Artificial Intelligence-Based Medical and Health Service Data Processing and Analysis System

Abstract

The present invention discloses a medical and health service data processing and analysis system based on artificial intelligence, belonging to the technical field of medical and health service data processing, and including a user module, a user analysis module, and a monitoring and analysis module; the user module is used for users to upload corresponding physiological data and store the disease record data of users; the user analysis module is used for analyzing the disease record data of users to determine the user scope; merging and classifying according to the user scope to obtain each status classification; setting corresponding status criteria and adjustment models for each status classification; obtaining the disease record data of users, determining the status classification of users according to the disease record data, and identifying the status criteria corresponding to the status classification; obtaining the physiological data of users and the corresponding detection device information; analyzing through the adjustment model to obtain corresponding monitoring optimization data; the monitoring and analysis module is used for evaluating the physical status of users to obtain an evaluation result and sending the evaluation result to the user module.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical and health service data processing, and specifically relates to a medical and health service data processing and analysis system based on artificial intelligence. Background Art

[0002] With the rapid development of technology, traditional medical detection methods, such as physiological information monitoring, have been widely popularized, enabling users to obtain their physiological parameters such as heart rate, blood pressure, and blood sugar in real time. However, although these detection technologies are becoming increasingly mature, users often only obtain simple numerical displays and cannot deeply understand the health implications or potential risks behind these data; for devices with intelligent analysis functions, such as smart wearable devices, they are often not accepted by some users due to price, accuracy, etc. The above users generally purchase single-function detection devices for blood pressure, body temperature, heart rate, etc. for detection, resulting in their judgment being based solely on their own experience, with great uncertainty; based on this, in order to solve the above problems, the present invention provides a medical and health service data processing and analysis system based on artificial intelligence. Summary of the Invention

[0003] In order to solve the problems existing in the above solutions, the present invention provides a medical and health service data processing and analysis system based on artificial intelligence.

[0004] The object of the present invention can be achieved by the following technical solutions:

[0005] A medical and health service data processing and analysis system based on artificial intelligence includes a user module, a user analysis module, and a monitoring and analysis module;

[0006] The user module is used for users to upload corresponding physiological data and store the disease record data of users.

[0007] The user analysis module is used to analyze the disease record data of users to determine the user scope; perform merging and classification according to the user scope to obtain each status classification, and mark the matching scope corresponding to each status classification; set corresponding status standards and adjustment models for each status classification; establish a status matching library, and store each status classification and the corresponding status standards and matching scopes into the status matching library; obtain the disease record data of users, input the disease record data into the status matching library for matching to determine the status classification of users, and identify the status standards corresponding to the status classification; obtain the physiological data of users and the corresponding detection device information; perform analysis through the adjustment model to obtain the corresponding monitoring optimization data.

[0008] Further, the method for merging and classification according to the user scope includes:

[0009] Set up a disease classification map, divide the user scope through the disease classification map, and obtain each unit class; assign corresponding disease labels to each unit class, and set several merging sets according to the disease labels corresponding to each unit class;

[0010] Set corresponding test sets for each merging set, and perform merging analysis on each unit class within the merging set through the test sets to obtain each status classification.

[0011] Furthermore, the method for performing merging analysis on each unit class within the merging set through the test set includes:

[0012] Step SA1: Select several test data from the test set to test each unit class, and obtain the test results corresponding to each unit class;

[0013] Step SA2: Calculate the difference values between the test results of each unit class, and determine whether the merging requirements are met according to the difference values;

[0014] When it is determined that the merging requirements are not met, classify the corresponding unit classes into different initial classifications;

[0015] When it is determined that the merging requirements are met, classify the corresponding unit classes into the same initial classification;

[0016] And so on, until all unit classes are allocated to the corresponding initial classifications;

[0017] Step SA3: Select several new test data from the test set to test each unit class within each initial classification, and determine whether each unit class meets the merging requirements;

[0018] When it is determined that the merging requirements are not met, classify the corresponding unit classes that do not meet the merging requirements into new initial classifications;

[0019] When it is determined that the merging requirements are met, no corresponding operation is performed;

[0020] And so on, until all unit classes within each initial classification meet the merging requirements;

[0021] Step SA4: Loop step SA3 until the entire training set is tested, and mark each initial classification as a status classification.

