WO2019091111A1 - 管控药物的特征获取方法、电子装置及计算机可读存储介质 - Google Patents

管控药物的特征获取方法、电子装置及计算机可读存储介质 Download PDF

Info

Publication number
WO2019091111A1
WO2019091111A1 PCT/CN2018/090906 CN2018090906W WO2019091111A1 WO 2019091111 A1 WO2019091111 A1 WO 2019091111A1 CN 2018090906 W CN2018090906 W CN 2018090906W WO 2019091111 A1 WO2019091111 A1 WO 2019091111A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
controlled drug
feature
controlled
model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2018/090906
Other languages
English (en)
French (fr)
Inventor
周瑜
阮晓雯
徐亮
肖京
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An Technology Shenzhen Co Ltd
Original Assignee
Ping An Technology Shenzhen Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ping An Technology Shenzhen Co Ltd filed Critical Ping An Technology Shenzhen Co Ltd
Publication of WO2019091111A1 publication Critical patent/WO2019091111A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
    • G16H20/10—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Definitions

  • the present application relates to the field of communications technologies, and in particular, to a method for acquiring features of a controlled drug, an electronic device, and a computer readable storage medium.
  • the present application provides a method for acquiring characteristics of a controlled drug, an electronic device, and a computer readable storage medium.
  • the present application provides a method for acquiring a feature of a controlled drug, which is applied to an electronic device, and the method includes:
  • a feature table of the specific controlled drug is returned by the feature extraction model.
  • the present application further provides an electronic device including a memory and a processor, wherein the memory stores a feature acquiring system of a controlled drug that can be run on the processor, and the characteristics of the controlled drug The step of implementing the feature acquisition method of the controlled drug as described above when the system is executed by the processor.
  • the present application further provides a computer readable storage medium, wherein the computer readable storage medium stores a feature acquisition system for a controlled drug, and the feature acquisition system of the controlled drug can be at least one processor. Executing to cause the at least one processor to perform the steps of the feature acquisition method of the controlled drug as described above.
  • the electronic device, the feature acquiring method of the controlled drug, and the computer readable storage medium proposed by the present application first obtain the original case data of the insurance institution database and the medical institution database; and then, the original case The data is pre-processed to obtain a pre-processed data set; secondly, a classification model is established according to the forward characteristics, the nature characteristics and the backward characteristics of the controlled drugs; and then, the data of the controlled drugs is extracted by the classification model; Then, a specific controlled drug feature extraction model is established according to the unique characteristics of the specific controlled drug; finally, the feature table of the specific controlled drug is returned by the feature extraction model.
  • the feature acquisition method of the controlled drug and the computer readable storage medium proposed by the present application the analysis of the big data can quickly acquire the feature sets of various controlled drugs, and greatly improve the feature extraction of different controlled drugs in different regions and regions. Speed and accuracy are more convenient, accurate and faster than the prior art.
  • FIG. 1 is a schematic diagram of an optional application environment of each embodiment of the present application.
  • FIG. 2 is a schematic diagram of an optional hardware architecture of an electronic device of the present application.
  • FIG. 3 is a schematic diagram of a program module of a first embodiment of a feature acquisition system for a controlled drug of the present application
  • FIG. 4 is a schematic flow chart of a first embodiment of a method for acquiring characteristics of a controlled drug of the present application
  • FIG. 5 is a schematic flow chart of a second embodiment of a method for acquiring characteristics of a controlled drug of the present application
  • FIG. 6 is a schematic flow chart of a third embodiment of a method for acquiring characteristics of a controlled drug of the present application
  • FIG. 7 is a schematic flow chart of a fourth embodiment of a method for acquiring characteristics of a controlled drug of the present application.
  • FIG. 8 is a schematic flow chart of a fifth embodiment of a method for acquiring characteristics of a controlled drug of the present application.
  • FIG. 1 it is a schematic diagram of an optional application environment of each embodiment of the present application.
  • the present application is applicable to an application environment 1, which includes, but is not limited to, an insurance institution database 10 (referred to as “insurance institution” in FIG. 1) and a medical institution database 11 (in FIG. 1).
  • an insurance institution database 10 referred to as "insurance institution” in FIG. 1
  • a medical institution database 11 in FIG. 1
  • the network 12 and the electronic device 13 it should be noted that in other embodiments, the insurance institution database 10 and the medical institution database 11 may be an insurance institution server or a medical institution server.
  • the electronic device 13 may be a mobile device, a smart phone, a notebook computer, a digital broadcast receiver, a PDA (Personal Digital Assistant), a PAD (Tablet), etc., and a desktop device such as a desktop computer.
  • a fixed terminal such as a notebook, a server, or the like.
  • the electronic device 13 is used as an application server.
  • the application server may be a computing device such as a rack server, a blade server, a tower server, or a rack server.
  • the application server may be a stand-alone server or a server cluster composed of multiple servers.
  • the database the implementation of different professional companies are different, the main database type is Oracle, there will also be various types of databases such as PostgreSQL, MySQL.
  • the network 12 may be an intranet, an Internet, a Global System of Mobile communication (GSM), a Wideband Code Division Multiple Access (WCDMA), a 4G network, Wireless or wired networks such as 5G networks, Bluetooth, Wi-Fi, and call networks.
  • the insurance institution database 10 and the medical institution database 11 are communicatively coupled to one or more of the electronic devices 13 (only one of which is shown) through the network 12 to enable the electronic device 13 to pass
  • the network 12 performs data transmission and interaction with the insurance institution database 10 and the medical institution database 11.
  • FIG. 2 it is a schematic diagram of an optional hardware architecture of the electronic device 13 of the present application.
  • the electronic device 13 may include, but is not limited to, a memory 130, a network interface 131, and a processor 132 that are communicably connected to each other through a system bus. It is noted that FIG. 2 only shows the electronic device 13 having the components 130-132, but it should be understood that not all illustrated components may be implemented, and more or fewer components may be implemented instead.
  • the memory 130 includes at least one type of readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (eg, SD or DX memory, etc.), random access Memory (RAM), static random access memory (SRAM), read only memory (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), magnetic memory, magnetic disk, optical disk, and the like.
  • the memory 130 may be an internal storage unit of the electronic device 13, such as a hard disk or a memory of the electronic device 13.
  • the memory 130 may also be an external storage device of the electronic device 13, such as a plug-in hard disk equipped on the electronic device 13, a smart memory card (SMC), and a secure digital device. (Secure Digital, SD) card, flash card, etc.
  • SMC smart memory card
  • SD Secure Digital
  • the memory 130 may also include both an internal storage unit of the electronic device 13 and an external storage device thereof.
  • the memory 130 is generally used to store an operating system installed in the electronic device 13 and various types of application software, such as program codes of the feature acquisition system 200 for controlling drugs.
  • the memory 130 can also be used to temporarily store various types of data that have been output or are to be output.
  • the network interface 131 may include a wireless network interface or a wired network interface, which is generally used to establish a communication connection between the electronic device 13 and other electronic devices.
  • the processor 132 may be a Central Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip in some embodiments.
  • the processor 132 is typically used to control the overall operation of the electronic device 13.
  • the processor 132 is configured to run program code or processing data stored in the memory 130, such as the feature acquiring system 200 that runs the controlled drug.
  • the present application proposes a feature acquisition system 200 for a controlled drug.
  • FIG. 3 it is a program module diagram of the first embodiment of the feature acquisition system 200 of the present invention.
  • the controlled drug feature acquisition system 200 includes a series of computer program instructions stored on the memory 130 that, when executed by the processor 132, can implement the features of various embodiments of the present application. Get the operation.
  • the feature acquisition system 200 that governs medications can be divided into one or more modules based on the particular operations implemented by the various portions of the computer program instructions. For example, in FIG. 3, the feature acquisition system 200 of the controlled drug may be divided into an acquisition module 201, a processing module 202, an establishment module 203, a calling module 204, a storage module 205, and a display module 206. among them:
  • the obtaining module 201 is configured to obtain original case data of the insurance institution database 10 and the medical institution database 11.
  • the insurance institution database 10 can be a database of the insurance company of the city.
  • the case data in the database mainly includes a policy and a receipt
  • the medical institution database 11 can be For the database/server of hospitals and medical centers in the city, according to the different medical insurance policies, the original case data of hospitals, medical centers, etc. including provinces, municipalities, districts and other regions can be obtained according to the needs.
  • the processing module 202 is configured to perform pre-processing on the original case data to obtain a pre-processed data set.
  • the preprocessing method includes:
  • the unified quantification method is mainly for enumerated data, such as the patient's response to specific viruses and diseases (immunity, high resistance, resistance, sensation).
  • the unified unit of measurement is mainly for numerical data, such as the dose use of the controlled drug. , package, gram, milliliter, etc. as a unit, the classification of data of different nature uses a unified unit of measurement, for example, the measurement of milliliters, liters is uniformly measured in milliliters as a unit of measurement, and the measurement of milligrams, grams, kilograms is uniformly measured in milligrams.
