Intelligent processing system for endocrine disease diagnosis data

Abstract

The invention discloses an intelligent processing system for endocrine disease diagnosis data, which belongs to the technical field of endocrine disease diagnosis data processing and comprises a hospital end, a cloud end and a user end, the cloud comprises a storage module and a reference analysis module; the reserve module is used for performing reserve analysis and establishing a problem reserve library; the benchmark analysis module is used for configuring a baseline model for the user side and configuring a question and answer library for the corresponding user side according to the disease condition data; the hospital end comprises an information gathering module; the information summarization module is used for summarizing information of each patient to obtain hospital detection data, and integrating the hospital detection data and the patient information into disease condition data; the user side comprises an acquisition module and a home analysis module; the acquisition module is used for acquiring home detection data of a user; and the home analysis module is used for performing real-time analysis on the home detection data through a preset baseline model to obtain a baseline risk standard, and analyzing the home detection data according to the baseline risk standard to obtain a home analysis result.

Description

Technical Field

[0001] The present invention belongs to the technical field of endocrine disease diagnosis data processing, and in particular is an endocrine disease diagnosis data intelligent processing system. Background Art

[0002] Primary Aldosteronism (PA) is an endocrine hypertension characterized by excessive autonomic aldosterone secretion. Long-term uncontrolled PA can lead to irreversible target organ damage, including myocardial hypertrophy, renal failure, and severe arrhythmias. Although aldosterone receptor antagonists such as spironolactone can effectively control the disease by inhibiting the effects of aldosterone, PA patients require long-term home monitoring of key physiological parameters such as blood pressure and serum potassium, and medication regimens should be adjusted based on dynamic changes to maintain metabolic balance. However, existing home monitoring systems have fundamental flaws in data processing and patient self-management support. As a result, even if patients have the original data generated by self-testing equipment, they are still unable to independently complete disease status assessment and risk prediction. Specifically, the following technical bottlenecks are manifested: although existing home monitoring equipment such as smart blood pressure monitors and blood potassium test strips can provide real-time display functions for blood pressure, blood potassium and other values, their data output is only in the form of isolated parameters (such as "blood potassium 3.2mmol / L" and "blood pressure 145 / 92mmHg"), and are not combined with the patient's individualized baseline, medication records and PA-specific pathophysiological mechanisms for dynamic analysis; resulting in the inability to fully utilize the test data. At the same time, fixed home comparisons can easily delay treatment opportunities due to non-individualized analysis.

[0003] Based on this, in order to solve the above problems, the present invention provides an intelligent processing system for endocrine disease diagnosis data. Summary of the Invention

[0004] In order to solve the problems existing in the above solutions, the present invention provides an intelligent processing system for endocrine disease diagnosis data.

[0005] The purpose of the present invention can be achieved through the following technical solutions: An intelligent processing system for endocrine disease diagnosis data, including hospital side, cloud side and user side; The cloud includes a reserve module and a benchmark analysis module; The reserve module is used to summarize in real time various reserve questions that home patients have about the processing of test data; determine the basis for answering each reserve question, and mark the corresponding basis label according to whether the hospital needs to provide answer data for the answer based on the basis for answering; and establish a question reserve library after summarizing the reserve questions and the basis for answering.

[0006] The benchmark analysis module is used to configure a corresponding baseline model for the corresponding user terminal based on the received disease condition data, and the baseline model is used to generate a corresponding baseline risk standard based on the patient test data; and configure a question and answer library for the corresponding user terminal based on the disease condition data.

[0007] Furthermore, the configuration of the baseline model includes: Obtain various patient information, classify the patients according to the patient information, obtain several patient classifications, and set a corresponding baseline model for each patient classification; Identify the patient information and hospital test data corresponding to the disease condition data, match the patient classification to which the corresponding patient belongs based on the patient information, match the corresponding baseline model based on the patient classification, and verify and adjust the baseline model based on the hospital test data; Deploy the verified and adjusted baseline model on the user side.

