital patients after laboratory examinations in the hospital.

[0047] As another example, the patient data uploaded by outpatients may include but are not limited to: the patient's personal data, treatment data, exercise data, diet data, heart rate data, etc. Among them, exercise data can be uploaded through external devices such as smart watches or bracelets worn by patients; diet data can be taken and manually uploaded using a dedicated APP installed on a mobile terminal such as a mobile phone used by the patient; treatment data can also be recorded and uploaded daily using a dedicated APP installed on the patient's mobile terminal, mainly including treatment methods, blood sugar monitoring data, finger blood data, high dose, basal rate, and drug data.

[0048] In one embodiment of the present application, the data visualization management system also includes a data acquisition terminal, which collects the patient's exercise data, diet data and heart rate data, and uploads the exercise data, diet data and heart rate data to a data receiving terminal.

[0049] The data visualization management system itself can also be equipped with a corresponding data collection terminal, and various data collection devices can be set up at the data collection terminal to collect the exercise data, diet data and heart rate data of each patient, and upload these data to the data receiving terminal of the data visualization management system.

[0050] After receiving the patient data uploaded by the patient through the data receiving end, the data visualization management system can classify the patient data through the data classification module to establish a corresponding patient data set. Specifically, the patient data can be parsed to identify the various feature fields contained therein, and the patient data can be classified according to the feature fields to establish a patient data set.

[0051] As an example, the patient data includes treatment data, and the data classification module obtains a treatment plan feature identifier by parsing the treatment data, determines the patient category according to the treatment plan feature identifier, and classifies the patient data according to the patient category to obtain a patient data set. In one embodiment, the treatment data uploaded by the patient only includes the drug category, drug dosage, and drug use time. By obtaining the treatment plan based on the treatment data uploaded by the patient, it can be clarified whether the patient is only undergoing simple drug treatment.

[0052] After the data visualization management system establishes the corresponding patient data set through the data classification module, it can use the patient data set through the processing module to call the trained auxiliary diagnosis and treatment model to output the diagnosis and treatment recommendation plan. The data visualization management system can use a combination of machine learning, artificial intelligence or data mining to analyze the patient data set and generate corresponding treatment methods, treatment plans or patient risk levels and other diagnosis and treatment recommendations.

[0053] In one embodiment of the present application, the obtained patient data set includes the current blood glucose value, and the processing module of the data visualization management system includes a blood glucose risk level identification submodule, a model building submodule, a model calling submodule and a recommendation scheme output submodule.

[0054] Among them, the blood sugar risk level identification submodule is mainly used to determine the patient's blood sugar risk level according to the current blood sugar value. For example, if the patient's current blood sugar value is very low, it means that the patient is at risk of hypoglycemia. According to the degree of danger of hypoglycemia, the corresponding patient's blood sugar risk level can be determined; if the patient's current blood sugar value is very high, it means that the patient is at risk of hyperglycemia. According to the degree of danger of hyperglycemia, the corresponding patient's blood sugar risk level can be determined.

[0055] According to the diabetes guidelines, the relationship between the patient's current blood sugar level and blood sugar risk is shown in Table 1 below:

[0056] Table 1

[0057] Risk of very low blood sugar Risk of hypoglycemia normal Risk of high blood sugar Risk of very high blood sugar BG<50 50≤BG<70 70≤BG<180 180≤BG<300 300≤BG TBR>1% TBR>4% TIR>70% TAR>20% TAR>5%

[0058] In Table 1, BG represents the current blood glucose value, that is, the blood glucose concentration, TIR represents the proportion of time when blood glucose is within the normal range, TBR represents the proportion of time when blood glucose is below the normal range, and TAR represents the proportion of time when blood glucose is above the normal range.

