Customized treatment scheme recommendation method and system based on big data mining

Abstract

The invention discloses a customized treatment scheme recommendation method and system based on big data mining. The customized treatment scheme recommendation method comprises the following steps: collecting medical treatment data samples, and establishing a medical database; extracting sample feature keywords from the medical database; a customized treatment scheme model is trained according to the sample features and the test cases, and the model outputs a customized treatment scheme and gives a treatment effect prediction trend; and adjusting the customized treatment scheme according to the treatment effect prediction trend and the actual treatment trend. The medical treatment data sample is mined based on big data, and the customized treatment scheme is given according to the sample and the patient physical condition training model, so that the treatment effect is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for recommending customized treatment plans based on big data mining. Background Art

[0002] In the past, treatment was mostly based on the doctor's personal experience. With the development of computers, it has become easy to collect various medical data, and the significant improvement in computing power has made it possible to process and analyze these huge data sets. The development of technologies such as high-performance computing and cloud computing has provided strong technical support for big data mining. Treatment plans based on big data mining provide more accurate customized treatment plans by drawing on the experience of multiple successful cases and combining the patient's own basic conditions, thereby improving the treatment effect. In view of this, the present invention provides a method and system for recommending customized treatment plans based on big data mining. Summary of the invention

[0003] The present invention provides a customized treatment plan recommendation method based on big data mining, comprising:

[0004] S10, collect medical treatment data samples and build a medical database;

[0005] S20, extracting sample feature keywords from the medical database;

[0006] S30. Train a customized treatment plan model based on sample characteristics and test cases. The model outputs a customized treatment plan and gives a prediction trend of treatment effect.

[0007] S40. Adjust the customized treatment plan based on the predicted trend of treatment effect and the actual treatment trend.

[0008] As described above, a customized treatment plan recommendation method based on big data mining is used, in which the collected medical treatment data samples need to use data cleaning and standardization techniques to remove noise and inconsistency in the samples to ensure data quality.

[0009] As described above, a customized treatment plan recommendation method based on big data mining is described, wherein sample feature keywords are extracted based on given features, and based on the given features, descriptions related to the features are extracted as keywords from relevant documents of the sample.

[0010] In the above-mentioned customized treatment plan recommendation method based on big data mining, training the customized treatment plan model is specifically divided into the following sub-steps:

[0011] Communicate with patients to summarize test cases and treatment influencing factors;

[0012] Match the test case features with the features of the sample set in the database to form a feature group;

[0013] Match the samples with the most similar features to the test cases in the feature population set to form a historical sample set, so as to find the corresponding treatment plan set;

[0014] Prioritizing treatment options in a treatment option set;

[0015] Select the treatment plan with the highest priority for optimization and get a customized treatment plan;

[0016] Calculate treatment effect prediction trends based on impact factors.

[0017] As described above, a customized treatment plan recommendation method based on big data mining is provided, wherein the condition for a sample to enter a feature group set is that the features of the sample match the features of a test case, and if any feature of the test case exists in the sample features, the sample enters the feature group set.

[0018] As described above, a customized treatment plan recommendation method based on big data mining is described, in which the condition for a sample to enter the historical sample set is that a sample contains all the features of the test case, and then enters the historical sample set.

[0019] As described above, a customized treatment plan recommendation method based on big data mining is provided, wherein the optimized treatment plan is to replace the options in the plan that do not meet the conditions one by one until all options meet the conditions.

[0020] The present invention also provides a customized treatment plan recommendation system based on big data mining, including: a database module, an extraction module, a training module, and an adjustment module.

[0021] Database module: used to collect medical treatment data samples, process the samples using data cleaning and standardization techniques, and form a structured medical database.

[0022] Extraction module: used to extract feature description keywords of samples in the database and extract feature keywords of test cases.

[0023] Training module: used to match the test case features with the features of the sample set in the database to form a feature group; match the samples in the feature group set that are most similar to the test case features to form a historical sample set, so as to find the corresponding treatment plan set; prioritize the treatment plans in the treatment plan set; select the treatment plan with the highest priority for optimization to obtain a customized treatment plan.

[0024] Adjustment module: used to adjust customized treatment plans based on the predicted trend of treatment effects and the actual treatment trend.

[0025] The present invention also provides a computer storage medium, comprising at least one memory and at least one processor;

[0026] A memory for storing one or more program instructions;

[0027] A processor is used to run one or more program instructions to execute a customized treatment plan recommendation method based on big data mining.

[0028] The beneficial effects achieved by the present invention are as follows: the present invention mines medical treatment data samples based on big data, and provides customized treatment plans according to the samples and the patient's physical condition training model, thereby improving the treatment effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0030] Figure 1 This is a flow chart of a customized treatment plan recommendation method based on big data mining provided in Example 1 of the present application.

[0031] Figure 2 This is a schematic diagram of a customized treatment plan recommendation system based on big data mining provided in Example 2 of the present application. DETAILED DESCRIPTION

[0032] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0033] Embodiment 1

[0034] like Figure 1 As shown, Embodiment 1 of the present application provides a customized treatment plan recommendation method based on big data mining, including:

[0035] S10. Collect medical treatment data samples and build a medical database.