  • unified performance form is mainly for multi-expression data, such as date data, which can be expressed as YYYY-MM-DD or other forms such as MM-DD-YYYY, which is unified by YYYY-MM-DD. Formal identity.
  • the cleaning model includes a plurality of judgment criteria, for example, clearing garbled data in the original data, clearing data that is obviously not in accordance with common sense or obviously comparing the preset conditions, and clearing data irrelevant to the influence of the controlled drug use. Clear the duplicate data and clear the data of patients with serious data loss;
  • a filter is created based on each criterion, and the data is filtered synchronously/asynchronously with a plurality of said filters.
  • the establishing module 203 is configured to establish a classification model according to the forward characteristics, the nature features, and the backward characteristics of the controlled drugs.
  • the classification model is established based on the time dimension, and the reference point is selected at a specific time point, and the data acquired at the reference point is used as a natural feature, and the data before the reference point is used as a forward feature, and the data obtained after the reference point is used.
  • a backward feature As a backward feature
  • data of the general characteristics of the controlled drugs in different time periods are selected as training samples to train the classification model
  • the classification model may provide a set of attributes for the decision tree classifier, and the decision tree model may classify the data by making a series of decisions based on the attribute set; selecting the tree classifier, selecting the tree classifier to use and Similar techniques for decision tree classifiers classify data.
  • the selection tree contains special selection nodes, and the selection node has multiple branches.
  • artificial neural networks case inferences, nearest neighbor methods, support vector machines, and random forests can be used.
  • the calling module 204 is configured to extract data of the controlled medication by using the classification model.
  • the data classified by the classification model has a label for controlling drugs, for example, "insulin”, “glucose” and the like are extracted, and at this time, all the data are data related to the controlled drugs. ;
  • a controlled drug database can be established for the extracted data to facilitate subsequent use and query.
  • the establishing module 203 is further configured to establish a specific controlled drug feature extraction model according to the unique characteristics of the specific controlled drug.
  • different controlled drugs have different characteristics, including but not limited to application scenarios, reimbursement conditions and reimbursement quotas according to medical insurance policies, drug use status, specificity of different places, etc., according to different controlled drugs
  • the unique characteristics of the specific controlled drug feature extraction model can be used to extract the characteristics of the specific controlled drug.
  • diabetes has evolved from a geriatric disease to a disease that is common to children, young people, and the elderly. Although it is the same disease, the management of different age groups is different. For example, different time for disease management, different time for treatment, different physical conditions, different blood glucose management standards, because of these differences, different ages It has different characteristics in the use of insulin. Therefore, it is possible to establish the characteristic extraction basis of insulin based on age as a characteristic feature. Based on age, the expression is established according to the different usage conditions of different ages of insulin, and the feature extraction model is established based on the expression. According to the feature extraction model, all insulin-related data in the controlled drug database can be found.
  • the feature extraction model may be a classification model, such as a two-category model, and the two-category model is defined to include two categories, a positive category and a negative category, respectively representing specific controlled drugs and other drugs, such as a positive class. Insulin, the negative class is other controlled drugs.
  • the two-category model is used to predict (classify) the test samples in the controlled drug database, and sometimes there are cases of classification errors. For example, when the characteristic features of the two controlled drugs appear intersection, the feature extraction model needs to be corrected. For example, a classification scoring model can be added.
  • each object in the positive class is referred to as a positive instance
  • each object in the negative class is referred to as a negative instance.
  • the situation arises as follows: If an instance is a positive class and is predicted to be a positive class, it is called a true positive (TP), and if the instance is a negative class is predicted to be a positive class, it is called a false positive class (False postiVe , FP).
  • TP true positive
  • FP false positive class
  • the instance is a negative class is predicted to be a negative class
  • TN true negative class
  • a positive instance is predicted to be a negative class (false negative, FN).
  • TP positive instance prediction is positive class number
  • FN positive instance prediction is negative class number
  • FP negative instance prediction The number of positive classes
  • TN The negative instance is predicted to be the number of negative classes.
  • the classification scoring model can also calculate the following three parameters: sensitivity: the proportion of instances in the positive class that are correctly predicted to be positive, ie TP/(TP+FN). Specificity: The proportion of instances in the negative class that are correctly predicted to be negative, ie TN/(TN+FP). Positive Predictive Value (PPV): The proportion of positive instances in the instance predicted to be positive, ie TP/(TP+FP).
  • the classification results can be evaluated. According to the evaluation results, the classification situation can be understood and the feature extraction model can be optimized.
  • the invoking module 204 is further configured to return a feature table of the specific controlled drug by using the feature extraction model.
  • the features affecting the use of a particular controlled drug include a plurality of features, and a profile of a particular controlled drug can be established based on the data returned by the feature, including a plurality of influencing factors for the use of the particular controlled drug.
  • the data in the feature table may include information about patient-predicted pre-existing parameters, including patient condition parameters, such as a history of drug allergy or sensitivity, other currently administered drugs in the patient's tissue.
  • patient condition parameters such as blood pressure, heart rate, heart rate, temperature, blood oxygen, respiratory rate, or ventilation frequency
  • Parameters usage of controlled drugs, frequency of use, cost, time of drug use, current drug, drug class, drug allergy history and sensitivity
  • comorbid case data including acute comorbidities and chronic comorbidities, eg acute comorbidities Including hypoglycemia and hyperglycemia, chronic complications include eye lesions, kidney disease, neuropathy, cardiovascular disease, and foot lesions.
  • the storage module 205 is configured to store data acquired by the acquisition module 201, the processing module 202, the establishment module 203, and the calling module 204.
  • the storage module 205 can be used to store a feature table of a specific controlled drug.
  • the storage module 205 includes a readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (eg, SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), magnetic memory, magnetic disk, optical disk, and the like.
  • the display module 206 is configured to display an intermediate result and a final result in the working process of the feature acquiring system 200 of the controlled drug.
  • the display module 206 includes a display device such as an LCD or an LED, and the display module 206 can be used to display the specific control. A list of the characteristics of the drug.
  • the present application also proposes a method for acquiring characteristics of a controlled drug.
  • FIG. 4 it is a schematic flowchart of the first embodiment of the method for acquiring characteristics of the controlled drug of the present application.
  • the order of execution of the steps in the flowchart shown in FIG. 4 may be changed according to different requirements, and some steps may be omitted.
  • step S110 the original case data of the insurance institution database 10 and the medical institution database 11 are obtained.
  • the insurance institution database 10 can be a database of the insurance company of the city.
  • the case data in the database mainly includes a policy and a receipt
  • the medical institution database 11 can be For the hospitals and medical centers of this city, according to the different medical insurance policies, the original case data of hospitals, medical centers, etc. including provinces, municipalities, districts and other regions can be obtained as needed.
  • step S120 the original case data is preprocessed to obtain a preprocessed data set.
  • the preprocessing manner includes sorting, normalizing, denoising, etc. the original case data in chronological order.
  • Step S130 establishing a classification model according to the forward characteristics, the nature characteristics and the backward characteristics of the controlled drugs.
  • the classification model is established based on the time dimension, and the reference point is selected at a specific time point, and the data acquired at the reference point is used as a natural feature, and the data before the reference point is used as a forward feature, and the data obtained after the reference point is used.
  • a backward feature As a backward feature
  • data of the general characteristics of the controlled drugs in different time periods are selected as training samples to train the classification model
  • the classification model may provide a set of attributes for the decision tree classifier, and the decision tree model may classify the data by making a series of decisions based on the attribute set; selecting the tree classifier, selecting the tree classifier to use and Similar techniques for decision tree classifiers classify data.
  • the selection tree contains special selection nodes, and the selection node has multiple branches.
  • artificial neural networks case inferences, nearest neighbor methods, support vector machines, and random forests can be used.
  • Step S140 extracting data of the controlled drug by using the classification model.
  • the data classified by the classification model has a label for controlling drugs, for example, "insulin”, “glucose” and the like are extracted, and at this time, all the data are data related to the controlled drugs. ;
  • Step S150 establishing a specific control drug feature extraction model according to the unique characteristics of the specific controlled drug.
  • different controlled drugs have different characteristics, including but not limited to application scenarios, reimbursement conditions and reimbursement quotas according to medical insurance policies, drug use status, specificity of different places, etc., according to different controlled drugs
