Treatment cost prediction method, system, device and storage medium based on machine learning

Abstract

The present invention relates to a method, system, device and storage medium for predicting treatment costs based on machine learning. It includes: obtaining basic information data and diagnostic information data of a patient; inputting the basic information data and diagnostic information data into a pre-trained disease diagnosis model to obtain the type of disease of the patient; according to the type of disease, retrieving a data subset of different treatment plans under the type of disease from a hospital medical record database; comparing the basic information data and diagnostic data of the patient with the basic information data and diagnostic information data of the medical record in the data subset, matching the closest medical record of the patient in the data subset of different treatment plans, and obtaining the treatment plan and treatment cost of the medical record; outputting different treatment plans of the patient and the treatment cost of each treatment plan according to the treatment plan and treatment cost of the case. This application has important application value in improving doctor-patient relationship.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical technology, and more specifically, to a method, system, device and storage medium for predicting treatment costs based on machine learning. Background Art

[0002] Currently, when patients and their families are faced with treatment plan decisions in clinical scenarios, they can only passively rely on doctors' advice due to the lack of relevant professional knowledge. It is difficult for them to fully understand the advantages and disadvantages of various clinical treatment pathways in a short period of time and make decisions that are more in line with their own interests. In addition, the existing clinical treatment plans have high requirements for doctors' clinical experience, which poses a great challenge to doctors with insufficient clinical experience, and clinicians may find it difficult to take care of patients' own needs. Summary of the invention

[0003] To solve the above problems, the present application establishes a treatment cost prediction method, system, device and storage medium based on machine learning.

[0004] The purpose of this application is to provide a method for predicting treatment costs based on machine learning, and the specific method steps include:

[0005] Obtain the patient's basic information and diagnostic information;

[0006] Inputting the basic information data and the diagnostic information data into a pre-trained disease diagnosis model to obtain the patient's disease type;

[0007] According to the disease type, retrieve a data subset of different treatment plans under the disease type from a hospital medical record database, wherein the data subset of different treatment plans under the disease type from the hospital medical record database is obtained by performing cluster analysis on the hospital medical record database;

[0008] Compare the basic information data and diagnostic data of the patient with the basic information data and diagnostic information data of the medical records in the data subset, match the closest medical records of the patient in the data subsets with different treatment plans, and obtain the treatment plan and treatment cost of the medical record;

[0009] According to the treatment plan and treatment cost of the case, the different treatment plans for the patient and the treatment cost of each treatment plan are output.

[0010] Furthermore, the training process of the pre-trained disease diagnosis model is as follows: obtaining the basic information data and diagnosis information data of the patient as a training set, training the machine learning model, comparing the predicted disease type with the actual disease type, generating a loss value, back-propagating, optimizing the disease diagnosis model, and obtaining the pre-trained disease diagnosis model;

[0011] Optionally, the basic information data includes information data such as age, gender, admission time, discharge time, etc., and the diagnostic information data includes diagnostic information data such as admission diagnosis, examination and testing data, nursing diagnosis, will data, discharge diagnosis, etc.

[0012] Furthermore, the algorithm used by the pre-trained disease diagnosis model is selected from a variety of machine learning algorithms, and the machine learning algorithm is selected from one or more of the following machine learning algorithms: support vector machine, random forest, XGBoost, LightGBM.

[0013] Furthermore, the process of constructing the data subsets of different treatment plans under the disease category in the hospital medical record database includes: obtaining all medical record sets under the disease category in the hospital medical record database, performing cluster analysis based on differences in treatment plans, and obtaining data subsets of different treatment plans under the disease category;

[0014] Preferably, the cluster analysis may adopt FCM fuzzy clustering or K-Means clustering; optionally, the hospital medical record database is a dynamically updated database.

[0015] Furthermore, the method steps also include correcting the treatment plan and treatment costs of the obtained medical records to obtain corrected treatment plans and treatment costs of the medical records, and outputting different treatment plans for patients and the treatment costs of each treatment plan based on the corrected treatment plans and treatment costs of the cases. The correction of the treatment plans and treatment costs of the medical records includes correction of the treatment plans and / or correction of the treatment costs of the medical records. The correction of the treatment costs is to restore the medical insurance exemption costs in the medical records to full self-paid costs to obtain the treatment costs of the corrected medical records; the correction of the treatment plan is to match and correct the drugs and drug prices in the medical records with the drugs and drug prices in the hospital pharmacy to obtain the corrected treatment plan of the medical record.