[0059] In actual operation, each blood sugar risk can be graded according to Table 2:

[0060] Table 2

[0061] type grade normal R0 Risk of very low blood sugar R1 Risk of hypoglycemia R2 Risk of high blood sugar R3 Risk of very high blood sugar R4

[0062] Next, according to the classification in Table 2 and considering the actual degree of danger of each blood sugar risk, the corresponding level of each blood sugar risk can be divided according to Table 3, that is, the blood sugar risk level:

[0063] Table 3

[0064] grade Blood sugar risk level R1 and R4 High risk R2 Medium risk R3 Lower risk R0 Low risk

[0065] In one embodiment of the present application, the blood sugar risk level identification submodule includes:

[0066] The blood glucose risk level identification unit calls the trained blood glucose prediction model according to the current blood glucose value to output the patient's blood glucose risk level.

[0067] The data visualization management system can collect a large number of patients' blood sugar data as samples in advance and train the corresponding blood sugar prediction models. After obtaining the patient's current blood sugar value, these blood sugar prediction models can be called to analyze the current blood sugar value to obtain the corresponding patient blood sugar risk level, such as the various blood sugar risk levels in Table 3 above.

[0068] In one embodiment of the present application, the blood sugar risk level identification submodule further includes:

[0069] The blood glucose prediction model training unit uses the processed blood glucose data of the patient to train the blood glucose prediction model, wherein different blood glucose prediction models are obtained by training blood glucose data under different modes.

[0070] The data visualization management system collects a large amount of different patients' own blood sugar data in advance. After preprocessing these blood sugar data, these blood sugar data can be classified according to different modes, and different blood sugar prediction models can be trained using the blood sugar data in different modes. This can improve the prediction accuracy of the blood sugar prediction model to a certain extent. For example, blood sugar data can be classified according to high blood sugar concentration and low blood sugar concentration, and a large amount of blood sugar data with high blood sugar concentration can be used to train a blood sugar prediction model specifically for predicting high blood sugar, and a large amount of blood sugar data with low blood sugar concentration can be used to train a blood sugar prediction model specifically for predicting low blood sugar. After obtaining the patient's current blood sugar value, if the current blood sugar value is high, the blood sugar prediction model specifically for predicting high blood sugar is called, and if the current blood sugar value is low, the blood sugar prediction model specifically for predicting low blood sugar is called.

[0071] The model building submodule is mainly used to build and train auxiliary diagnosis and treatment models based on patient data sets. For example, a large number of patient data sets with labeled category labels can be used as samples to train a machine learning classification model as an auxiliary diagnosis and treatment model. The auxiliary diagnosis and treatment model will be described in more detail below.

[0072] The model calling submodule is mainly used to call auxiliary diagnosis and treatment models according to patient data and the patient's blood sugar risk level. In actual operation, the trained auxiliary diagnosis and treatment models can include multiple ones. According to the characteristic fields of the patient data and the patient's actual blood sugar risk level, the most suitable auxiliary diagnosis and treatment model can be found from multiple auxiliary diagnosis and treatment models and called.

[0073] The recommended solution output submodule is mainly used to obtain the recommended solution output by the auxiliary diagnosis and treatment model. After obtaining the corresponding recommended solution by using the auxiliary diagnosis and treatment model, the recommended solution can be obtained through the recommended solution output submodule and output to the display end of the data visualization management system or the doctor client for display.

[0074] In actual operation, different auxiliary diagnosis and treatment models can be trained to output different diagnosis and treatment recommendations. For example, for the treatment method of a patient, a treatment method diagnosis and treatment model can be constructed to output the recommended target treatment method for the patient; for the treatment plan of a patient, a treatment plan diagnosis and treatment model can be constructed to output the recommended target treatment plan for the patient; for the risk level of a patient, a risk level diagnosis and treatment model can be constructed to output the corresponding patient risk level.

[0075] In one embodiment of the present application, the characteristic fields of the patient data set include gender, age, height, weight, waist circumference, eating habits, exercise status, heart rate, course of disease, complications and drug dosage; the auxiliary diagnosis and treatment model includes a treatment method diagnosis and treatment model; the diagnosis and treatment recommendation plan includes a target treatment method; the treatment method diagnosis and treatment model is expressed as:

[0076] O s1 =f s1 (z1)

[0077] Among them, O s1 represents the output target treatment method, z1 represents the input feature field, and f s1 It is a machine learning classification model, which represents the treatment method diagnosis model.