  • the unique characteristics of the specific controlled drug feature extraction model can be used to extract the characteristics of the specific controlled drug.
  • diabetes has evolved from a geriatric disease to a disease that is common to children, young people, and the elderly. Although it is the same disease, the management of different age groups is different. For example, different time for disease management, different time for treatment, different physical conditions, different blood glucose management standards, because of these differences, different ages It has different characteristics in the use of insulin. Therefore, it is possible to establish the characteristic extraction basis of insulin based on age as a characteristic feature. Based on age, the expression is established according to the different usage conditions of different ages of insulin, and the feature extraction model is established based on the expression. According to the feature extraction model, all insulin-related data in the controlled drug database can be found.
  • the feature extraction model may be a classification model, such as a two-category model, and the two-category model is defined to include two categories, a positive category and a negative category, respectively representing specific controlled drugs and other drugs, such as a positive class. Insulin, the negative class is other controlled drugs.
  • Step S160 returning a feature table of the specific controlled drug by the feature extraction model.
  • the features affecting the use of a particular controlled drug include a plurality of features, and a profile of a particular controlled drug can be established based on the data returned by the feature, including a plurality of influencing factors for the use of the particular controlled drug.
  • the data in the feature table may include information about patient-predicted pre-existing parameters, including patient condition parameters, such as a history of drug allergy or sensitivity, other currently administered drugs in the patient's tissue.
  • patient condition parameters such as blood pressure, heart rate, heart rate, temperature, blood oxygen, respiratory rate, or ventilation frequency
  • Parameters usage of controlled drugs, frequency of use, cost, time of drug use, current drug, drug class, drug allergy history and sensitivity
  • comorbid case data including acute comorbidities and chronic comorbidities, eg acute comorbidities Including hypoglycemia and hyperglycemia, chronic complications include eye lesions, kidney disease, neuropathy, cardiovascular disease, and foot lesions.
  • the pre-processing step further includes the following steps:
  • step S210 the original case data is sorted in chronological order.
  • step S220 the sorted data is normalized.
  • the normalization process includes: a unified quantization method, a unified measurement unit, and a unified expression form.
  • the unified quantification method is mainly for enumerated data, such as the patient's response to specific viruses and diseases (immunity, high resistance, resistance, sensation).
  • the unified unit of measurement is mainly for numerical data, such as the dose use of the controlled drug. , package, gram, milliliter, etc. as a unit, the classification of data of different nature uses a unified unit of measurement, for example, the measurement of milliliters, liters is uniformly measured in milliliters as a unit of measurement, and the measurement of milligrams, grams, kilograms is uniformly measured in milligrams.
  • unified performance form is mainly for multi-expression data, such as date data, which can be expressed as YYYY-MM-DD or other forms such as MM-DD-YYYY, which is unified by YYYY-MM-DD. Formal identity.
  • step S230 the normalized data is denoised.
  • the denoising includes establishing a cleaning model to filter data, and deleting data that does not meet the criteria for judging:
  • the cleaning model includes a plurality of judgment criteria, for example, clearing garbled data in the original data, clearing data that is obviously not in accordance with common sense or obviously comparing the preset conditions, and clearing data irrelevant to the influence of the controlled drug use. Clear the duplicate data and clear the data of patients with severe data loss.
  • a filter is established according to each criterion, and data is filtered synchronously/asynchronously with a plurality of said filters.
  • Feature acquisition method for controlled drugs feature acquisition method for controlled drugs; feature acquisition method for controlled drugs; feature acquisition method for controlled drugs; feature acquisition method for controlled drugs; feature acquisition method for controlled drugs
  • FIG. 6 is a schematic flow chart of a third embodiment of a method for acquiring characteristics of a controlled drug of the present application.
  • the step S150 of the first embodiment of the method for acquiring the characteristics of the controlled drug includes the following steps:
  • Step S510 establishing a two-category model according to the unique characteristics of the specific controlled drug.
  • the two-category model is defined to include two categories, a positive class and a negative class, respectively representing specific controlled drugs and other drugs, such as a positive class of insulin and a negative class of other controlled drugs.
  • Step S520 classifying the test samples in the controlled drug database using the two-category model.
  • each object in the positive class is called a positive instance
  • each object in the negative class is called a negative instance.
  • TP true positive
  • FP positive class
  • TN true negative class
  • FN negative class
  • step S530 the classification result is evaluated using a classification scoring model to optimize the two classification model.
  • a classification error occurs, for example, when the unique characteristics of the two controlled drugs appear intersection, the feature extraction model needs to be performed at this time. Corrected. For example, a classification scoring model can be added.
  • TP positive instance prediction is positive number
  • FN positive instance prediction is negative class number
  • FP negative instance prediction The number of positive classes
  • TN The negative instance is predicted to be the number of negative classes.
  • the classification scoring model can also calculate the following three parameters: sensitivity: the proportion of instances in the positive class that are correctly predicted to be positive, ie TP/(TP+FN). Specificity: The proportion of instances in the negative class that are correctly predicted to be negative, ie TN/(TN+FP). Positive Predictive Value (PPV): The proportion of positive instances in the instance predicted to be positive, ie TP/(TP+FP).
  • the classification results can be evaluated. According to the evaluation results, the classification situation can be understood, and the feature extraction model (two classification model) can be optimized.
  • FIG. 7 is a schematic flow chart of a fourth embodiment of a method for acquiring characteristics of a controlled drug of the present application.
  • the method for acquiring the characteristics of the controlled drug further includes the following steps after the step S140 of the first embodiment:
  • Step S610 establishing a controlled drug database for the extracted data.
  • a controlled drug database can be established for the extracted data to facilitate subsequent use and query.
  • the database the implementation of different professional companies are different, the main database type is Oracle, there will also be various types of databases such as PostgreSQL, MySQL.
  • FIG. 8 it is a schematic flowchart of a fifth embodiment of a method for acquiring characteristics of a controlled drug of the present application.
  • the steps S710-S760 of the feature acquisition method of the controlled drug are similar to the steps S110-S160 of the first embodiment, except that the method further includes steps S770-S780.
  • the order of execution of the steps in the flowchart shown in FIG. 10 may be changed according to different requirements, and some steps may be omitted.
  • step S710 the original case data of the insurance institution database 10 and the medical institution database 11 are obtained.
  • Step S720 preprocessing the original case data to obtain a preprocessed data set.
  • Step S730 establishing a classification model according to the forward characteristics, the nature characteristics and the backward characteristics of the controlled drugs.
  • Step S740 extracting data of the controlled drug by using the classification model.
  • Step S750 establishing a specific control drug feature extraction model according to the unique characteristics of the specific controlled drug.
  • Step S760 returning a feature table of the specific controlled drug by the feature extraction model.
  • Step S770 storing the feature table.
  • the storage module 205 can be used to store a feature table of a particular controlled drug.
  • the storage module 205 includes a readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (eg, SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), magnetic memory, magnetic disk, optical disk, and the like.
  • a readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (eg, SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), magnetic memory, magnetic disk, optical disk, and the like.
  • ROM read only memory
  • Step S780 displaying the feature table to the user.
  • the intermediate result and the final result in the working process of the feature acquisition system 200 of the controlled drug are displayed to the user through the display device.
  • the electronic device, the feature acquiring method of the controlled drug, and the computer readable storage medium proposed by the present application first obtain the original case data of the insurance institution database and the medical institution database; and then, the original case The data is pre-processed to obtain a pre-processed data set; secondly, a classification model is established according to the forward characteristics, the nature characteristics and the backward characteristics of the controlled drugs; and then, the data of the controlled drugs is extracted by the classification model; Then, a specific controlled drug feature extraction model is established according to the unique characteristics of the specific controlled drug; finally, the feature table of the specific controlled drug is returned by the feature extraction model.
  • the feature acquisition method of the controlled drug and the computer readable storage medium proposed by the present application the analysis of the big data can quickly acquire the feature sets of various controlled drugs, and greatly improve the feature extraction of different controlled drugs in different regions and regions. Speed and accuracy are more convenient, accurate and faster than the prior art.
  • the technical solution of the present application which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM, disk,
  • a storage medium such as ROM/RAM, disk
  • the optical disc includes a number of instructions for causing a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the methods described in the various embodiments of the present application.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • Epidemiology (AREA)
  • Data Mining & Analysis (AREA)
  • Primary Health Care (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Pathology (AREA)
  • Databases & Information Systems (AREA)
  • Chemical & Material Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Medicinal Chemistry (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

