Medicine data classification system and classification method thereof

Abstract

The invention discloses a medicine data classification system, which comprises a cloud server in which a central processing unit is arranged; the data importing system is used for importing hospital patient information, diagnosis information and medication information; and the medicine management system is used for maintaining basic information of medicines and realizing classified statistics of label information, name information, specification information and unit information of medicines. According to the medicine data classification system and the classification method thereof, management of patient information, diagnosis information and medicine use information can be realized, medicine classification can be realized based on the medicine information and the diagnosis information, and an effective data basis is provided for metering calculation or big data analysis of medicines.

Description

Technical Field

[0001] The present invention relates to the technical field of data classification and grading, and in particular to a medical data classification system and a classification method thereof. Background Art

[0002] Data classification can be understood as knowing which data (actually metadata, or more accurately, fields) belongs to which business scope, that is, category, from the perspective of business or industry. Data classification is different from data classification. For most companies, data classification is more from the perspective of meeting regulatory requirements. Data classification belongs to the field of data security, and it may be more appropriate to call it sensitivity level. Some data in an enterprise have a high degree of confidentiality, some have a low degree of confidentiality, some can be made public, and some cannot be made public. Data with different sensitivity levels are subject to different protection strategies when used internally, and the degree of openness to external sharing is also different.

[0003] The identification of sensitive data is the most fundamental issue in data security management. There are many methods for identifying sensitive data, but each method has its own advantages and disadvantages.

[0004] However, the existing data classification and grading are deficient in terms of security and privacy protection; at the same time, the existing data classification and grading methods do not distinguish between real-time data and existing data, which will lead to additional data identification costs; in addition, the existing technologies are all based on general models to perform data classification or grading, and cannot perform personalized classification based on the user's active requirements, cannot meet the user's specific classification requirements, the classification cannot be further refined, and the data value cannot be mined to the greatest extent. Summary of the invention

[0005] 1. Technical issues to be resolved

[0006] In view of the deficiencies of the prior art, the present invention provides a medical data classification system and a classification method thereof.

[0007] (II) Technical solution

[0008] To achieve the above object, the present invention provides the following technical solution: a medical data classification system, comprising:

[0009] A cloud server, wherein a central processing unit is provided in the cloud server;

[0010] Data import system, used to import hospital patient information, diagnosis information and medication information;

[0011] The drug management system is used to maintain basic drug information and realize classified statistics of drug label information, name information, specification information, and unit information;

[0012] The drug classification system is used to establish a binary matrix of medication information and diagnosis information, and classify drugs based on the binary matrix;

[0013] The classification and grading system also includes: a data receiving subsystem, a data identification subsystem, a data screening subsystem, a data classification subsystem and a data grading subsystem;

[0014] The data receiving subsystem provides a plurality of APR interface terminals, through which the data receiving subsystem receives the data to be processed from a plurality of data sources, the plurality of data sources including a database and a data terminal; the data identification subsystem is connected to the data receiving subsystem to identify the attributes of the data to be processed; the data screening subsystem receives input screening criteria, and based on the screening criteria, screens the data to be processed from the data to be processed; the data classification subsystem is used to classify the data to be processed, and the data grading subsystem performs data grading and value assessment on the data to be processed based on the classification results;

[0015] Sensitive data system, used to analyze sensitive data and build relevant models.

[0016] Preferably, after receiving the first data to be processed from the data terminal, the data receiving subsystem sends the first data to be processed to the data identifying subsystem.

[0017] Preferably, the process of classifying drugs by the drug classification module is as follows:

[0018] S1: Obtain diagnostic information and medication information, and pre-process the diagnostic information and medication information, including word form normalization and unified replacement of medication name naming methods;

[0019] S2: Based on the correspondence between the diagnosis information and the medication information, a binary matrix of the medication information and the diagnosis information is established;

[0020] S3: Based on the binary matrix of medication information and diagnosis information, the similarity distance between drugs is calculated, and the distance matrix between drugs is established;

[0021] S4: Classify drugs based on the distance matrix between drugs.

[0022] Preferably, the sensitive data system includes:

[0023] Build a sensitive data classification model;

[0024] Inputting the sensitive data features into the sensitive data classification model to obtain a sensitive data classification result;

[0025] Build a sensitive data classification model;

[0026] Inputting the sensitive data classification result, the sensitive data characteristics and the preset data security level into the sensitive data classification model to obtain a sensitive data classification result;

[0027] The sensitive data is classified and identified according to the sensitive data classification result, the sensitive data characteristics and the sensitive data grading result.