[0008] Furthermore, a question-answer library is configured for the corresponding user terminal based on the disease condition data, including: Identify the various reserve problems in the problem reserve library, and analyze whether patients will have corresponding reserve problems during the home testing process based on disease condition data; Obtain the basis for answering the reserve questions, supplement the basis for answering according to the disease condition data, and configure the question and answer library according to the basis for answering the reserve questions.

[0009] The hospital side includes an information aggregation module; The information aggregation module is used to aggregate information of each patient, obtain hospital test data of the corresponding patient, obtain patient information, integrate the hospital test data and patient information into disease condition data; and evaluate whether the disease condition data can meet the patient's home application needs; When the assessment does not satisfy the requirements for home application, supplementary processing is performed on the disease condition data until the disease condition data satisfies the requirements for home application; When the assessment meets the requirements of home application, the disease condition data is sent to the benchmark analysis module in the cloud.

[0010] Furthermore, we evaluate whether the corresponding disease condition data can meet the home application needs of the corresponding patients, including: Obtaining the patient classification and baseline model information corresponding to the corresponding patient from the cloud-based benchmark analysis module based on the patient information; Calibrate hospital test data based on baseline model information to determine whether the hospital test data meets the verification and adjustment requirements of the baseline model; When the hospital test data does not meet the verification adjustment requirements of the baseline model, the disease condition data is evaluated as not meeting the home application requirements of the corresponding patient; When the hospital detection data meets the verification adjustment requirement of the baseline model, a corresponding user-side question and answer library is identified, a basis label of each answer in the question and answer library is identified, and an answer calibration basis is determined according to the basis label; the hospital detection data is calibrated according to the answer calibration basis, and a corresponding calibration result is obtained, when the calibration passes, it is evaluated that the corresponding disease condition data meets the home application requirement of the corresponding patient; when the calibration does not pass, it is evaluated that the corresponding disease condition data does not meet the home application requirement of the corresponding patient.

[0011] Further, judging whether the hospital detection data meets the verification adjustment requirement of the baseline model comprises: A verification calibration model is established, and the expression of the verification calibration model is: ; In the formula: (s, BV) is input data, s represents hospital detection data, and BV is a verification adjustment requirement of the baseline model; s→BV represents that the hospital detection data meets the verification adjustment requirement of the baseline model; and the output data is a verification calibration value YT (s, BV), and the verification calibration value is 1 or 0; The hospital detection data and the verification adjustment requirement are integrated as input data and input to the verification calibration for analysis, and a corresponding verification calibration value is obtained; When the verification calibration value is 1, the hospital detection data meets the verification adjustment requirement of the baseline model; When the verification calibration value is 0, the hospital detection data does not meet the verification adjustment requirement of the baseline model.

[0012] The user side comprises a collection module, a home analysis module and a question and answer module. The collection module is used for collecting home detection data of a user.

[0013] Further, the collection of the home detection data comprises: According to the baseline model, various data required to be collected at home by a user are obtained, a corresponding data collection list is generated, the data collection list is displayed to the user, the user collects data according to the data collection list, corresponding home detection data is obtained, and the home detection data is uploaded.

[0014] The home analysis module is used for performing real-time analysis on the home detection data through a preset baseline model, obtaining a baseline risk standard of the corresponding home detection data, and analyzing the home detection data according to the baseline risk standard, and obtaining a corresponding home analysis result.

[0015] The question and answer module is used for answering the detection question of the user based on the question and answer library.