本申请公开了一种管控药物的特征获取方法，应用于电子装置，所述方法包括：获取保险机构数据库及医疗机构数据库的原始病例数据；对所述原始病例数据进行预处理，得到预处理后的数据集；根据管控药物的前向性特征、本性特征及后向性特征建立分类模型；通过所述分类模型提取所述管控药物的数据；根据特定管控药物的特有特征建立特定管控药物特征提取模型；及通过所述特征提取模型返回所述特定管控药物的特征表。本申请还提供一种电子装置及计算机可读存储介质。本申请通过对大数据进行分析能够快速获取各种管控药物的特征集合，极大地提高不同地区、不同管控药物的特征提取速度、准确性。

Description

管控药物的特征获取方法、电子装置及计算机可读存储介质
本申请要求于2017年11月10日提交中国专利局，申请号为201711107651.0、发明名称为“管控药物的特征获取方法、电子装置及计算机可读存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及通信技术领域，尤其涉及一种管控药物的特征获取方法、电子装置及计算机可读存储介质。
背景技术
随着信息技术的告诉发展，行业数据量急剧增长，对于医药、气象等需要存储大量的检测信息的行业来说，更是面临这对大数据进行快速分析的挑战。例如，在医药行业，很多的药物因为具有一定的危害而被严格管控，对这些管控药物使用情况常常难以把握。目前管控药物违规的筛选，主要根据相应业务逻辑而来，先根据使用规则，抽取合规单据，再做减法得到违规单据。但不同药品对应的业务逻辑几乎都不大相同，而且不同地区管控药名单之间还存在一些差异，这样就导致切换到不同地区，不同管控药物在提取特征的时候往往容易忽略、轻视一些重要性特征。
因此，亟需提出一种方法快速、方便地获取各种管控药物的特征。
发明内容
有鉴于此，本申请提出一种管控药物的特征获取方法、电子装置及计算机可读存储介质，通过对大数据进行分析能够快速获取各种管控药物的特征集合，极大地提高不同地区、不同管控药物的特征提取速度、准确性。
首先，为实现上述目的，本申请提出一种管控药物的特征获取方法，该方法应用于电子装置，所述方法包括：
获取保险机构数据库及医疗机构数据库的原始病例数据；
对所述原始病例数据进行预处理，得到预处理后的数据集；
根据管控药物的前向性特征、本性特征及后向性特征建立分类模型；
通过所述分类模型提取所述管控药物的数据；
根据特定管控药物的特有特征建立特定管控药物特征提取模型；及
通过所述特征提取模型返回特定管控药物的特征表。
此外，为实现上述目的，本申请还提供一种电子装置，包括存储器、处理器，所述存储器上存储有可在所述处理器上运行的管控药物的特征获取系统，所述管控药物的特征获取系统被所述处理器执行时实现如上述的管控药物的特征获取方法的步骤。
进一步地，为实现上述目的，本申请还提供一种计算机可读存储介质， 所述计算机可读存储介质存储有管控药物的特征获取系统，所述管控药物的特征获取系统可被至少一个处理器执行，以使所述至少一个处理器执行如上述的管控药物的特征获取方法的步骤。
相较于现有技术，本申请所提出的电子装置、管控药物的特征获取方法及计算机可读存储介质，首先，获取保险机构数据库及医疗机构数据库的原始病例数据；然后，对所述原始病例数据进行预处理，得到预处理后的数据集；其次，根据管控药物的前向性特征、本性特征及后向性特征建立分类模型；之后，通过所述分类模型提取所述管控药物的数据；然后，根据特定管控药物的特有特征建立特定管控药物特征提取模型；最后，通过所述特征提取模型返回特定管控药物的特征表。通过本申请所提出的电子装置、管控药物的特征获取方法及计算机可读存储介质，对大数据进行分析能够快速获取各种管控药物的特征集合，极大地提高不同地区、不同管控药物的特征提取速度、准确性，相对于现有技术更见便捷、准确、迅速。
附图说明
图1是本申请各个实施方式一可选的应用环境示意图；
图2是本申请电子装置一可选的硬件架构的示意图；
图3是本申请管控药物的特征获取系统第一实施方式的程序模块示意图；
图4是本申请管控药物的特征获取方法第一实施方式的流程示意图；
图5是本申请管控药物的特征获取方法第二实施方式的流程示意图；
图6是本申请管控药物的特征获取方法第三实施方式的流程示意图；
图7是本申请管控药物的特征获取方法第四实施方式的流程示意图；
图8是本申请管控药物的特征获取方法第五实施方式的流程示意图。
本申请目的的实现、功能特点及优点将结合实施方式，参照附图做进一步说明。
具体实施方式
为了使本申请的目的、技术方案及优点更加清楚明白，以下结合附图及实施方式，对本申请进行进一步详细说明。应当理解，此处所描述的具体实施方式仅用以解释本申请，并不用于限定本申请。基于本申请中的实施方式，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施方式，都属于本申请保护的范围。
需要说明的是，在本申请中涉及“第一”、“第二”等的描述仅用于描述目的，而不能理解为指示或暗示其相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括至少一个该特征。另外，各个实施方式之间的技术方案可以相互结合，但是必须是以本领域普通技术人员能够实现为基础，当技术方案的结合出现相互矛盾或无法实现时应当认为这种技术方案的结合不存在，也不在本申请要求的保护范围之内。
参阅图1所示，是本申请各个实施方式一可选的应用环境示意图。
在本实施方式中，本申请可应用于应用环境1，所述应用环境1包括，但不仅限于，保险机构数据库10(图1中简称为“保险机构”)、医疗机构数据库11(图1中简称为“医疗机构”)、网络12及电子装置13，需要说明的是，在其他实施方式中，保险机构数据库10、医疗机构数据库11也可以是保险机构服务器、医疗机构服务器。
在一实施方式中，所述电子装置13可以是移动电话、智能电话、笔记本电脑、数字广播接收器、PDA(个人数字助理)、PAD(平板电脑)等等的可移动设备，以及诸如台式计算机、笔记本、服务器等等的固定终端，本实施方式以电子装置13为应用服务器进行说明。所述应用服务器可以是机架式服务器、刀片式服务器、塔式服务器或机柜式服务器等计算设备，该应用服务器可以是独立的服务器，也可以是多个服务器所组成的服务器集群。所述数据库，各专业公司的实现方式不同，主要的数据库类型为Oracle，也会存在PostgreSQL、MySQL等类型的各种数据库。所述网络12可以是企业内部网(Intranet)、互联网(Internet)、全球移动通讯系统(Global System of Mobile communication，GSM)、宽带码分多址(Wideband Code Division Multiple Access，WCDMA)、4G网络、5G网络、蓝牙(Bluetooth)、Wi-Fi、通话网络等无线或有线网络。
在一实施方式中，所述保险机构数据库10和医疗机构数据库11通过所述网络12与一个或多个所述电子装置13(图中仅示出一个)通信连接，以使电子装置13能够通过网络12与所述保险机构数据库10和医疗机构数据库11进行数据传输和交互。
参阅图2所示，是本申请电子装置13一可选的硬件架构的示意图。
在一实施方式中，所述电子装置13可包括，但不仅限于，可通过系统总线相互通信连接存储器130、网络接口131、处理器132。需要指出的是，图2仅示出了具有组件130-132的电子装置13，但是应理解的是，并不要求实施所有示出的组件，可以替代的实施更多或者更少的组件。
在一实施方式中，所述存储器130至少包括一种类型的可读存储介质，所述可读存储介质包括闪存、硬盘、多媒体卡、卡型存储器(例如，SD或DX存储器等)、随机访问存储器(RAM)、静态随机访问存储器(SRAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、可编程只读存储器(PROM)、磁性存储器、磁盘、光盘等。在一些实施方式中，所述存储器130可以是所述电子装置13的内部存储单元，例如该电子装置13的硬盘或内存。在另一些实施方式中，所述存储器130也可以是所述电子装置13的外部存储设备，例如该电子装置13上配备的插接式硬盘，智能存储卡(Smart Media Card,SMC)，安全数字(Secure Digital,SD)卡，闪存卡(Flash Card)等。当然，所述存储器130还可以既包括所述电子装置13的内部存储单元也包括其外部存储设备。本实施方式中，所述存储器130通常用于存储安装于所述电子装置13的操作系统和各类应用软件，例如管控药物的特征获取系统200的程序代码 等。此外，所述存储器130还可以用于暂时地存储已经输出或者将要输出的各类数据。
在一实施方式中，所述网络接口131可包括无线网络接口或有线网络接口，该网络接口131通常用于在所述电子装置13与其他电子设备之间建立通信连接。