[0028] Preferably, the construction of the sensitive data classification model includes: preprocessing the sensitive data; dividing the preprocessed sensitive data into multiple sets, cyclically selecting any one set as the first test set, and the remaining sets as the first training sets; setting the neural network model and parameters to construct an initial sensitive data classification model; training and testing the sensitive data classification model according to the first training set and the first test set, performing incremental neural network learning, and continuously adjusting the sensitive data classification model and parameters until the output result meets the preset target.

[0029] Preferably, the setting of the neural network model and parameters to construct an initial sensitive data classification model includes: setting the number of nodes in the neural network input layer, the number of nodes in the output layer, the number of hidden layers, and the number of nodes in each hidden layer; setting the initial connection weights from the neural network input layer to the hidden layer, and from the hidden layer to the output layer, to construct an initial sensitive data classification model.

[0030] A method for classifying medical data comprises the following steps:

[0031] Step 1: Data receiving: used to receive data to be identified, the data to be identified comes from a database and / or a data terminal, the database is one or more of a plurality of classification sub-databases; the data terminal includes a mobile terminal and a desktop terminal;

[0032] Step 2: Data identification: used to identify the attributes of the data to be identified from the central processor, and store the identified data to be identified in the corresponding classification sub-database;

[0033] Step 3: Data screening: used to screen the criteria and screen the data to be processed from the data to be identified;

[0034] Step 4: Data classification: used to classify the data to be processed;

[0035] Step 5: Data classification: Based on the classification results of the data classification, the data to be processed is classified and value assessed.

[0036] Preferably, before the data screening step, the method further includes a user parameter input step; when the user parameter input step is activated, the data identification step is skipped.

[0037] (III) Beneficial effects

[0038] Compared with the prior art, the present invention provides a medical data classification system and classification method thereof, which have the following beneficial effects:

[0039] 1. This medical data classification system and its classification method can realize the management of patient information, diagnosis information and medication information, and can realize drug classification based on drug information and diagnosis information, providing an effective data basis for drug measurement calculation or big data analysis.

[0040] 2. A medical data classification system and classification method thereof, which constructs a sensitive data classification model; inputs sensitive data features into the sensitive data classification model to obtain sensitive data classification results; constructs a sensitive data grading model; inputs sensitive data classification results, sensitive data features and preset data security levels into the sensitive data grading model to obtain sensitive data grading results; and classifies and identifies sensitive data based on the sensitive data classification results, sensitive data features and sensitive data grading results. DETAILED DESCRIPTION

[0041] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] Embodiment 1:

[0043] A medical data classification system, comprising:

[0044] A cloud server, wherein a central processing unit is provided in the cloud server;

[0045] Data import system, used to import hospital patient information, diagnosis information and medication information;

[0046] The drug management system is used to maintain basic drug information and realize classified statistics of drug label information, name information, specification information, and unit information;

[0047] The drug classification system is used to establish a binary matrix of medication information and diagnosis information, and classify drugs based on the binary matrix;

[0048] The classification and grading system also includes: a data receiving subsystem, a data identification subsystem, a data screening subsystem, a data classification subsystem and a data grading subsystem;

[0049] The data receiving subsystem provides multiple APR interface terminals, through which the data receiving subsystem receives the data to be processed from multiple data sources, including databases and data terminals; the data identification subsystem is connected to the data receiving subsystem to identify the attributes of the data to be processed; the data screening subsystem receives the input screening criteria, and based on the screening criteria, screens the data to be processed from the data to be processed; the data classification subsystem is used to classify the data to be processed, and the data grading subsystem performs data grading and value assessment on the data to be processed based on the classification results;

[0050] Sensitive data system, used to analyze sensitive data and build relevant models.

[0051] In the embodiment of the present invention, after receiving the first data to be processed from the data terminal, the data receiving subsystem sends the first data to be processed to the data identifying subsystem.

[0052] In the embodiment of the present invention, the process of the drug classification module classifying drugs is as follows:

[0053] S1: Obtain diagnostic information and medication information, and pre-process the diagnostic information and medication information, including word form normalization and unified replacement of medication name naming methods;

[0054] S2: Based on the correspondence between the diagnosis information and the medication information, a binary matrix of the medication information and the diagnosis information is established;

[0055] S3: Based on the binary matrix of medication information and diagnosis information, the similarity distance between drugs is calculated, and the distance matrix between drugs is established;

[0056] S4: Classify drugs based on the distance matrix between drugs.