[0016] Compared with the prior art, the beneficial effects of the present application are: The endocrine disease diagnosis data intelligent processing system of the present application breaks through the limitation of the existing household monitoring equipment which only provides isolated parameters. The system can combine the individualized baseline of the patient, the medication record and the pathophysiological mechanism specific to primary aldosteronism (PA), dynamically analyze the key physiological indicator data such as blood pressure and blood potassium obtained by household monitoring equipment such as intelligent blood pressure meter and blood potassium test paper instrument, and conduct individualized analysis combined with the dynamically obtained baseline risk standard, which helps the patient to discover the change of the disease in time, take corresponding measures, and avoid irreversible target organ damage such as myocardial hypertrophy, renal failure and serious arrhythmia caused by disease delay. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0018] Figure 1 The present application is a schematic diagram of the principle. DETAILED DESCRIPTION

[0019] The technical solutions of the present application will be described below in conjunction with the embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0020] As shown in Figure 1 An endocrine disease diagnosis data intelligent processing system, comprising a hospital end, a cloud end and a user end; The cloud end comprises a storage module and a baseline analysis module. The reserve module is used to aggregate various problems of home patients on detection data processing in real time, such as inconsistency between blood potassium detection results and hospital check, whether the blood potassium value is normal, whether the drug is invalid, and other various problems, and mark as a reserve problem after deduplication; determine the answer basis required for answering each reserve problem, such as the inconsistency between the blood potassium detection result and the hospital check, the answer basis of which includes hospital detection method information, patient application method information, various error causes and results, etc.; according to whether the answer basis needs to be answered by the hospital to provide answer data, mark the corresponding basis tag for the corresponding answer basis, such as hospital data supplement providing tag and normal tag, that is, for the answer basis that can be directly obtained, it is directly obtained, and for the data that needs to be provided by the hospital for the actual situation of the patient in the future, the answer basis in the problem database is not complete, and needs to be supplemented by subsequent data; the reserve problem and the answer basis are aggregated to establish a problem reserve library.

[0021] The baseline analysis module is used to configure a corresponding baseline model for a corresponding user terminal according to received disease condition data, and the baseline model is used to establish a prediction model of a patient individualized "blood potassium-blood pressure-aldosterone" dynamic baseline standard. The core function is to learn the physiological parameter fluctuation range of the patient in the normal state through historical data, and to provide a baseline for subsequent abnormal detection and risk warning; that is, the baseline model is used to generate a corresponding baseline risk standard according to patient detection data, and the baseline risk standard is a standard range of a corresponding analysis item, such as a blood potassium baseline of 3.5-4.5 mmol / L and a blood pressure baseline of 120-130 / 80-85 mmHg; according to the disease condition data, determine the reserve problems that the patient may have in the subsequent home process, and configure a corresponding question and answer library for the corresponding user terminal according to the corresponding reserve problem and the answer basis.

[0022] In one embodiment, the configuration of the baseline model includes: Obtain various patient information that may have, classify the patients according to the patient information, and obtain a plurality of patient classifications. The classification standard is that patients in the same category can apply the same baseline model or adaptively adjust the baseline model, that is, the baseline model can adapt to all patients in the patient classification. The classification can be based on clustering algorithms such as K-means clustering; Set a corresponding baseline model for each patient classification; Identify the patient information corresponding to the disease condition data and the hospital detection data, match the patient classification to which the patient belongs according to the patient information, match the corresponding baseline model according to the patient classification, and verify and adjust the baseline model according to the hospital detection data, that is, form a verification set of the baseline model according to the hospital detection data, verify through the verification set, and adjust according to the verification result.

[0023] In one embodiment, the patient is classified according to the patient information, and various existing methods can also be used to classify the patient first to form several initial classifications, and then the simulation diagnosis method is used to determine whether the same baseline model can be applied, and the classification is adjusted to obtain the patient classification.

[0024] In one embodiment, the baseline model is constructed by analyzing the patient's historical data (such as blood pressure, blood potassium, medication records, etc.) to build individualized normal physiological parameter fluctuation range (such as blood potassium baseline of 3.5-4.5 mmol / L, blood pressure baseline of 120-130 / 80-85 mmHg); can be established in many ways, Exemplary baseline model setting process: Data collection and preprocessing: Data types: including blood pressure, blood potassium, medication records (such as spironolactone dose), dietary intake (such as high potassium food), activity level, etc.