在一实施方式中，所述处理器132在一些实施方式中可以是中央处理器(Central Processing Unit，CPU)、控制器、微控制器、微处理器、或其他数据处理芯片。该处理器132通常用于控制所述电子装置13的总体操作。本实施方式中，所述处理器132用于运行所述存储器130中存储的程序代码或者处理数据，例如运行所述的管控药物的特征获取系统200等。
至此，己经详细介绍了本申请相关设备的硬件结构和功能。下面，将基于上述介绍提出本申请的各个实施方式。
首先，本申请提出一种管控药物的特征获取系统200。
参阅图3所示，是本申请管控药物的特征获取系统200第一实施方式的程序模块图。
在一实施方式中，所述管控药物的特征获取系统200包括一系列的存储于存储器130上的计算机程序指令，当该计算机程序指令被处理器132执行时，可以实现本申请各实施方式的特征获取操作。在一些实施方式中，基于该计算机程序指令各部分所实现的特定的操作，管控药物的特征获取系统200可以被划分为一个或多个模块。例如，在图3中，所述管控药物的特征获取系统200可以被分割成获取模块201、处理模块202、建立模块203、调用模块204、存储模块205及显示模块206。其中：
所述获取模块201，用于获取保险机构数据库10及医疗机构数据库11的原始病例数据。
具体地，接收保险机构数据库10及各级医疗机构数据库11的所有病例数据，保险机构数据库10可为本市保险公司的数据库，该数据库中的病例数据主要包括保单、收据，医疗机构数据库11可为本市的医院、医疗中心的数据库/服务器，根据各地医保政策的不同，可以根据需要获取任意包括省、市、区及其他区域的医院、医疗中心等的原始病例数据。
所述处理模块202，用于对所述原始病例数据进行预处理，得到预处理后的数据集。
具体地，预处理方式包括：
1，将所述原始病例数据按照时间顺序排序；也可以对数据按时间分段，比如[Ti，Ti+1]，i＝0,1…N-1；
2，对排序后的数据进行规范化处理，包括：统一量化方式、统一计量单位、统一表现形式。统一量化方式主要针对枚举型数据，例如患者对特定病毒、疾病的反应{免疫，高抗，抗，感}四种情况；统一计量单位主要针对数值型数据，例如管控药物使用的剂量使用副、包、克、毫升等作为单位，对 不同的性质的数据归类采用统一计量单位，例如对毫升、升的计量统一采用毫升作为计量单位，对毫克、克、千克的计量统一采用毫克作为计量单位；统一表现形式主要针对多表现形式的数据，例如日期型数据，既可以表示为YYYY-MM-DD，也可以表示为MM-DD-YYYY等其它形式，在此统一以YYYY-MM-DD的形式统一标识。
3，将规范化处理后的数据进行去噪，建立清洗模型对数据进行筛选，将明显不符合判断标准的数据删除：
所述清洗模型包括多条判断标准，例如，对原始数据中的乱码数据进行清除，对明显不符合常识或者对比预设条件明显错误的数据进行清除，将对管控药物使用影响无关的数据进行清除，对重复数据进行清除，对数据缺失严重的患者的数据进行清除；
根据每个标准建立一个筛选器，用多个所述筛选器同步/异步对数据进行筛选。
所述建立模块203，用于根据管控药物的前向性特征、本性特征及后向性特征建立分类模型。
具体地，以时间维度为依据建立分类模型，选取特定时间点参考点，在所述参考点获取的数据作为本性特征，将参考点之前的数据作为前向性特征，将参考点之后获得的数据作为后向性特征；
具体地，选取不同时段管控药物通用性特征的数据作为训练样本对分类模型进行训练；
具体地，分类模型可为决策树分类器,决策树模型提供一个属性集合，决策树通过在属性集的基础上作出一系列的决策，将数据分类；选择树分类器，选择树分类器使用与决策树分类器相似的技术对数据进行分类。与决策树不同的是，选择树中包含特殊的选择节点，选择节点有多个分支。另外，还可使用人工神经网络、案例推论、最近邻居法、支持向量机及随机森林等分类器。
所述调用模块204用于通过所述分类模型提取管控药物的数据。
具体地，通过所述分类模型分类后的数据都具有管控药物的标签，比如，“胰岛素”，“葡萄糖”等管控药物都被提取出来，此时，所有的数据都为与管控药物相关的数据；
具体地，可以针对提取出来的数据建立管控药物数据库，以方便后续使用及查询。
在一实施方式中，所述建立模块203还用于根据特定管控药物的特有特征建立特定管控药物特征提取模型。
具体地，不同地管控药物具有不同的特性，所述特性包括但不限于应用场景、根据医保政策规定的报销条件及报销额度、药物的使用状况、不同地方的特异性等，根据不同管控药物的特有特征建立特定管控药物特征提取模型就可以提取出所述特定管控药物的特征。例如，糖尿病已经由原来的老年病发展为儿童、年轻人、老年人都会得的病。虽然是同一种病，但是不同年 龄段的管理确是不同的，比如，用于疾病管理的时间不同、治疗达标的时间不同、身体状况不同、血糖管理标准不同，因为这些不同点，不同的年龄在胰岛素使用上具有不同地特点，因此可以根据年龄作为特有特征建立胰岛素的特征提取依据，以年龄为基础，根据不同年龄胰岛素不同的使用情况建立表达式，以该表达式为基础建立特征提取模型，根据该特征提取模型就可以找到管控药物数据库中所有与胰岛素有关的数据。
具体地，所述的特征提取模型可为分类模型，比如可为二分类模型，定义所述二分类模型包括正类和负类两个类别，分别代表特定管控药物及其他药物，比如正类为胰岛素，负类为其它管控药物。
使用二分类模型对管控药物数据库中的测试样本进行预测(分类)，有时会出现分类错误的情况，比如当两种管控药物的特有特征出现交集，此时对于该特征提取模型需要进行修正。例如，可以增加分类打分模型。
具体地，使用二分类模型对管控药物数据库中的测试样本进行分类，将正类中的每一个对象称为正实例，将负类中的每一个对象称为负实例，此时会有四种情况出现，具体如下：如果一个实例是正类并被预测为正类，称之为真正类(True positive，TP)，如果实例是负类被预测为正类，称之为假正类(False postiVe，FP)。相应的，相应地，如果实例是负类被预测成负类，称之为真负类(True negative，TN)，正实例被预测成负类则为假负类(false negative，FN)。
在所述分类打分模型中，以TP、FN、FP、TN分别表示对应分类的数目，即：TP:正实例预测为正类数目；FN:正实例预测为负类数目；FP:负实例预测为正类的数目；TN:负实例预测为负类的数目。
另外，所述分类打分模型还可以计算下列三种参数：敏感性(sensitivity):正类中正确预测为正类的实例比例，即TP/(TP+FN)。特异性(specificity):负类中被正确预测为负类的实例比例，即TN/(TN+FP)。阳性预测值(Positive Predictive Value，PPV):预测为正类的实例中，正实例占的比例，即TP/(TP+FP)。
根据以上分类打分模型可以对分类的结果进行评价，根据评价结果可以对分类情况进行了解，并对特征提取模型进行优化。
在一实施方式中，所述调用模块204还用于通过所述特征提取模型返回特定管控药物的特征表。
具体地，影响使用特定管控药物的特征包括多种，根据特征特其模型返回的数据可以建立特定管控药物的特征表，在所述特征表中包括对该特定管控药物使用的多种影响因素。例如，以胰岛素为例，特征表中的数据可以包括：病人相关的预存参数的信息，该参数包括病人状况参数，例如药物过敏史或敏感历、当前在病人的组织中的其它目前服用的药物、年龄、体重、身高、肾功能或肝功能检测数据；病人身体当前状况相关的参数(即，病人状况参数)，例如血压、心率、心律、温度、血氧、呼吸率或换气频率；药物参数，管控药物的使用量、使用频率、花费费用、药物使用时间、当前药物、药物类别、药物过敏历和敏感度；合并症案例数据，包括急性合并症及慢性合并 症，例如，急性合并症包括低血糖和高血糖，慢性合并症包括眼睛病变、肾脏病变、神经病变、心脏血管病变及足部病变等。
所述存储模块205用于存储获取模块201、处理模块202、建立模块203、调用模块204获取的数据，例如，存储模块205可用于存储特定管控药物的特征表。所述存储模块205包括可读存储介质，包括闪存、硬盘、多媒体卡、卡型存储器(例如，SD或DX存储器等)、随机访问存储器(RAM)、静态随机访问存储器(SRAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、可编程只读存储器(PROM)、磁性存储器、磁盘、光盘等。
所述显示模块206用于显示管控药物的特征获取系统200工作过程中的中间结果及最终结果，所述显示模块206包括LCD、LED等显示设备，所述显示模块206可用于显示所述特定管控药物的特征表。
此外，本申请还提出一种管控药物的特征获取方法。