[0016] Furthermore, the correction of the treatment plan is to match and correct the drugs and drug prices in the medical record with the drugs and drug prices in the hospital drug store to obtain the treatment plan of the corrected medical record. When the drugs in the treatment plan and the drugs in the hospital drug store match each other, the drug prices are updated to obtain the treatment plan and treatment costs of the medical record after the price correction; when the drugs in the treatment plan and the drugs in the hospital drug store do not match each other, the drugs are replaced with drugs of the same type to obtain the treatment plan and treatment costs of the medical record after the drugs and prices are corrected.

[0017] Optionally, when replacing drugs of the same type, drugs with the same generic name should be considered first, and drugs with the same function should be considered secondarily.

[0018] Furthermore, the method step also includes, before outputting different treatment plans for the patient and the treatment costs of each treatment plan according to the treatment plan and treatment costs of the case, selecting the settlement type of the patient, and outputting different treatment plans and the treatment costs of each treatment plan under the settlement type of the patient according to the settlement type of the patient, the treatment plan and treatment costs of the case;

[0019] Optionally, the settlement type includes ordinary patients, medical insurance patients or public medical care patients.

[0020] The purpose of this application is to provide a treatment cost prediction system based on machine learning, the prediction system comprising:

[0021] An acquisition unit, used to acquire basic information data and diagnostic information data of the patient;

[0022] A diagnosis unit, used to input the basic information data and the diagnosis information data into a pre-trained disease diagnosis model to obtain the type of disease of the patient;

[0023] A retrieving unit, configured to retrieve, according to the disease type, a data subset of different treatment plans under the disease type from a hospital medical record database, wherein the data subset of different treatment plans under the disease type from the hospital medical record database is obtained by performing cluster analysis on the hospital medical record database;

[0024] A matching unit, used to compare the basic information data and diagnostic data of the patient with the basic information data and diagnostic information data of the medical records in the data subset, match the closest medical records of the patient in the data subsets with different treatment plans, and obtain the treatment plan and treatment cost of the medical records;

[0025] The output unit is used to output different treatment plans for the patient and the treatment cost of each treatment plan according to the treatment plan and treatment cost of the case.

[0026] The purpose of this application is to provide a device for predicting treatment costs based on machine learning, the prediction device comprising: a memory and a processor;

[0027] The memory is used to store program instructions;

[0028] The processor is used to call program instructions, and when the program instructions are executed, the above-mentioned treatment cost prediction method based on machine learning is implemented.

[0029] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned method for predicting treatment costs based on machine learning.

[0030] Advantages of this application:

[0031] 1. This application can quickly diagnose the patient's disease based on the patient's basic information and diagnostic information, and quickly match the treatment plan and treatment costs of the relevant medical records in the hospital database based on the diagnosis results, so as to provide more clinical decisions for patients and doctors in the early stage;

[0032] 2. This application predicts the patient's treatment plan and treatment costs at an early stage, allowing patients to participate in clinical decision-making based on their own conditions. The transparency of diagnosis and treatment will greatly avoid possible conflicts between doctors and patients;

[0033] 3. This application takes into account various situations of hospital treatment cost forecasting, including relevant deviation factors of existing medical record costs, actual situation of drug warehouse, drug updates, price fluctuations caused by volume purchases, etc., and provides more accurate treatment cost forecasts in combination with patient settlement types;

[0034] 4. The disease diagnosis model provided by this application helps doctors to determine the type of disease of patients more quickly and accurately. It makes full use of the valuable intangible asset of the hospital medical record database to provide doctors with different treatment plans for similar diseases in the past, thereby reducing reliance on doctors' personal experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0036] Figure 1 is a schematic flow chart of a method for predicting treatment costs based on machine learning provided by an embodiment of the present invention;

[0037] Figure 2 is a schematic diagram of a treatment cost prediction system based on machine learning provided by an embodiment of the present invention;

[0038] Figure 3 It is a schematic diagram of a treatment cost prediction device based on machine learning provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0040] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0041] The following will be combined with the drawings in 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0042] Figure 1 The present invention provides a method for predicting treatment costs based on machine learning. Specifically, the method includes the following steps:

[0043] 101: Obtain the patient's basic information and diagnostic information;

[0044] In one embodiment, the basic information data includes information data such as age, gender, admission time, discharge time, etc., and the diagnostic information data includes diagnostic information data such as admission diagnosis, examination and testing data, nursing diagnosis, will data, discharge diagnosis, etc.