[0025] Data cleaning: remove noise data (such as abnormal values caused by sensor errors), fill in missing values (such as filling in unrecorded blood potassium values by interpolation method).

[0026] Data standardization: normalize different dimension data (such as blood pressure unit mmHg, blood potassium unit mmol / L) to a unified range (such as 0-1), eliminate dimension influence.

[0027] Time series analysis: establish baseline model through historical data to predict future data changes. For example, use moving average model to calculate the average blood potassium value of the patient in the past 7 days as the current baseline.

[0028] Model training and optimization: Feature selection: screen key features (such as blood potassium, blood pressure, spironolactone dose) from multi-source data to reduce redundant information.

[0029] Parameter tuning: adjust model hyperparameters (such as moving average window size, regularization coefficient of regression model) through grid search or random search to improve prediction accuracy.

[0030] Cross-validation: divide the data into training set and test set to evaluate the performance of the model on unseen data (such as mean square error MSE, mean absolute error MAE).

[0031] In one embodiment, the question and answer library is configured according to the disease condition data of the corresponding user terminal, the analysis of each reserve question in the question reserve library is carried out according to the disease condition data, whether the corresponding reserve question exists in the patient during the home process is evaluated, and the corresponding reserve question and answer basis are determined, the answer basis is supplemented according to the disease condition data, and the question and answer library is configured according to the reserve question and answer basis.

[0032] The hospital end is used by the hospital side, comprising an information aggregation module; The information aggregation module is used to aggregate information of each patient, such as blood sodium, blood potassium, blood glucose, blood lipids, urine biochemical indicators, plasma aldosterone concentration, plasma renin activity, saline infusion test data, adrenal gland B-ultrasound, adrenal gland MRI, doctor's diagnosis data and various related data. The above collected data is integrated and marked as hospital detection data, that is, the hospital detection data is composed of each single data and is the data within a continuous time; patient information is obtained, which includes age, gender, medical history and other information related to risk assessment benchmark determination, and the hospital detection data and the patient information are integrated into disease condition data; whether the corresponding disease condition data meets the home application needs of the corresponding patient is evaluated; When the evaluation does not meet the home application needs, the disease condition data of the patient is supplemented until the disease condition data meets the home application needs; When the evaluation meets the home application needs, the disease condition data is sent to the benchmark analysis module of the cloud end.

[0033] In one embodiment, whether the corresponding disease condition data meets the home application needs of the corresponding patient is evaluated, comprising: Obtaining patient classification and baseline model information corresponding to the corresponding patient classification from the benchmark analysis module of the cloud end, the baseline model information including required input data, output result information, verification and adjustment requirements and other related information, the verification and adjustment requirements being used to analyze whether the hospital detection data meets the verification and adjustment requirements of the baseline model as a verification set; calibrating the hospital detection data according to the baseline model information to determine whether the hospital detection data meets the verification and adjustment requirements of the baseline model; When the hospital detection data does not meet the verification and adjustment requirements of the baseline model, it is evaluated that the corresponding disease condition data does not meet the home application needs of the corresponding patient; When the hospital detection data meets the verification and adjustment requirements of the baseline model, the question and answer library of the patient user end is identified, the basis labels of each answer basis in the question and answer library are identified, the answer basis that needs to be supplemented by the hospital data is determined according to the basis labels, and is marked as an answer calibration basis; the hospital detection data is calibrated according to the answer calibration basis to determine whether the hospital detection data includes all data required by the hospital to provide the answer calibration basis, and a corresponding calibration result is obtained. When the calibration is passed, it is evaluated that the corresponding disease condition data meets the home application needs of the corresponding patient; when the calibration is not passed, it is evaluated that the corresponding disease condition data does not meet the home application needs of the corresponding patient.

[0034] In one embodiment, whether the hospital detection data meets the verification and adjustment requirements of the baseline model can be directly judged based on the existing mode.