参阅图4所示，是本申请管控药物的特征获取方法第一实施方式的流程示意图。在本实施方式中，根据不同的需求，图4所示的流程图中的步骤的执行顺序可以改变，某些步骤可以省略。
步骤S110，获取保险机构数据库10及医疗机构数据库11的原始病例数据。
具体地，接收保险机构数据库10及各级医疗机构数据库11的所有病例数据，保险机构数据库10可为本市保险公司的数据库，该数据库中的病例数据主要包括保单、收据，医疗机构数据库11可为本市的医院、医疗中心，根据各地医保政策的不同，可以根据需要获取任意包括省、市、区及其他区域的医院、医疗中心等的原始病例数据。
步骤S120，对所述原始病例数据进行预处理，得到预处理后的数据集。
具体地，所述预处理方式包括将所述原始病例数据按照时间顺序排序、规范化、去噪等。
步骤S130，根据管控药物的前向性特征、本性特征及后向性特征建立分类模型。
具体地，以时间维度为依据建立分类模型，选取特定时间点参考点，在所述参考点获取的数据作为本性特征，将参考点之前的数据作为前向性特征，将参考点之后获得的数据作为后向性特征；
具体地，选取不同时段管控药物通用性特征的数据作为训练样本对分类模型进行训练；
具体地，分类模型可为决策树分类器,决策树模型提供一个属性集合，决策树通过在属性集的基础上作出一系列的决策，将数据分类；选择树分类器，选择树分类器使用与决策树分类器相似的技术对数据进行分类。与决策树不同的是，选择树中包含特殊的选择节点，选择节点有多个分支。另外，还可使用人工神经网络、案例推论、最近邻居法、支持向量机及随机森林等分类器。
步骤S140，通过所述分类模型提取所述管控药物的数据。
具体地，通过所述分类模型分类后的数据都具有管控药物的标签，比如，“胰岛素”，“葡萄糖”等管控药物都被提取出来，此时，所有的数据都为与管控药物相关的数据；
步骤S150，根据特定管控药物的特有特征建立特定管控药物特征提取模型。
具体地，不同地管控药物具有不同的特性，所述特性包括但不限于应用场景、根据医保政策规定的报销条件及报销额度、药物的使用状况、不同地方的特异性等，根据不同管控药物的特有特征建立特定管控药物特征提取模型就可以提取出所述特定管控药物的特征。例如，糖尿病已经由原来的老年病发展为儿童、年轻人、老年人都会得的病。虽然是同一种病，但是不同年龄段的管理确是不同的，比如，用于疾病管理的时间不同、治疗达标的时间不同、身体状况不同、血糖管理标准不同，因为这些不同点，不同的年龄在胰岛素使用上具有不同地特点，因此可以根据年龄作为特有特征建立胰岛素的特征提取依据，以年龄为基础，根据不同年龄胰岛素不同的使用情况建立表达式，以该表达式为基础建立特征提取模型，根据该特征提取模型就可以找到管控药物数据库中所有与胰岛素有关的数据。
具体地，所述的特征提取模型可为分类模型，比如可为二分类模型，定义所述二分类模型包括正类和负类两个类别，分别代表特定管控药物及其他药物，比如正类为胰岛素，负类为其它管控药物。
步骤S160，通过所述特征提取模型返回特定管控药物的特征表。
具体地，影响使用特定管控药物的特征包括多种，根据特征特其模型返回的数据可以建立特定管控药物的特征表，在所述特征表中包括对该特定管控药物使用的多种影响因素。例如，以胰岛素为例，特征表中的数据可以包括：病人相关的预存参数的信息，该参数包括病人状况参数，例如药物过敏史或敏感历、当前在病人的组织中的其它目前服用的药物、年龄、体重、身高、肾功能或肝功能检测数据；病人身体当前状况相关的参数(即，病人状况参数)，例如血压、心率、心律、温度、血氧、呼吸率或换气频率；药物参数，管控药物的使用量、使用频率、花费费用、药物使用时间、当前药物、药物类别、药物过敏历和敏感度；合并症案例数据，包括急性合并症及慢性合并症，例如，急性合并症包括低血糖和高血糖，慢性合并症包括眼睛病变、肾脏病变、神经病变、心脏血管病变及足部病变等。
如图5所示，是本申请管控药物的特征获取方法的第二实施方式的流程示意图。本实施方式中，所述管控药物的特征获取方法第一实施方式中的步骤S120中，所述预处理步骤还包括以下步骤：
步骤S210，将所述原始病例数据按照时间顺序排序。
具体地，将所述原始病例数据按照时间顺序排序，也可以对数据按时间分段，比如[Ti，Ti+1]，i＝0,1…N-1。
步骤S220，对排序后的数据进行规范化处理。
具体地，所述规范化处理包括：统一量化方式、统一计量单位、统一表现形式。统一量化方式主要针对枚举型数据，例如患者对特定病毒、疾病的反应{免疫，高抗，抗，感}四种情况；统一计量单位主要针对数值型数据，例如管控药物使用的剂量使用副、包、克、毫升等作为单位，对不同的性质的数据归类采用统一计量单位，例如对毫升、升的计量统一采用毫升作为计量单位，对毫克、克、千克的计量统一采用毫克作为计量单位；统一表现形式主要针对多表现形式的数据，例如日期型数据，既可以表示为YYYY-MM-DD，也可以表示为MM-DD-YYYY等其它形式，在此统一以YYYY-MM-DD的形式统一标识。
步骤S230，将规范化处理后的数据进行去噪。
具体地，所述去噪包括建立清洗模型对数据进行筛选，将明显不符合判断标准的数据删除：
所述清洗模型包括多条判断标准，例如，对原始数据中的乱码数据进行清除，对明显不符合常识或者对比预设条件明显错误的数据进行清除，将对管控药物使用影响无关的数据进行清除，对重复数据进行清除，对数据缺失严重的患者的数据进行清除。
具体地，根据每个标准建立一个筛选器，用多个所述筛选器同步/异步对数据进行筛选。
管控药物的特征获取方法管控药物的特征获取方法管控药物的特征获取方法管控药物的特征获取方法
如图6所示，是本申请管控药物的特征获取方法的第三实施方式的流程示意图。本实施方式中，所述管控药物的特征获取方法第一实施方式的步骤S150，具体包括以下步骤：
步骤S510，根据特定管控药物的特有特征建立二分类模型。
具体地，定义所述二分类模型包括正类和负类两个类别，分别代表特定管控药物及其他药物，比如正类为胰岛素，负类为其它管控药物。
步骤S520，使用二分类模型对管控药物数据库中的测试样本进行分类。
具体地，将正类中的每一个对象称为正实例，将负类中的每一个对象称为负实例，此时会有四种情况出现，具体如下：如果一个实例是正类并被预测为正类，称之为真正类(True positive，TP)，如果实例是负类被预测为正类，称之为假正类(False postiVe，FP)。相应的，相应地，如果实例是负类被预测成负类，称之为真负类(True negative，TN)，正实例被预测成负类则为假负类(false negative，FN)。
步骤S530，使用分类打分模型对分类的结果进行评价，以对所述二分类模型进行优化。
具体地，使用二分类模型对管控药物数据库中的测试样本进行预测(分类)，有时会出现分类错误的情况，比如当两种管控药物的特有特征出现交集，此时对于该特征提取模型需要进行修正。例如，可以增加分类打分模型。
在所述分类打分模型中，以TP、FN、FP、TN分别表示对应分类的数目， 即：TP:正实例预测为正类数目；FN:正实例预测为负类数目；FP:负实例预测为正类的数目；TN:负实例预测为负类的数目。
另外，所述分类打分模型还可以计算下列三种参数：敏感性(sensitivity):正类中正确预测为正类的实例比例，即TP/(TP+FN)。特异性(specificity):负类中被正确预测为负类的实例比例，即TN/(TN+FP)。阳性预测值(Positive Predictive Value，PPV):预测为正类的实例中，正实例占的比例，即TP/(TP+FP)。
根据以上分类打分模型可以对分类的结果进行评价，根据评价结果可以对分类情况进行了解，并对特征提取模型(二分类模型)进行优化。
如图7所示，是本申请管控药物的特征获取方法的第四实施方式的流程示意图。本实施方式中，所述管控药物的特征获取方法在第一实施方式的步骤S140之后还包括以下步骤：
步骤S610，针对提取出来的数据建立管控药物数据库。
具体地，可以针对提取出来的数据建立管控药物数据库，以方便后续使用及查询。所述数据库，各专业公司的实现方式不同，主要的数据库类型为Oracle，也会存在PostgreSQL、MySQL等类型的各种数据库。
参阅图8所示，是本申请管控药物的特征获取方法第五实施方式的流程示意图。在本实施方式中，所述管控药物的特征获取方法的步骤S710-S760与第一实施例的步骤S110-S160相类似，区别在于该方法还包括步骤S770-S780。根据不同的需求，图10所示的流程图中的步骤的执行顺序可以改变，某些步骤可以省略。
步骤S710，获取保险机构数据库10及医疗机构数据库11的原始病例数据。
步骤S720，对所述原始病例数据进行预处理，得到预处理后的数据集。
步骤S730，根据管控药物的前向性特征、本性特征及后向性特征建立分类模型。