[0022] Furthermore, the calculation method of the difference value includes:

[0023] Assume two unit classes as A and B respectively, obtain the test results corresponding to unit class A and unit class B, set corresponding difference items according to the test results, adjust the two test results according to each difference item, and obtain the difference analysis data corresponding to unit class A and unit class B respectively;

[0024] Mark the data of each difference item in the difference analysis data as CYAi and CYBi respectively, where i represents the corresponding difference item, i = 1, 2, ……, n, and n is a positive integer;

[0025] According to the formula Calculate the single value of each difference item;

[0026] In the formula: DYi is the single value; ψ is the preset representative value;

[0027] According to the formula Calculate the corresponding difference value; In the formula: PH is the difference value.

[0028] The monitoring and analysis module is used to evaluate the physical state of the user, obtain the monitoring optimization data of the user and the corresponding state standards; evaluate the physical state of the user according to the monitoring optimization data and the state standards, obtain the corresponding evaluation results, and send the evaluation results to the user module.

[0029] Further, the method for evaluating the physical state of the user according to the monitoring optimization data and the state standards includes:

[0030] Identify the data of each monitoring item in the monitoring optimization data, mark the data of the monitoring item as GTj, j = 1, 2, ……, m, and m is a positive integer; identify the data of each standard item in the state standards, and mark the data of the standard item as BGj;

[0031] According to the formula Calculate the corresponding physical evaluation value;

[0032] In the formula: GPR is the physical evaluation value; βj is the proportionality coefficient, and the value range is 0 < βj ≤ 1; PX(*) is a custom function, and * is the input data of the custom function;

[0033] When the physical evaluation value is greater than the threshold X1, the evaluation status is abnormal;

[0034] When the physical evaluation value is not greater than the threshold X1, the evaluation status is normal.

[0035] Further, the expression of the custom function is:

[0036]

[0037] Further, it also includes a medication monitoring module. The medication monitoring module is used to monitor the user's medication, establish a platform material library, and the platform material library is used to store each alternative target, candidate drug, influencing collection item, and medication judgment standard;

[0038] Determine the monitoring target and the corresponding medication information, and match the corresponding influencing collection item and medication judgment standard from the platform material library according to the medication information corresponding to the monitoring target;

[0039] Collect the corresponding medication status monitoring data according to the influencing collection items; generate the corresponding medication status monitoring curve according to the medication status monitoring data;

[0040] Establish a medication judgment model according to the medication judgment criteria, and perform real-time analysis on the medication status monitoring curve through the medication judgment model to obtain the corresponding medication judgment value;

[0041] When a medication judgment value equal to 1 is not obtained within the preset time, perform corresponding warning prompts.

[0042] Furthermore, the expression of the medication judgment model is:

[0043]

[0044] In the formula: s is the medication status monitoring curve, and the output data is the medication judgment value 1 or 0.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] Through the mutual cooperation between the user module, the user analysis module and the monitoring analysis module, the intelligent analysis of the user's physiological data is realized, and the problem that it is difficult for users to accurately analyze the physiological data based on their own experience is solved; by setting up a medication monitoring module, the monitoring personnel assisting the user can monitor the user's medication situation and urge the user to take medicine on time. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0048] Figure 1 It is the principle block diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0050] Such as Figure 1As shown in the figure, an artificial intelligence-based medical and health service data processing and analysis system includes a user module, a user analysis module, a monitoring analysis module, and a medication monitoring module;

[0051] The user module is used for each user to register or log in, upload physiological data detected by the user, such as blood pressure, blood sugar, body temperature, heart rate, etc.; and receive information sent by other modules in the follow-up, and display the received information to the user; and the user needs to fill in corresponding disease record data in the user module, such as gender, age, disease treatment situation, medications taken, and other relevant data;

[0052] The user analysis module is used to analyze the disease record data of the user, understand the actual physical condition of the user, and achieve targeted analysis of the user in the follow-up; the detailed process is as follows:

[0053] According to various current physiological monitoring requirements, obtain various possible user situations, combine the business scope of the platform party to set various user situations that meet the service scope of the platform party, and form a user scope; perform merging and classification according to the user scope to obtain each status classification, and mark the matching scope corresponding to each status classification; the matching scope is the summary scope corresponding to each disease record data corresponding to this status classification;

[0054] According to the differences of each status classification, set corresponding status standards and adjustment models for each status classification. The status standard is to set detection standards for this situation according to the user situation corresponding to this status classification, such as standard values of blood pressure, heart rate, etc.; the adjustment model is used to correct the detection data of various detection devices on the current market. Specifically, the platform party analyzes the differences in the detection of users in different status classifications by different detection devices, and can use artificial test simulations, test data of corresponding detection devices, etc. for analysis, and then establish a corresponding training set manually. The training set includes input data and output data. The input data is detection data and detection device information, and the output data is the corrected detection information; reduce the detection error caused by the user's self-detection; establish a corresponding adjustment model based on neural networks such as CNN networks or DNN networks, train through the established training set, and analyze through the adjusted model after successful training.