[0035] In one embodiment, determining whether the hospital detection data meets the verification adjustment requirement of the baseline model comprises: A verification calibration model is established, and the verification calibration model is used to determine whether the corresponding hospital detection data meets the verification requirement of the baseline model. The expression of the verification calibration model is: In the formula, (s, BV) is input data, s represents hospital detection data, and BV represents the verification adjustment requirement of the baseline model. s→BV represents that the hospital detection data meets the verification adjustment requirement of the baseline model. The corresponding historical data is marked to train the training set, the hospital detection data including the data required by the verification adjustment requirement is calibrated one by one, that is, when the hospital detection data including all the data required by the verification adjustment requirement is considered to meet the verification adjustment requirement of the baseline model, the required various data types are determined according to the verification adjustment requirement, and the hospital detection data is analyzed and judged based on the determined data types. Then, the corresponding training set is established according to the above manner to train. The output data is a verification calibration value YT(s, BV), and the verification calibration value is 1 or 0. The hospital detection data and the verification adjustment requirement are integrated as input data to be input into the verification calibration for analysis to obtain a corresponding verification calibration value. When the verification calibration value is 1, the hospital detection data meets the verification adjustment requirement of the baseline model. When the verification calibration value is 0, the hospital detection data does not meet the verification adjustment requirement of the baseline model.

[0036] The user terminal is used by a user and comprises a collection module, a home analysis module, and a question and answer module. The collection module is used to collect various related data of the user's home detection. The user can be guided to collect data according to the input data required by the baseline model, such as data detected by various home detection devices, data detected by clinics, community hospitals, and the like. The data can be uploaded in various ways such as photographing, transmission, and manual input. The corresponding data is marked as home detection data.

[0037] In one embodiment, the detection error of the user is corrected by analyzing the collection method of the user, such as the error caused by the device.

[0038] The home analysis module is used to analyze the home detection data in real time by using the preset baseline model to obtain a baseline risk standard of the corresponding home detection data. The home detection data is analyzed according to the baseline risk standard to obtain a corresponding home analysis result.

[0039] ​The question and answer module is used for answering the detected question of the user based on a question and answer library, and existing question and answer technology is used for answering, such as commonly matching a corresponding stored question and an answering basis from the question and answer library according to the question of the user, generating corresponding answering data according to the answering basis, or using current intelligent question and answer technology for answering.

[0040] In one embodiment, the user terminal also has a function of communicating with the hospital terminal, such as communicating with corresponding medical staff.

[0041] The above formulas are all calculated by removing the dimension and taking the numerical value, the formula is obtained by collecting a large amount of data to simulate the software to obtain a formula closest to the real situation, and the preset parameters and the preset threshold in the formula are set by the person skilled in the art according to the actual situation or obtained by a large amount of data simulation.

[0042] The above embodiments are only used to illustrate the technical method of the present application but not limit the present application, although the present application is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. An endocrine disease diagnosis data intelligent processing system, characterized by, The hospital end, the cloud end and the user end are included; The cloud end includes a reserve module and a benchmark analysis module; the hospital end includes an information summary module; and the user end includes a collection module and a home analysis module; The reserve module is used for real-time summarizing various reserve problems of home patients in detection data processing, determining answer bases required for solving each reserve problem, marking corresponding basis tags according to whether the hospital needs to provide answer data for solving, and establishing a problem reserve library after summarizing the reserve problems and the answer bases; The benchmark analysis module is used for configuring a corresponding baseline model for the user end according to received disease condition data, generating a corresponding baseline risk standard according to patient detection data, and configuring a question and answer library for the user end according to the disease condition data; The information summary module is used for information summarizing of each patient, obtaining hospital detection data of the corresponding patient, acquiring patient information, integrating the hospital detection data and the patient information into disease condition data, and evaluating whether the disease condition data can meet the home application needs of the patient; When the evaluation does not meet the home application needs, the disease condition data is supplemented until the disease condition data meets the home application needs; When the evaluation meets the home application needs, the disease condition data is sent to the benchmark analysis module of the cloud end; The collection module is used for collecting home detection data of the user; The home analysis module is used for real-time analysis of the home detection data through a preset baseline model, obtaining a baseline risk standard of the corresponding home detection data, and analyzing the home detection data according to the baseline risk standard to obtain a corresponding home analysis result.