[0078] like Figure 3 The figure shows a schematic diagram of the treatment method diagnosis model. The data visualization management system can collect a large amount of sample data with treatment method labels in advance as a training set. These sample data contain various feature fields in the patient data set, and establish the treatment method diagnosis model by training the machine learning classification model. After the feature fields of a patient's patient data set are input into the treatment method diagnosis model, the model analyzes and processes the data to obtain a classification result, and the output patient treatment method can be determined based on the classification result.

[0079] Taking diabetic patients as an example, treatment methods can include the following categories: drugs: drug therapy alone; MDI (combined with drugs): multiple daily injections of mealtime insulin and basal insulin combined with drug therapy; CSII (combined with drugs): continuous subcutaneous insulin infusion combined with drug therapy; SAP (combined with drugs): sensor-enhanced insulin pump combined with drug therapy; APS (combined with drugs): artificial pancreas system combined with drug therapy.

[0080] The characteristic field z1 may include: diabetes type, diabetes course, age, gender, eating habits, blood sugar, current and past history of hypoglycemia risk, BMI, diabetes complications, combined hypoglycemic drugs, insulin secretion function, creatinine-blood, creatinine-urine, alanine aminotransferase, aspartate aminotransferase, lipoprotein status, triglycerides, blood pressure, diabetes symptoms, and past insulin dosages.

[0081] In actual operation, some feature fields and treatment methods can be normalized first to generate corresponding structured data as follows: diabetes type, diabetes course, age, gender, blood sugar, BMI, combined hypoglycemic drugs, creatinine-blood, creatinine-urine, alanine aminotransferase, aspartate aminotransferase, lipoprotein status, triglycerides, blood pressure, past insulin doses and treatment methods. The following parts of the feature fields are used as free text data: current and past history of hypoglycemia risk, diabetic complications, insulin secretion function and diabetic symptoms. Then, a pre-trained natural language processing model (such as the ClinicalBERT model) is used to process the free text data. The processed structured data and free text data are encoded using the one-hot encoding rule to obtain multi-classification labels. Afterwards, the feature field z1 and the treatment method O can be combined into a single-hot encoding model. s1 Substitute the support vector machine model for machine learning training and fit the feature vectors in the model to construct a treatment method diagnosis model.

[0082] In another embodiment of the present application, the characteristic fields of the patient data set include disease classification, disease course, age, gender, eating habits, blood sugar, blood pressure, complications and drug use; the auxiliary diagnosis and treatment model includes a treatment plan diagnosis and treatment model; the diagnosis and treatment recommendation plan includes a target treatment plan; the treatment plan diagnosis and treatment model is expressed as:

[0083] O s2 =f s2 (z2)

[0084] Among them, O s2 represents the output target treatment plan, z2 represents the input feature field, and f s2 It is a machine learning classification model, representing the treatment plan diagnosis model.

[0085] like Figure 4 The figure shows a schematic diagram of the treatment plan diagnosis model. Similarly, the data visualization management system can collect a large amount of sample data with treatment plan labels in advance as a training set. These sample data contain various feature fields in the patient data set, and establish the treatment plan diagnosis model by training the machine learning classification model. After the feature fields of a patient's patient data set are input into the treatment plan diagnosis model, the model analyzes and processes the data to obtain a classification result, and the output patient treatment plan can be determined based on the classification result.

[0086] Taking diabetic patients as an example, the treatment plan may include: basal rate, large dose, medication amount and laboratory test list.

[0087] The characteristic field z2 may include: diabetes classification, diabetes course, age, gender, eating habits, blood sugar, current and past history of hypoglycemia risk, BMI, diabetes complications, combined hypoglycemic drugs, insulin secretion function, creatinine-blood, creatinine-urine, alanine aminotransferase, aspartate aminotransferase, lipoprotein status, triglycerides, blood pressure, diabetes symptoms, treatment methods, and past insulin doses.