[0055] Sort out each status classification, matching scope, and status standard, and establish a corresponding status matching library;

[0056] Obtain the disease record data of the user, input the disease record data into the status matching library for matching, determine the status classification of the user, obtain the physiological data of the user and the detection device information used by the user for detection, analyze through the corresponding adjustment model, and obtain the corrected physiological data, marked as monitored and optimized data.

[0057] The method of merging and classifying according to the user scope includes:

[0058] The platform first sets a disease classification map, which is set based on the finest classification criteria. For example, for hypertension disease, it will be divided into finer levels according to various variable parameters. Exemplarily, corresponding classifications are set according to parameters such as gender, age group, weight range, etc., mainly based on whether the parameter range has the same impact on the disease; specifically, it is set by the platform.

[0059] The user scope is segmented through the disease classification map to obtain the user scope corresponding to each classification, which is marked as a unit class; that is, various user situations in the unit class can be analyzed using the same analysis criteria.

[0060] Corresponding disease labels are assigned to each unit class according to the disease type differences, which can be a single disease or a combination of multiple diseases; several merging sets are set according to the disease labels corresponding to each unit class. That is, first, they are classified according to each single disease label, and then the unit classes with combined disease labels including this disease label are assigned to this classification. That is, for the unit classes with combined disease labels, they will appear in multiple merging sets.

[0061] Corresponding test sets are set for each merging set. The test sets are various preset ones used to test whether each unit class can be regarded as the same situation for subsequent disease analysis. Generally, they are corresponding physiological data. The physiological data of each user within the unit class is the same and is the physiological data in the test set. Analyze whether the physical analysis results are the same under the same physiological data.

[0062] Through the test set, combined analysis is carried out on each unit class within the merging set, and the unit classes that meet the merging requirements are merged to obtain each status classification.

[0063] The method of carrying out combined analysis on each unit class within the merging set through the test set includes:

[0064] Step SA1: Select several test data from the test set to test each unit class, and obtain the test results corresponding to each unit class. The test results are determined according to various existing disease and physical condition analysis results. That is, a large amount of historical evaluation data for evaluating users according to their physiological status is obtained in advance, and then data matching is carried out according to the user situation, physiological status, etc. to obtain the corresponding historical evaluation data under this test condition, and data statistics are carried out to identify various evaluation results and the proportion corresponding to each evaluation result; test various user situations within the unit class, obtain the evaluation results and proportions corresponding to each user situation, and conduct summary statistics to obtain the various evaluation results and result proportions corresponding to this unit class. Integrate and mark the various evaluation results and result proportions corresponding to the unit class as the test results.

[0065] Step SA2: Calculate the difference values between the test results of each unit class, and determine whether the merging requirements are met based on the difference values. When the difference value is greater than the threshold X1 or ψ exists in the difference values, it is determined that the merging requirements are not met; otherwise, it is determined that the merging requirements are met;

[0066] When it is determined that the merging requirements are not met, classify the corresponding unit classes into different initial classifications;

[0067] When it is determined that the merging requirements are met, classify the corresponding unit classes into the same initial classification;

[0068] And so on, until all unit classes are assigned to the corresponding initial classifications; a single one forms an initial classification independently;

[0069] Step SA3: Select several new test data from the test set to test each unit class within each initial classification, and evaluate whether each unit class meets the merging requirements;

[0070] When it is determined that the merging requirements are not met, classify the corresponding unit classes that do not meet the merging requirements into new initial classifications;

[0071] When it is determined that the merging requirements are met, no corresponding operation is performed;

[0072] And so on, until all unit classes within each initial classification meet the merging requirements;

[0073] Step SA4: Loop step SA3 until the entire training set is tested, and mark each initial classification as a status classification.

[0074] The calculation methods of the difference values include:

[0075] Assume two unit classes as A and B respectively, obtain the test results corresponding to unit class A and unit class B, count various evaluation results that exist in both test results, mark them as difference items, adjust the two test results according to each difference item, and obtain the difference analysis data corresponding to unit class A and unit class B respectively; that is, for a difference item that does not exist in a certain test result, supplement it therein, but the corresponding difference item data is 0;

[0076] Mark the data of each difference item in the difference analysis data as CYAi and CYBi respectively, that is, the corresponding result ratio; i represents the corresponding difference item, i = 1, 2,..., n, and n is a positive integer;

[0077] According to the formula Calculate the single value of each difference item; in the formula: DYi is the single value; ψ is a preset representative value, non - numerical, directly indicating that it cannot be merged;

[0078] According to the formula Calculate the corresponding difference value; where: PH is the difference value.