[0057] In an embodiment of the present invention, the sensitive data system includes:

[0058] Build a sensitive data classification model;

[0059] Input the sensitive data features into the sensitive data classification model to obtain the sensitive data classification results;

[0060] Build a sensitive data classification model;

[0061] Input the sensitive data classification results, sensitive data characteristics and preset data security levels into the sensitive data classification model to obtain the sensitive data classification results;

[0062] Sensitive data is classified and identified based on the sensitive data classification results, sensitive data characteristics and sensitive data grading results.

[0063] In an embodiment of the present invention, a sensitive data classification model is constructed, including: preprocessing sensitive data; dividing the preprocessed sensitive data into multiple sets, cyclically selecting any one set as a first test set, and the remaining sets as a first training set; setting a neural network model and parameters to construct an initial sensitive data classification model; training and testing the sensitive data classification model according to the first training set and the first test set, performing incremental neural network learning, and continuously adjusting the sensitive data classification model and parameters until the output result meets the preset target.

[0064] In an embodiment of the present invention, a neural network model and parameters are set to construct an initial sensitive data classification model, including: setting the number of nodes in the neural network input layer, the number of nodes in the output layer, the number of hidden layers, and the number of nodes in each hidden layer; setting the initial connection weights from the neural network input layer to the hidden layer, and from the hidden layer to the output layer, to construct an initial sensitive data classification model.

[0065] In an embodiment of the present invention, the sensitive data classification model is trained and tested according to the first training set and the first test set, incremental neural network learning is performed, and the sensitive data classification model and parameters are continuously adjusted until the output result meets the preset target, including: inputting the first training set into the sensitive data classification model, and training the sensitive data classification model multiple times;

[0066] Input the first test set into the sensitive data classification model after multiple trainings, and confirm whether the output results have classification errors or whether new classes need to be added based on the output results of the test set; if the classification is wrong, use the steepest descent method to adjust the weights of the sensitive data classification model until the classification results meet expectations;

[0067] If a new class is to be added, an output node is added to the original sensitive data classification model, the new connection weights from the hidden layer to the output layer are randomly initialized, and the number of new hidden layer nodes is confirmed by gradually adding hidden layer nodes. According to the actual results and expected results, the sensitive data classification model and parameters are adjusted to build a sensitive data grading model, including: taking the sensitive data classification results, sensitive data features and preset data security levels as the sample set, dividing the sample set into multiple sets, cyclically selecting any one set as the second test set, and the remaining sets as the second training sets.

[0068] Setting neural network model parameters and constructing an initial sensitive data classification model; training and testing the sensitive data classification model according to the second training set and the second test set, and continuously adjusting the sensitive data classification model parameters until the output result meets the preset target; training and testing the sensitive data classification model according to the second training set and the second test set, and continuously adjusting the sensitive data classification model parameters until the output result meets the preset target, including: inputting the second training set into the sensitive data classification model and training the sensitive data classification model multiple times; inputting the second test set into the sensitive data classification model after multiple trainings, and confirming whether the output result has a classification error based on the output result of the test set; if the classification is wrong, executing the steepest descent method to adjust the weights of the sensitive data classification model until the classification result meets expectations.

[0069] Embodiment 2:

[0070] A method for classifying medical data comprises the following steps:

[0071] Step 1: Data receiving: used to receive the data to be identified, the data to be identified comes from a database and / or a data terminal, the database is one or more of a plurality of classification sub-databases; the data terminal includes a mobile terminal and a desktop terminal;

[0072] Step 2: Data identification: used to identify the attributes of the data to be identified from the central processor, and store the identified data to be identified in the corresponding classification sub-database;

[0073] Step 3: Data screening: used to screen the criteria and select the data to be processed from the data to be identified;

[0074] Step 4: Data classification: used to classify the data to be processed;

[0075] Step 5: Data classification: Based on the classification results of data classification, data classification and value assessment are performed on the data to be processed.