[0088] In actual operation, some feature fields can be normalized first to generate corresponding structured data as follows: diabetes type, diabetes course, age, gender, blood sugar, BMI, combined hypoglycemic drugs, creatinine-blood, creatinine-urine, alanine aminotransferase, aspartate aminotransferase, lipoprotein status, triglycerides, blood pressure, past insulin doses and treatment methods. The following parts of the feature fields are used as free text data: current and past history of hypoglycemia risk, diabetic complications, insulin secretion function and diabetic symptoms. Then, the free text data is processed using a pre-trained natural language processing model, and the processed structured data and free text data are encoded using the one-hot encoding rule to obtain multi-classification labels. In addition, a neural network model containing multiple hidden layers (for example, 16 hidden layers) can be constructed, using the ReLU function as the activation function, the mean square error as the loss function in the training process, and the Adam optimizer can be selected as the optimizer for training the neural network. The feature field z2 and the treatment plan O s2 By substituting the neural network model for deep learning training and updating the weights of each parameter in the model, a treatment plan diagnosis model can be constructed.

[0089] In another embodiment of the present application, the characteristic fields of the patient data set include disease classification, disease course, age, gender, eating habits, blood sugar, blood pressure, complications, past medical history, drug use and lipoprotein status; the auxiliary diagnosis and treatment model includes a risk level diagnosis and treatment model; the diagnosis and treatment recommendation plan includes the patient risk level; the risk level diagnosis and treatment model is expressed as:

[0090] O s3 =f s3 (z3)

[0091] Among them, O s3 represents the output patient risk level, z3 represents the input feature field, and f s3 It is a machine learning classification model, representing the risk level diagnosis and treatment model.

[0092] like Figure 5 The figure shows a schematic diagram of the risk level diagnosis and treatment model. Similarly, the data visualization management system can collect a large amount of sample data with risk level labels in advance as a training set. These sample data contain various feature fields in the patient data set, and establish the risk level diagnosis and treatment model by training the machine learning classification model. After the feature fields of a patient's patient data set are input into the risk level diagnosis and treatment model, the model analyzes and processes the data to obtain a classification result, and the output patient risk level can be determined based on the classification result.

[0093] Also taking diabetic patients as an example, the patient risk levels may include: high risk, medium risk, lower risk and low risk.

[0094] The characteristic field z3 may include: blood glucose value, TIR, TBR, TAR, diabetes classification, diabetes course, age, gender, eating habits, blood glucose, current and past history of hypoglycemia risk, BMI, diabetes complications, combined hypoglycemic drugs, insulin secretion function, creatinine-blood, creatinine-urine, alanine aminotransferase, aspartate aminotransferase, lipoprotein status, triglycerides, blood pressure, diabetes symptoms and past insulin dosage and other fields.

[0095] In actual operation, some feature fields and patient risk levels can be normalized first to generate corresponding structured data as follows: diabetes type, diabetes course, age, gender, blood sugar, BMI, combined hypoglycemic drugs, creatinine-blood, creatinine-urine, alanine aminotransferase, aspartate aminotransferase, lipoprotein status, triglycerides, blood pressure, past insulin doses, and patient risk level. The following parts of the feature fields are used as free text data: current and past history of hypoglycemia risk, diabetic complications, insulin secretion function, and diabetic symptoms. Then, the pre-trained natural language processing model is used to process the free text data. The processed structured data and free text data are encoded using the one-hot encoding rule to obtain multi-classification labels. Afterwards, the feature field z3 and the patient risk level O can be combined into a single-hot encoding model. s3 Substitute the support vector machine model for machine learning training and fit the feature vectors in the model to construct a risk level diagnosis and treatment model.

[0096] In one embodiment of the present application, after obtaining the target treatment method, target treatment plan and patient risk level for the current patient, the data visualization management system can determine whether the patient's risk level is too high, such as high risk or medium risk. If the patient's risk level is too high, the data visualization management system can generate corresponding diagnosis and treatment recommendation information based on the target treatment method and target treatment plan, and send the diagnosis and treatment recommendation information to the patient client, thereby reminding and advising the patient.