[0079] The monitoring and analysis module is used to analyze the user's monitoring and optimization data, obtain the user's monitoring and optimization data and the corresponding status standards; evaluate the user's physical status according to the monitoring and optimization data and the status standards, obtain the corresponding evaluation results, and send the evaluation results to the user module.

[0080] The method for evaluating the user's physical status according to the monitoring and optimization data and the status standards includes:

[0081] Mark the corresponding data items such as blood pressure and heart rate in the monitoring and optimization data as monitoring items, and mark the corresponding data as monitoring item data. The monitoring items correspond to the corresponding data in the status standards, and mark the data items in the status standards as standard items; identify each monitoring item data in the monitoring and optimization data, and mark the monitoring item data as GTj, j = 1, 2,..., m, where m is a positive integer; identify each standard item data in the status standards, and mark the standard item data as BGj;

[0082] According to the formula Calculate the corresponding physical evaluation value;

[0083] Where: GPR is the physical evaluation value; βj is the proportionality coefficient, and the value range is 0 < βj ≤ 1; PX(*) is a custom function, and * is the input data of the custom function; calculate according to the actual situation of this monitoring item. If the difference between GTj and BGj is within the allowable range, its output is 0, because for each monitoring item, its actual requirement is a numerical interval, and within this interval, GTj is regarded as meeting the status standard requirements. Otherwise, the absolute value of the difference between the two is output; the expression is

[0084]

[0085] When the physical evaluation value is greater than the threshold X1, the evaluation status is abnormal;

[0086] When the physical evaluation value is not greater than the threshold X1, the evaluation status is normal.

[0087] The medication monitoring module is used to monitor the user's medication status. Because many users often need to take corresponding medications, especially for diseases such as hypertension and heart diseases, but many elderly users often forget to take their medications, and the corresponding caregivers cannot truly understand the actual medication situation. Therefore, for many elderly users and others, it is necessary to analyze whether they have taken their medications, so as to assist the caregivers in taking care; the detailed process is as follows:

[0088] Obtain the diseases corresponding to the medications that need to be monitored by the user, mark them as monitoring targets, which are set by the user, and set the corresponding medication information;

[0089] The platform sets various diseases that can be monitored for medication use, marks them as alternative targets, obtains various drugs that may be applied to each alternative target, marks them as candidate drugs, determines the physical changes of the alternative targets after taking the corresponding candidate drugs based on historical medication data, and then sets the corresponding influencing collection items, that is, in order to judge whether medication is taken, which physical data need to be collected, such as blood pressure, blood sugar, heart rate, etc.;

[0090] And determine the changes in the data of each influencing collection item of each alternative target after taking the corresponding candidate drugs according to the existing medical data or through methods such as testing and analysis by the platform, and then set the corresponding medication judgment criteria; such as blood pressure changes, etc.; Integrate each alternative target, candidate drug, influencing collection item, and medication judgment criteria to establish a corresponding platform material library;

[0091] Match the corresponding influencing collection items and medication judgment criteria from the platform material library according to the medication information corresponding to the monitoring target;

[0092] Perform real-time collection according to the influencing collection item data to obtain the corresponding medication status monitoring data; Generate the corresponding medication status monitoring curve according to the obtained medication status monitoring data, with the horizontal axis being time and the vertical axis being the data of the corresponding influencing collection item, forming multiple images, uniformly marked as the medication status monitoring curve;

[0093] Establish a medication judgment model according to the medication judgment criteria. The medication judgment model is used to analyze the medication status monitoring curve based on the medication judgment criteria to judge whether medication is taken, mainly judging according to the changes in the curve. Using the medication judgment criteria, the medication status monitoring curve can be judged based on the existing technology, such as establishing a corresponding intelligent model based on neural network for recognition and judgment. The expression of the medication judgment model is In the formula: s is the input data, that is, the medication status monitoring curve, and the output data is the medication judgment value 1 or 0;

[0094] Perform real-time analysis on the medication status monitoring curve through the medication judgment model to obtain the corresponding medication judgment value;

[0095] When a medication judgment value equal to 1 is not obtained within the preset time, corresponding warning prompts are given. The preset time is set by the user.

[0096] The above formulas are all calculated by removing the dimension and taking their numerical values. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters and preset thresholds in the formulas are set by those skilled in the art according to the actual situation or obtained through a large amount of data simulation.