[0045] 102: Input the basic information data and the diagnostic information data into a pre-trained disease diagnosis model to obtain the disease type of the patient;

[0046] In one embodiment, the training process of the pre-trained disease diagnosis model is as follows: obtaining the patient's basic information data and diagnosis information data as a training set, training the machine learning model, comparing the predicted disease type with the actual disease type, generating a loss value, back-propagating, optimizing the disease diagnosis model, and obtaining the pre-trained disease diagnosis model.

[0047] In one embodiment, the algorithm used by the pre-trained disease diagnosis model is selected from a variety of machine learning algorithms, and the machine learning algorithm is selected from one or more of the following machine learning algorithms: support vector machine, random forest, XGBoost, LightGBM.

[0048] 103: Retrieving data subsets of different treatment plans under the disease type from a hospital medical record database according to the disease type, wherein the data subsets of different treatment plans under the disease type from the hospital medical record database are obtained by performing cluster analysis on the hospital medical record database;

[0049] In one embodiment, the process of constructing data subsets of different treatment options under the disease category in the hospital medical record database includes: obtaining all medical record sets under the disease category in the hospital medical record database, performing cluster analysis based on differences in treatment options, and obtaining data subsets of different treatment options under the disease category. Preferably, the cluster analysis can use FCM fuzzy clustering or K-Means clustering; optionally, the hospital medical record database is a dynamically updated database.

[0050] 104: Compare the basic information data and diagnostic data of the patient with the basic information data and diagnostic information data of the medical records in the data subset, match the closest medical records of the patient in the data subsets with different treatment plans, and obtain the treatment plan and treatment cost of the medical records;

[0051] In one embodiment, the method steps also include correcting the treatment plan and treatment costs of the obtained medical record to obtain a corrected treatment plan and treatment costs of the medical record, and outputting different treatment plans for the patient and the treatment costs of each treatment plan based on the corrected treatment plan and treatment costs of the case. The correction of the treatment plan and treatment costs of the medical record includes the correction of the treatment plan and / or the correction of the treatment costs of the medical record. The correction of the treatment costs is to restore the medical insurance exemption costs in the medical record to full self-paid costs to obtain the treatment costs of the corrected medical record; the correction of the treatment plan is to match and correct the drugs and drug prices in the medical record with the drugs and drug prices in the hospital pharmacy to obtain a corrected treatment plan for the medical record.

[0052] In one embodiment, the correction of the treatment plan is to match and correct the drugs and drug prices in the medical record with the drugs and drug prices in the hospital drug store to obtain a corrected treatment plan for the medical record. When the drugs in the treatment plan match the drugs in the hospital drug store, the drug prices are updated to obtain the treatment plan and treatment costs for the medical record after the prices are corrected. When the drugs in the treatment plan match the drugs in the hospital drug store, the drugs are replaced with drugs of the same type to obtain the treatment plan and treatment costs for the medical record after the drugs and prices are corrected. Optionally, when replacing drugs of the same type, drugs with the same generic name are first considered, and drugs with the same function are secondly considered.

[0053] 105: Output different treatment plans for the patient and the treatment cost of each treatment plan according to the treatment plan and treatment cost of the case.

[0054] In a specific embodiment, the method steps also include selecting the patient's settlement type before outputting different treatment plans for the patient and the treatment costs of each treatment plan based on the treatment plan and treatment costs of the case, and outputting different treatment plans and the treatment costs of each treatment plan under the patient's settlement type based on the patient's settlement type, the treatment plan and treatment costs of the case; optionally, the settlement types include ordinary patients, medical insurance patients or public medical patients.