[0076] The formula for calculating the similarity distance between drugs is:

[0077]

[0078] Among them, D(di, dj) is the similarity distance between drug i and requirement j, di is the binary matrix vector of drug i's medication information and diagnosis information, and dj is the binary matrix vector of drug j's medication information and diagnosis information.

[0079] In an embodiment of the present invention, before the data screening step, the method further includes a user parameter input step; when the user parameter input step is activated, the data identification step is skipped.

[0080] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A medical data classification system, characterized in that: include: A cloud server, wherein a central processing unit is provided in the cloud server; Data import system, used to import hospital patient information, diagnosis information and medication information; The drug management system is used to maintain basic drug information and realize classified statistics of drug label information, name information, specification information, and unit information; The drug classification system is used to establish a binary matrix of medication information and diagnosis information, and classify drugs based on the binary matrix; The classification and grading system also includes: a data receiving subsystem, a data identification subsystem, a data screening subsystem, a data classification subsystem and a data grading subsystem; The data receiving subsystem provides a plurality of APR interface terminals, through which the data receiving subsystem receives the data to be processed from a plurality of data sources, the plurality of data sources including a database and a data terminal; the data identification subsystem is connected to the data receiving subsystem to identify the attributes of the data to be processed; the data screening subsystem receives input screening criteria, and based on the screening criteria, screens the data to be processed from the data to be processed; the data classification subsystem is used to classify the data to be processed, and the data grading subsystem performs data grading and value assessment on the data to be processed based on the classification results; Sensitive data system, used to analyze sensitive data and build relevant models.

2. A medical data classification system according to claim 1, characterized in that: After receiving the first data to be processed from the data terminal, the data receiving subsystem sends the first data to be processed to the data identifying subsystem.

3. A medical data classification system according to claim 2, characterized in that: The process of classifying drugs by the drug classification module is as follows: S1: Obtain diagnostic information and medication information, and pre-process the diagnostic information and medication information, including word form normalization and unified replacement of medication name naming methods; S2: Based on the correspondence between the diagnosis information and the medication information, a binary matrix of the medication information and the diagnosis information is established; S3: Based on the binary matrix of medication information and diagnosis information, the similarity distance between drugs is calculated, and the distance matrix between drugs is established; S4: Classify drugs based on the distance matrix between drugs.

4. A medical data classification system according to claim 1, characterized in that: The sensitive data systems include: Build a sensitive data classification model; Inputting the sensitive data features into the sensitive data classification model to obtain a sensitive data classification result; Build a sensitive data classification model; Inputting the sensitive data classification result, the sensitive data characteristics and the preset data security level into the sensitive data classification model to obtain a sensitive data classification result; The sensitive data is classified and identified according to the sensitive data classification result, the sensitive data characteristics and the sensitive data grading result.

5. A medical data classification system according to claim 4, characterized in that: The construction of the sensitive data classification model includes: preprocessing the sensitive data; dividing the preprocessed sensitive data into multiple sets, cyclically selecting any one set as a first test set, and the remaining sets as a first training set; setting a neural network model and parameters to construct an initial sensitive data classification model; training and testing the sensitive data classification model according to the first training set and the first test set, performing incremental neural network learning, and continuously adjusting the sensitive data classification model and parameters until the output result meets the preset target.

6. A medical data classification system according to claim 5, characterized in that: The setting of the neural network model and parameters to construct an initial sensitive data classification model includes: setting the number of nodes in the neural network input layer, the number of nodes in the output layer, the number of hidden layers, and the number of nodes in each hidden layer; setting the initial connection weights from the neural network input layer to the hidden layer, and from the hidden layer to the output layer, to construct an initial sensitive data classification model.

7. A method for classifying medical data, characterized in that: The following steps are involved: Step 1: Data receiving: used to receive data to be identified, the data to be identified comes from a database and / or a data terminal, the database is one or more of a plurality of classification sub-databases; the data terminal includes a mobile terminal and a desktop terminal; Step 2: Data identification: used to identify the attributes of the data to be identified from the central processor, and store the identified data to be identified in the corresponding classification sub-database; Step 3: Data screening: used to screen the criteria and screen the data to be processed from the data to be identified; Step 4: Data classification: used to classify the data to be processed; Step 5: Data classification: Based on the classification results of the data classification, the data to be processed is classified and value assessed.

8. A medical data classification method according to claim 7, characterized in that: Before the data screening step, the method further includes a user parameter input step; when the user parameter input step is activated, the data identification step is skipped.

Images