[0097] As an example, when the data visualization management system generates corresponding diagnosis and treatment recommendation information based on the target treatment method and the target treatment plan, it can perform corresponding text integration operations on the target treatment method and the target treatment plan according to the user's reading habits, thereby obtaining corresponding diagnosis and treatment recommendation information. For example, if a diabetic patient's patient risk level obtained by the risk level diagnosis and treatment model is high risk, the target treatment method obtained by the treatment method diagnosis and treatment model is continuous subcutaneous insulin infusion combined with drug therapy, and the target treatment plan obtained by the treatment plan diagnosis and treatment model is daily infusion of A units of insulin, daily oral diabetes drug X two tablets and daily oral diabetes drug Y one tablet, then the corresponding text integration operation can be performed to obtain the diagnosis and treatment recommendation information "After diagnosis and evaluation, you are a high-risk patient for diabetes, and it is recommended to use continuous subcutaneous insulin infusion combined with drug therapy, daily infusion of A units of insulin, daily oral diabetes drug X two tablets and daily diabetes drug Y one tablet".

[0098] For the vast number of diabetic patients inside and outside the hospital, the data visualization management system provided in the embodiment of the present application can be regarded as a unified management platform for diabetes treatment. The working diagram of the unified management platform is as follows: Figure 6 As shown. Figure 6In the system, diabetic patients inside and outside the hospital can upload their own patient data to the unified management platform, which will pre-analyze and classify the patient data, establish the corresponding diabetic patient data set, and then call the trained auxiliary diagnosis and treatment model according to the diabetic patient data set to output the diagnosis and treatment recommendation plan suitable for the patient, such as insulin treatment plan, etc.

[0099] The unified management platform for diabetes treatment can be divided into two parts: the hospital unified management platform and the hospital unified management platform. When the in-hospital patient goes to the hospital, the doctor can diagnose the patient according to the patient's admission situation and determine the treatment method for the in-hospital patient: simple drugs, MDI (combined drugs), CSII (combined drugs), SAP (combined drugs) and APS (combined drugs), etc. After the in-hospital patient enters the data and performs a physical examination, the corresponding patient information, treatment data and laboratory test order information and other data can be obtained. The doctor can make a diagnosis based on these data, estimate the patient's risk level, and give a specific treatment plan. In addition, the treatment method, patient risk level and treatment plan can also be obtained by intelligent algorithms, for example, the treatment method diagnosis and treatment model, treatment plan diagnosis and treatment model and risk level diagnosis and treatment model described above can be used. The treatment method, patient risk level and treatment plan obtained by the intelligent algorithm belong to the machine-generated diagnosis and treatment recommendation plan, and the doctor can manually evaluate the diagnosis and treatment recommendation plan to decide whether to adopt it.

[0100] like Figure 7 The figure shows a flow chart of obtaining a recommended diagnosis and treatment plan through an intelligent algorithm. Based on the existing treatment plan and the patient data set established based on the uploaded patient data, the intelligent algorithm is used to determine the treatment method, patient risk level and treatment plan recommended to the doctor. The doctor can evaluate the treatment method, patient risk level and treatment plan recommended by the intelligent algorithm to determine whether the plan needs to be modified. For example, after the intelligent algorithm outputs the patient's risk level, the doctor confirms whether the patient has a high risk. If so, the doctor can formulate an emergency plan for the risk. If it is confirmed that there is no risk or the risk is low, the doctor can ignore it. After the intelligent algorithm outputs the recommended treatment plan, the doctor can choose to implement the treatment plan or modify the treatment plan.

[0101] The working diagram of the unified management platform outside the hospital is as follows Figure 8As shown, each out-of-hospital patient can upload various types of patient data, such as exercise data, diet data, treatment data, and patient information, to the out-of-hospital unified management platform through their own patient clients, such as mobile phones, smart watches, smart bracelets and other devices. Through the analysis and judgment of the patient data set by intelligent algorithms, the treatment methods, patient risk levels and treatment plans recommended to doctors are obtained. If the patient's risk level is found to be high, the out-of-hospital unified management platform can send reminders and related suggestions to the patient. Specifically, the out-of-hospital unified management platform can send reminders and related suggestions to the patient client through dedicated APP push, public account push or SMS. Similar to the in-hospital unified management platform, when the intelligent algorithm outputs the recommended treatment plan, the doctor can choose to implement the treatment plan or modify the treatment plan.