[0055] Figure 2 A treatment cost prediction system based on machine learning is provided in an embodiment of the present invention, comprising:

[0056] An acquisition unit 201 is used to acquire basic information data and diagnostic information data of a patient;

[0057] Diagnosis unit 202, used to input the basic information data and the diagnosis information data into a pre-trained disease diagnosis model to obtain the disease type of the patient;

[0058] A retrieving unit 203 is used to retrieve data subsets of different treatment plans under the disease type from a hospital medical record database according to the disease type, wherein the data subsets of different treatment plans under the disease type from the hospital medical record database are obtained by performing cluster analysis on the hospital medical record database;

[0059] A matching unit 204 is used to compare the basic information data and diagnosis data of the patient with the basic information data and diagnosis data of the medical records in the data subset, match the closest medical records of the patient in the data subsets with different treatment plans, and obtain the treatment plan and treatment cost of the medical records;

[0060] The output unit 205 is used to output different treatment plans for the patient and the treatment cost of each treatment plan according to the treatment plan and treatment cost of the case.

[0061] Figure 3 A device for predicting treatment costs based on machine learning is provided in an embodiment of the present invention, comprising: a memory and a processor;

[0062] The memory is used to store program instructions;

[0063] The processor is used to call program instructions, and when the program instructions are executed, the above-mentioned treatment cost prediction method based on machine learning is implemented.

[0064] One object of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for predicting treatment costs based on machine learning.

[0065] The validation results of this validation example show that assigning intrinsic weights to indications can moderately improve the performance of the method relative to the default setting.

[0066] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0067] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0068] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0069] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0070] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0071] A person skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. The above-mentioned storage medium can be a read-only memory, a disk or an optical disk, etc.

[0072] The above is a detailed introduction to a computer device provided by the present invention. For a person skilled in the art, according to the concept of the embodiments of the present invention, there may be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for predicting treatment costs based on machine learning, characterized in that: The specific steps include: Obtain the patient's basic information and diagnostic information; Inputting the basic information data and the diagnostic information data into a pre-trained disease diagnosis model to obtain the patient's disease type; According to the disease type, retrieve a data subset of different treatment plans under the disease type from a hospital medical record database, wherein the data subset of different treatment plans under the disease type from the hospital medical record database is obtained by performing cluster analysis on the hospital medical record database; Wherein, the hospital medical record database is a dynamically updated database; The basic information data and diagnostic data of the patient are compared with the basic information data and diagnostic information data of the medical records in the data subset, and the medical records of the patient in the data subset with different treatment plans are matched to obtain the medical records closest to each other, and the treatment plans and treatment costs of the medical records are obtained; wherein, the processing of the treatment plans and treatment costs of the obtained medical records also includes correction, and the correction of the treatment plans and treatment costs of the medical records is obtained by correction, and the different treatment plans and treatment costs of each treatment plan of the patient are output according to the correction of the treatment plans and treatment costs of the medical records, and the correction of the treatment plans and treatment costs of the medical records includes the correction of the treatment plans of the medical records and the correction of the treatment costs, and the correction of the treatment costs is to restore the medical insurance exemption fees in the medical records to the full self-paid fees, and obtain the treatment costs of the corrected medical records; The correction of the treatment plan is to match and correct the drugs and drug prices in the medical record with the drugs and drug prices in the hospital drug store to obtain the treatment plan of the corrected medical record. When the drugs in the treatment plan and the drugs in the hospital drug store match each other, the drug prices are updated to obtain the treatment plan and treatment costs of the medical record after the price correction; when the drugs in the treatment plan and the drugs in the hospital drug store do not match each other, the drugs are replaced with drugs of the same type to obtain the treatment plan and treatment costs of the medical record after the drugs and prices are corrected; When replacing similar drugs, drugs with the same generic name should be considered first, followed by drugs with the same function; Output different treatment plans for the patient and the treatment cost of each treatment plan according to the treatment plan and treatment cost of the medical record and / or the revised medical record and treatment cost; Before outputting the different treatment plans and the treatment costs of each treatment plan for the patient according to the treatment plans and treatment costs in the medical record, select the patient's settlement type, and output the different treatment plans and the treatment costs of each treatment plan under the patient's settlement type according to the patient's settlement type, the treatment plans and treatment costs in the medical record.

2. The method for predicting treatment costs based on machine learning according to claim 1, characterized in that: The training process of the pre-trained disease diagnosis model is as follows: obtaining the patient's basic information data and diagnosis information data as a training set, training the machine learning model, comparing the predicted disease type with the actual disease type, generating a loss value, back-propagating, optimizing the disease diagnosis model, and obtaining the pre-trained disease diagnosis model.