[0097] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An artificial intelligence-based medical and health service data processing and analysis system, characterized in that, It includes a user module, a user analysis module, and a monitoring and analysis module; The user module is used for users to upload corresponding physiological data and store the disease record data of users; The user analysis module is used to analyze the disease record data of users to determine the user scope; perform merging and classification according to the user scope to obtain each status classification, and mark the matching scope corresponding to each status classification; set corresponding status criteria and adjustment models for each status classification; establish a status matching library, and store each status classification and the corresponding status criteria and matching scope of each status classification into the status matching library; obtain the disease record data of users, input the disease record data into the status matching library for matching to determine the status classification of users, and identify the status criteria corresponding to the status classification; obtain the physiological data of users and the corresponding detection device information; perform analysis through the adjustment model to obtain corresponding monitoring optimization data; The monitoring and analysis module is used to evaluate the physical status of users, obtain the monitoring optimization data of users and the corresponding status criteria; evaluate the physical status of users according to the monitoring optimization data and the status criteria to obtain the corresponding evaluation results, and send the evaluation results to the user module; The method for merging and classification according to the user scope includes: Set a disease classification diagram, divide the user scope through the disease classification diagram to obtain each unit class; mark each unit class with the corresponding disease label, and set several merging sets according to the disease labels corresponding to each unit class; Set corresponding test sets for each merging set, and perform merging analysis on each unit class within the merging set through the test sets to obtain each status classification; The method for performing merging analysis on each unit class within the merging set through the test sets includes: Step SA1: Select several test data from the test set to test each unit class to obtain the test results corresponding to each unit class; Step SA2: Calculate the difference values between the test results of each unit class, and judge whether the merging requirements are met according to the difference values; When it is determined that the merging requirements are not met, classify the corresponding unit classes into different initial classifications; When it is determined that the merging requirements are met, classify the corresponding unit classes into the same initial classification; And so on until all unit classes are allocated to the corresponding initial classifications; Step SA3: Select several new test data from the test set to test each unit class within each initial classification, and judge whether each unit class meets the merging requirements; When it is determined that the merging requirements are not met, classify the corresponding unit classes that do not meet the merging requirements into new initial classifications; When it is determined that the merging requirements are met, no corresponding operation is performed; And so on until the merging requirements are met among all unit classes within each initial classification; Step SA4: Loop step SA3 until all the training sets are tested, and mark each initial classification as a status classification; The calculation method of the difference value includes: Assume two unit classes as A and B respectively, obtain the test results corresponding to unit class A and unit class B, set corresponding difference items according to the test results, adjust the two test results according to each difference item to obtain the difference analysis data corresponding to unit class A and unit class B respectively; Mark the data of each difference item in the difference analysis data as CYAi and CYBi respectively, where i represents the corresponding difference item, i = 1, 2, ……, n, and n is a positive integer; According to the formula Calculate the single value of each difference term; In the formula: DYi is a single value; ψ is a preset representative value; According to the formula calculate the corresponding difference value; where: PH is the difference value; The method for evaluating the physical state of a user according to the monitoring optimization data and the state standard includes: Identify the data of each monitoring item in the monitoring optimization data, and mark the monitoring item data as GTj, j = 1, 2, ……, m, where m is a positive integer; identify the data of each standard item in the state standard, and mark the standard item data as BGj; According to the formula Calculate the corresponding physical assessment value; In the formula: GPR is the physical evaluation value; βj is a proportionality coefficient, and the value range is 0 < βj ≤ 1; PX(*) is a custom function, and * is the input data of the custom function; When the physical evaluation value is greater than the threshold X1, the evaluation state is abnormal; When the physical evaluation value is not greater than the threshold X1, the evaluation state is normal; The expression of the custom function is: 。 2. The medical and health service data processing and analysis system based on artificial intelligence according to claim 1, wherein It further includes a medication monitoring module, which is used to monitor the user's medication and establish a platform material library. The platform material library is used to store each alternative target, candidate drug, influencing collection item, and medication judgment criterion; Determine the monitoring target and the corresponding medication information, and match the corresponding influencing collection item and medication judgment criterion from the platform material library according to the medication information corresponding to the monitoring target; Collect the corresponding medication status monitoring data according to the influencing collection item data; generate the corresponding medication status monitoring curve according to the medication status monitoring data; Establish a medication judgment model according to the medication judgment criterion, and perform real-time analysis on the medication status monitoring curve through the medication judgment model to obtain the corresponding medication judgment value; When a medication judgment value equal to 1 is not obtained within the preset time, corresponding warning prompts are given.

3. The artificial intelligence-based medical and health service data processing and analysis system according to claim 2, wherein The expression of the medication judgment model is: ; In the formula: s is the medication status monitoring curve, and the output data is the medication judgment value 1 or 0.

Images