[0102] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated in one processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit, and the above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application.

[0103] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0104] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0105] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0106] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0107] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0108] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0109] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A data visualization management system, characterized in that: The system comprises: The data receiving end receives patient data uploaded by in-hospital or out-of-hospital patients; A data classification module receives the patient data and classifies the patient data to establish a patient data set; A processing module, calling a trained auxiliary diagnosis and treatment model according to the patient data set to output a diagnosis and treatment recommendation plan; The display terminal displays the recommended diagnosis and treatment plan.

2. The data visualization management system according to claim 1, characterized in that: The patient data includes treatment data. The data classification module obtains a treatment plan feature identifier by parsing the treatment data, determines a patient category according to the treatment plan feature identifier, and classifies the patient data according to the patient category to obtain the patient data set.

3. The data visualization management system according to claim 2, characterized in that: The patient data set includes a current blood glucose value, and the processing module includes: A blood sugar risk level identification submodule, which determines the patient's blood sugar risk level according to the current blood sugar value; A model building submodule, building and training the auxiliary diagnosis and treatment model according to the patient data set; A model calling submodule, calling the auxiliary diagnosis and treatment model according to the patient data and the patient's blood sugar risk level; The recommended solution output submodule obtains the recommended diagnosis and treatment solution output by the auxiliary diagnosis and treatment model.

4. The data visualization management system according to claim 3, characterized in that: The blood sugar risk level identification submodule includes: The blood glucose risk level identification unit calls a trained blood glucose prediction model according to the current blood glucose value to output the patient's blood glucose risk level.

5. The data visualization management system according to claim 4, characterized in that: The blood sugar risk level identification submodule also includes: The blood glucose prediction model training unit uses the processed blood glucose data of the patient to train the blood glucose prediction model, wherein different blood glucose prediction models are obtained by training blood glucose data under different modes.

6. The data visualization management system according to claim 3, characterized in that: The system further comprises a blood glucose monitoring device, which visually displays the current blood glucose value.

7. The data visualization management system according to any one of claims 1 to 6, characterized in that: The system also includes a data acquisition terminal, which collects the patient's exercise data, diet data and heart rate data, and uploads the exercise data, diet data and heart rate data to the data receiving terminal.

8. The data visualization management system according to claim 7, characterized in that: The characteristic fields of the patient data set include gender, age, height, weight, waist circumference, eating habits, exercise status, heart rate, course of disease, complications and drug dosage; the auxiliary diagnosis and treatment model includes a treatment method diagnosis and treatment model; the diagnosis and treatment recommendation plan includes a target treatment method; the treatment method diagnosis and treatment model is expressed as: O s1 = f s1 (z1) Among them, O s1 represents the output target treatment method, z1 represents the input feature field, f s1 It is a machine learning classification model, which represents the diagnosis and treatment model of the treatment method.

9. The data visualization management system according to claim 7, characterized in that: The characteristic fields of the patient data set include disease classification, disease course, age, gender, eating habits, blood sugar, blood pressure, complications and drug use; the auxiliary diagnosis and treatment model includes a treatment plan diagnosis and treatment model; the diagnosis and treatment recommendation plan includes a target treatment plan; the treatment plan diagnosis and treatment model is expressed as: Oh s2 =f s2 (z2) Among them, O s2 represents the output target treatment plan, z2 represents the input feature field, f s2 It is a machine learning classification model, which represents the treatment plan diagnosis model.

10. The data visualization management system according to claim 7, characterized in that: The characteristic fields of the patient data set include disease classification, disease course, age, gender, eating habits, blood sugar, blood pressure, complications, past medical history, drug use and lipoprotein status; the auxiliary diagnosis and treatment model includes a risk level diagnosis and treatment model; the diagnosis and treatment recommendation plan includes the patient risk level; the risk level diagnosis and treatment model is expressed as: O s3 =f s3 (z3) Among them, O s3 represents the output patient risk level, z3 represents the input feature field, f s3 It is a machine learning classification model, which represents the risk level diagnosis and treatment model.

Images