2. The endocrine disease diagnosis data intelligent processing system according to claim 1, characterized in that, The configuration of the baseline model includes: Obtaining various patient information, classifying patients according to the patient information, obtaining a plurality of patient classifications, and setting a corresponding baseline model for each patient classification; Identifying patient information and hospital detection data corresponding to the disease condition data, matching the patient classification to which the patient belongs according to the patient information, matching the corresponding baseline model according to the patient classification, and verifying and adjusting the baseline model according to the hospital detection data; Deploying the baseline model verified and adjusted in the user end.

3. The endocrine disease diagnosis data intelligent processing system according to claim 1, characterized in that, Configuring a question and answer library for the corresponding user end according to the disease condition data includes: Identifying each reserve problem in the problem reserve library, and analyzing whether the patient will have the corresponding reserve problem in the home detection process according to the disease condition data; Obtaining answer bases for judging the reserve problems, supplementing the answer bases according to the disease condition data, and configuring the question and answer library according to the answer bases of the reserve problems.

4. The endocrine disease diagnosis data intelligent processing system according to claim 1, characterized in that, Evaluating whether the corresponding disease condition data can meet the home application needs of the corresponding patient includes: Obtaining the patient classification corresponding to the patient and the baseline model information corresponding to the patient from the benchmark analysis module of the cloud end according to the patient information; Calibrating the hospital detection data according to the baseline model information, and judging whether the hospital detection data meets the verification and adjustment needs of the baseline model; When the hospital detection data does not meet the verification and adjustment needs of the baseline model, evaluating that the disease condition data does not meet the home application needs of the corresponding patient; When the hospital test data meets the verification and adjustment requirements of the baseline model, identify the question and answer library of the corresponding user end, identify the basis labels of each answer in the question and answer library, and determine the answer calibration basis according to the basis labels; calibrate the hospital test data according to the answer calibration basis to obtain the corresponding calibration results. When the calibration passes, evaluate the corresponding disease condition data to meet the home application requirements of the corresponding patient; when the calibration fails, evaluate the corresponding disease condition data to not meet the home application requirements of the corresponding patient.

5. The endocrine disease diagnosis data intelligent processing system according to claim 4, characterized in that, Determine whether the hospital's test data meets the verification and adjustment requirements of the baseline model, including: Establishing a validation calibration model; Integrate hospital test data and verification adjustment requirements as input data and input them into verification calibration for analysis to obtain corresponding verification calibration values, which are 1 or 0; When the validation calibration value is 1, the hospital test data meets the validation adjustment requirements of the baseline model; When the validation calibration value is 0, the hospital test data does not meet the validation adjustment requirements of the baseline model.

6. The endocrine disease diagnosis data intelligent processing system according to claim 5, wherein The expression for verifying the calibration model is: ; Where: (s, BV) is the input data, s represents the hospital test data, and BV is the verification adjustment requirement of the baseline model; s→BV means that the hospital test data meets the verification adjustment requirement of the baseline model; the output data is the verification calibration value YT(s, BV), and the verification calibration value is 1 or 0.

7. The endocrine disease diagnosis data intelligent processing system according to claim 1, characterized in that, Collection of home testing data, including: According to the baseline model, various data that users need to collect at home are obtained, a corresponding data collection list is generated, and the data collection list is displayed to the user. The user collects data according to the data collection list, obtains corresponding home detection data, and uploads the home detection data.

8. The endocrine disease diagnosis data intelligent processing system according to claim 1, characterized in that, The user terminal further includes a question-answering module, which is used to answer the user's detection questions based on the question-answering library.

Images