步骤S740，通过所述分类模型提取所述管控药物的数据。
步骤S750，根据特定管控药物的特有特征建立特定管控药物特征提取模型。
步骤S760，通过所述特征提取模型返回特定管控药物的特征表。
步骤S770，存储所述特征表。
具体地，存储模块205可用于存储特定管控药物的特征表。所述存储模块205包括可读存储介质，包括闪存、硬盘、多媒体卡、卡型存储器(例如，SD或DX存储器等)、随机访问存储器(RAM)、静态随机访问存储器(SRAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、可编程只读存储器(PROM)、磁性存储器、磁盘、光盘等。
步骤S780，向用户显示所述特征表。
具体地，通过显示设备将管控药物的特征获取系统200工作过程中的中间结果及最终结果显示给用户。
相较于现有技术，本申请所提出的电子装置、管控药物的特征获取方法及计算机可读存储介质，首先，获取保险机构数据库及医疗机构数据库的原始病例数据；然后，对所述原始病例数据进行预处理，得到预处理后的数据集；其次，根据管控药物的前向性特征、本性特征及后向性特征建立分类模型；之后，通过所述分类模型提取所述管控药物的数据；然后，根据特定管控药物的特有特征建立特定管控药物特征提取模型；最后，通过所述特征提取模型返回特定管控药物的特征表。通过本申请所提出的电子装置、管控药物的特征获取方法及计算机可读存储介质，对大数据进行分析能够快速获取各种管控药物的特征集合，极大地提高不同地区、不同管控药物的特征提取速度、准确性，相对于现有技术更见便捷、准确、迅速。
上述本申请实施方式序号仅仅为了描述，不代表实施方式的优劣。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施方式方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，空调器，或者网络设备等)执行本申请各个实施方式所述的方法。
以上仅为本申请的优选实施方式，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种管控药物的特征获取方法，应用于电子装置，其特征在于，所述方法包括：
    获取保险机构数据库及医疗机构数据库的原始病例数据；
    对所述原始病例数据进行预处理，得到预处理后的数据集；
    根据管控药物的前向性特征、本性特征及后向性特征建立分类模型；
    通过所述分类模型提取所述管控药物的数据；
    根据特定管控药物的特有特征建立特定管控药物特征提取模型；及
    通过所述特征提取模型返回所述特定管控药物的特征表。
  2. 如权利要求1所述的管控药物的特征获取方法，其特征在于，所述预处理包括以下步骤：
    将所述原始病例数据按照时间顺序排序；
    对排序后的数据进行规范化处理；及
    将规范化处理后的数据进行去噪。
  3. 如权利要求2所述的管控药物的特征获取方法，其特征在于，所述规范化处理包括如下步骤：
    针对枚举型数据统一量化方式；
    针对数值型数据统一计量单位；及
    针对多表现形式的数据统一表现形式。
  4. 如权利要求2所述的管控药物的特征获取方法，其特征在于，所述去噪步骤包括：
    建立清洗模型；及
    根据所述清洗模型对所述规范化处理后的数据进行筛选。
  5. 如权利要求1-4任一项所述的管控药物的特征获取方法，其特征在于，建立所述分类模型具体还包括步骤：
    以时间维度为依据建立分类模型；及
    选取不同时段管控药物通用性特征的数据作为训练样本对分类模型进行训练；
    其中，所述分类模型为决策树分类器。
  6. 如权利要求1-4任一项所述的管控药物的特征获取方法，其特征在于，通过所述分类模型提取所述管控药物的数据之后还包括如下步骤：
    针对提取出来的数据建立管控药物数据库。
  7. 如权利要求1-4任一项所述的管控药物的特征获取方法，其特征在于，所述特征提取模型为二分类模型，所述二分类模型包括正类和负类两个类别，所述正类和负类分别代表特定管控药物及其他药物。
  8. 如权利要求1-4任一项所述的管控药物的特征获取方法，其特征在于，所述特征获取的方法还包括以下步骤：
    存储各步骤的中间数据及所述特征表；及
    通过显示设备向用户显示所述特征表。
  9. 一种电子装置，其特征在于，所述电子装置包括存储器、处理器，所述存储器上存储有可在所述处理器上运行的管控药物的特征获取系统，所述管控药物的特征获取系统被所述处理器执行时实现如下步骤：
    获取保险机构数据库及医疗机构数据库的原始病例数据；
    对所述原始病例数据进行预处理，得到预处理后的数据集；
    根据管控药物的前向性特征、本性特征及后向性特征建立分类模型；
    通过所述分类模型提取所述管控药物的数据；
    根据特定管控药物的特有特征建立特定管控药物特征提取模型；及
    通过所述特征提取模型返回所述特定管控药物的特征表。
  10. 如权利要求9所述的电子装置，其特征在于，所述预处理包括以下步骤：
    将所述原始病例数据按照时间顺序排序；
    对排序后的数据进行规范化处理；及
    将规范化处理后的数据进行去噪。
  11. 如权利要求10所述的电子装置，其特征在于，所述规范化处理包括如下步骤：
    针对枚举型数据统一量化方式；
    针对数值型数据统一计量单位；及
    针对多表现形式的数据统一表现形式。
  12. 如权利要求10所述的电子装置，其特征在于，所述去噪步骤包括：
    建立清洗模型；及
    根据所述清洗模型对所述规范化处理后的数据进行筛选。
  13. 如权利要求9-12任一项所述的电子装置，其特征在于，建立所述分类模型具体还包括步骤：
    以时间维度为依据建立分类模型；及
    选取不同时段管控药物通用性特征的数据作为训练样本对分类模型进行训练；
    其中，所述分类模型为决策树分类器。
  14. 如权利要求9-12任一项所述的电子装置，其特征在于，通过所述分类模型提取所述管控药物的数据之后还包括如下步骤：
    针对提取出来的数据建立管控药物数据库。
  15. 如权利要求9-12任一项所述的电子装置，其特征在于，所述特征提取模型为二分类模型，所述二分类模型包括正类和负类两个类别，所述正类和负类分别代表特定管控药物及其他药物。
  16. 如权利要求9-12任一项所述的电子装置，其特征在于，所述特征获取的方法还包括以下步骤：
    存储各步骤的中间数据及所述特征表；及
    通过显示设备向用户显示所述特征表。
  17. 一种计算机可读存储介质，所述计算机可读存储介质存储有管控药物的特征获取系统，所述管控药物的特征获取系统可被至少一个处理器执行，以使所述至少一个处理器执行如下步骤：
    获取保险机构数据库及医疗机构数据库的原始病例数据；
    对所述原始病例数据进行预处理，得到预处理后的数据集；
    根据管控药物的前向性特征、本性特征及后向性特征建立分类模型；
    通过所述分类模型提取所述管控药物的数据；
    根据特定管控药物的特有特征建立特定管控药物特征提取模型；及
    通过所述特征提取模型返回所述特定管控药物的特征表。
  18. 如权利要求17所述的计算机可读存储介质，其特征在于，所述预处理包括以下步骤：
    将所述原始病例数据按照时间顺序排序；
    对排序后的数据进行规范化处理；及
    将规范化处理后的数据进行去噪。
  19. 如权利要求18所述的计算机可读存储介质，其特征在于，所述规范化处理包括如下步骤：
    针对枚举型数据统一量化方式；
    针对数值型数据统一计量单位；及
    针对多表现形式的数据统一表现形式。
  20. 如权利要求18所述的计算机可读存储介质，其特征在于，所述去噪步骤包括：
    建立清洗模型；及
    根据所述清洗模型对所述规范化处理后的数据进行筛选。
PCT/CN2018/090906 2017-11-10 2018-06-12 管控药物的特征获取方法、电子装置及计算机可读存储介质 Ceased WO2019091111A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711107651.0 2017-11-10
CN201711107651.0A CN108022653B (zh) 2017-11-10 2017-11-10 管控药物的特征获取方法、电子装置及计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2019091111A1 true WO2019091111A1 (zh) 2019-05-16