3. The method for predicting treatment costs based on machine learning according to claim 2, characterized in that: The basic information data includes age, gender, admission time, and discharge time; the diagnostic information data includes admission diagnosis, examination and testing data, nursing diagnosis, will data, and discharge diagnosis.

4. The method for predicting treatment costs based on machine learning according to claim 2, characterized in that: The algorithm used by the pre-trained disease diagnosis model is selected from a variety of machine learning algorithms, and the machine learning algorithm is selected from one or more of the following machine learning algorithms: support vector machine, random forest, XGBoost, and LightGBM.

5. The method for predicting treatment costs based on machine learning according to claim 1, characterized in that: The process of constructing the data subsets of different treatment plans under the disease category in the hospital medical record database includes: obtaining all medical record sets under the disease category in the hospital medical record database, performing cluster analysis based on the differences in treatment plans, and obtaining data subsets of different treatment plans under the disease category.

6. The method for predicting treatment costs based on machine learning according to claim 5, characterized in that: The cluster analysis adopts FCM fuzzy clustering or K-Means clustering.

7. The method for predicting treatment costs based on machine learning according to claim 1, characterized in that: The settlement types include ordinary settlement, medical insurance settlement or public medical settlement.

8. A treatment cost prediction system based on machine learning, characterized in that: include: An acquisition unit, used to acquire basic information data and diagnostic information data of the patient; A diagnosis unit, used to input the basic information data and the diagnosis information data into a pre-trained disease diagnosis model to obtain the type of disease of the patient; A retrieving unit, configured to retrieve, according to the disease type, a data subset of different treatment plans under the disease type from a hospital medical record database, wherein the data subset of different treatment plans under the disease type from the hospital medical record database is obtained by performing cluster analysis on the hospital medical record database; Wherein, the hospital medical record database is a dynamically updated database; A matching unit is used to compare the basic information data and diagnostic data of the patient with the basic information data and diagnostic information data of the medical record in the data subset, match the closest medical record of the patient in the data subset of different treatment plans, and obtain the treatment plan and treatment cost of the medical record; wherein, the processing of the obtained treatment plan and treatment cost of the medical record also includes correction, and the correction of the treatment plan and treatment cost of the medical record is obtained by correction, and the different treatment plans and treatment costs of each treatment plan of the patient are output according to the correction of the treatment plan and treatment cost of the medical record, and the correction of the treatment plan and treatment cost of the medical record includes the correction of the treatment plan of the medical record and the correction of the treatment cost, and the correction of the treatment cost is to restore the medical insurance exemption fee in the medical record to the full self-paid fee, so as to obtain the treatment cost of the corrected medical record; The correction of the treatment plan is to match and correct the drugs and drug prices in the medical record with the drugs and drug prices in the hospital drug store to obtain the treatment plan of the corrected medical record. When the drugs in the treatment plan and the drugs in the hospital drug store match each other, the drug prices are updated to obtain the treatment plan and treatment costs of the medical record after the price correction; when the drugs in the treatment plan and the drugs in the hospital drug store do not match each other, the drugs are replaced with drugs of the same type to obtain the treatment plan and treatment costs of the medical record after the drugs and prices are corrected; When replacing similar drugs, drugs with the same generic name should be considered first, followed by drugs with the same function; An output unit, used for outputting different treatment plans for the patient and the treatment cost of each treatment plan according to the treatment plan and treatment cost of the medical record and / or the revised medical record and treatment cost; Before outputting the different treatment plans for the patient and the treatment costs of each treatment plan based on the treatment plans and treatment costs in the medical record, select the patient's settlement type, and output the different treatment plans and the treatment costs of each treatment plan under the patient's settlement type based on the patient's settlement type, the treatment plans and treatment costs in the medical record.

9. A device for predicting treatment costs based on machine learning, characterized in that: include: Memory and processor; The memory is used to store program instructions; The processor is used to call program instructions, and when the program instructions are executed, the method for predicting treatment costs based on machine learning described in any one of claims 1-7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting treatment costs based on machine learning according to any one of claims 1 to 7 is implemented.

Images