Family

ID=62080543

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/090906 Ceased WO2019091111A1 (zh) 2017-11-10 2018-06-12 管控药物的特征获取方法、电子装置及计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN108022653B (zh)
WO (1) WO2019091111A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113488187A (zh) * 2021-08-03 2021-10-08 南通市第二人民医院 一种麻醉意外案例收集分析方法及系统

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108022653B (zh) * 2017-11-10 2022-03-18 平安科技（深圳）有限公司 管控药物的特征获取方法、电子装置及计算机可读存储介质
CN111933244A (zh) * 2020-08-17 2020-11-13 医渡云（北京）技术有限公司 药品数据编码方法、装置、计算机可读介质及电子设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090043611A1 (en) * 2007-08-07 2009-02-12 Walgreen Co. Interface system for displaying comprehensive patient medication record
SE1200508A1 (sv) * 2012-08-23 2014-02-24 Careligo Ab Patientövervakningsanordning
CN104487972A (zh) * 2014-09-29 2015-04-01 曹庆恒 一种多维度用药信息处理方法、系统和设备
CN108022653A (zh) * 2017-11-10 2018-05-11 平安科技（深圳）有限公司 管控药物的特征获取方法、电子装置及计算机可读存储介质

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5963453A (en) * 1996-11-25 1999-10-05 Medication Management, Inc. System and method for processing prescription medications
TW200629089A (en) * 2005-02-03 2006-08-16 jia-chang Liu Medicine composition analysis and diagnosis system
US20120137239A1 (en) * 2010-11-30 2012-05-31 Mckesson Corporation Systems and methods for managing drug therapeutic class information
CN106326657B (zh) * 2016-08-24 2018-10-19 北京叮叮关爱科技有限公司 服药计划的推荐方法及系统
CN106339599A (zh) * 2016-08-31 2017-01-18 杭州逸曜信息技术有限公司 用药合理性检测关联信息处理方法
CN106650214A (zh) * 2016-10-12 2017-05-10 四川农业大学 一种用于医护管理的取药、服药提示系统及方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090043611A1 (en) * 2007-08-07 2009-02-12 Walgreen Co. Interface system for displaying comprehensive patient medication record
SE1200508A1 (sv) * 2012-08-23 2014-02-24 Careligo Ab Patientövervakningsanordning
CN104487972A (zh) * 2014-09-29 2015-04-01 曹庆恒 一种多维度用药信息处理方法、系统和设备
CN108022653A (zh) * 2017-11-10 2018-05-11 平安科技（深圳）有限公司 管控药物的特征获取方法、电子装置及计算机可读存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113488187A (zh) * 2021-08-03 2021-10-08 南通市第二人民医院 一种麻醉意外案例收集分析方法及系统
CN113488187B (zh) * 2021-08-03 2024-02-20 南通市第二人民医院 一种麻醉意外案例收集分析方法及系统

Also Published As

Publication number Publication date
CN108022653A (zh) 2018-05-11
CN108022653B (zh) 2022-03-18

Similar Documents

Publication Publication Date Title
US11755566B2 (en) Managing data objects for graph-based data structures
US10430716B2 (en) Data driven featurization and modeling
WO2019091111A1 (zh) 管控药物的特征获取方法、电子装置及计算机可读存储介质
Nabi et al. Machine Learning Approach: Detecting Polycystic Ovary Syndrome & It's Impact on Bangladeshi Women.
Salehinejad et al. Novel machine learning model to improve performance of an early warning system in hospitalized patients: a retrospective multisite cross-validation study
Lee et al. Comparison of the new-poisoning mortality score and the modified early warning score for predicting in-hospital mortality in patients with acute poisoning
CN110459330A (zh) 药物适应症预测方法及装置
WO2019085464A1 (zh) 违规单据评分方法、违规单据评分装置及计算机可读存储介质
Yu et al. Healthcare-Event driven semantic knowledge extraction with hybrid data repository
CN118016307A (zh) 一种疾病预警方法、系统和相关装置
CN111710431A (zh) 一种识别同义诊断名称的方法、装置、设备及存储介质
CA3098994C (en) Managing data objects for graph-based data structures

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18875771

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 22/09/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 18875771

Country of ref document: EP

Kind code of ref document: A1