[0036] Medical treatment data samples are collected from multiple channels such as electronic health record systems, medical literature databases, and public health data platforms. Data cleaning and standardization techniques are used to remove noise and inconsistencies in the samples, ensure data quality, and form a structured medical database.

[0037] S20. Extract sample feature keywords from the medical database.

[0038] Assume that the number of samples in the medical database is n. For each sample i, extract sample feature description keywords based on the given features. Each sample has m features. The m feature attributes include disease category, disease cause, disease symptoms, allergy history, diagnosis result, treatment plan, treatment cost, reason for successful cure, reason for failed cure, etc. Each feature has j feature description keywords extracted. The formula for extracting feature description keywords is: It means extracting j descriptive keywords of m features in the i-th sample. Represents selection and feature t a The first j related keywords, Represents the extracted feature t a Keywords, TF(t a ,Z i ) represents the feature t a In the sample document Z i The frequency of occurrence in Represents the feature t a The inverse document frequency, N is the total number of words in the document, DF(t) is the number of words containing feature t a The number of documents.

[0039] The above extraction process is repeated to extract all the feature description keywords of n samples and store them in set S.

[0040] S30. A customized treatment plan model is trained based on sample characteristics and test cases. The model outputs a customized treatment plan and gives a predicted trend of treatment effect.

[0041] Training a customized treatment plan model is divided into the following sub-steps:

[0042] S31. Communicate with patients and summarize test cases and treatment influencing factors.

[0043] Before building a customized treatment plan model, you first need to communicate with the patient and record the communication content to form a test case.

[0044] Treatment influencing factors include communication feature vectors, medical history feature vectors, and physical factor feature vectors.

[0045] Communication between patients and doctors. Good communication can improve treatment compliance and effectiveness. The characteristics of communication can be obtained through the frequency and content of communication. The specific formula is C i ={c i1 , c i2 , c i3}, where Ci represents the communication feature vector of the ith patient, c i1 represents the communication frequency, c i2 Indicates the communication method. Different communication methods will affect the effect of information transmission. i3 Indicates the content of communication.

[0046] The patient's medical history is crucial to the choice of treatment plan. The specific formula is E i ={e i1 , e i2 , e i3 , e i4}, where E i represents the medical history feature vector of the i-th patient, e i1 Indicates previous illness, the illness the patient had in the past and its treatment, e i2 Indicates family medical history, including hereditary diseases, infectious diseases, i3 Indicates surgical history, the patient's past surgical experience and its results. Surgical history can help assess the patient's physical recovery ability and potential risks. i4 Indicates allergic condition.

[0047] The patient's physical factors will also affect the effect of treatment. The specific formula is D i ={d i1 , d i2 , d i3 , d i4}, where D i represents the physical factor feature vector of the i-th patient, d i1 represents the patient's age, d i2 represents the patient's gender, d i3 Indicates the patient's living habits, i4 Indicates the patient's physical condition.

[0048] S32. Extract test case features, match the test case features with features of a sample set in a database, and form a feature group.

[0049] Use the extraction formula in step S20 to extract the features of the test case. Suppose the number of test case features that are not empty after extraction is k, k≤m, and the feature set of the test case is f k , the keyword set of the test case is

[0050] The feature set of the test case is matched with the feature set of the sample. The specific formula is: R represents the feature group set, k represents the test case has k features to match, represents the indicator function, f brepresents the bth feature of the test case, Indicates the qth feature of the ith sample among n samples. If any feature of the test case does not exist in a certain sample, the indicator function δ=0, and the sample does not enter the feature population set R; if any feature of the test case exists in a certain sample, the indicator function δ=1, and the sample enters the feature population set R.

[0051] S33. Match the samples that are most similar to the test case features in the feature group set to form a historical sample set, so as to find the corresponding treatment plan set.

[0052] Match samples with all the features of the test case in the feature group set and enter the historical sample set.

[0053] Assume that there are z samples in the feature population set R, define the feature matching indicator function I, and if a sample contains all the features of the test case, it will enter the historical sample set. Specifically, the matching formula is W represents the historical sample set, f c is the cth feature of the test case, is the qth feature in the i-th sample in the feature population set R.

[0054] For samples that enter the historical sample set, their treatment plans enter the treatment plan set P.

[0055] S34. Prioritize the treatment plans in the treatment plan set.

[0056] Perform keyword matching on samples in the historical sample set. The higher the keyword matching score of the sample, the higher the priority of its treatment plan. Define the matching score function Score to calculate the sample keyword matching score.

[0057] Specifically, the formula is f represents the test case, W z represents the sample in the historical sample set, β represents the proportion score of the test case feature in the sample feature, α is the parameter for adjusting the weight of features and keywords, k represents the number of features of the test case, is a set of keywords that characterize the test case. It is a set of keywords of features in the sample that correspond to the features of the test case.

[0058] S35. Select the treatment plan with the highest priority for optimization and obtain a customized treatment plan.

[0059] Assume that the treatment plan set P = {p1, p2, ..., p o}, select the treatment plan p1 with the highest priority in the treatment plan set P, which contains x options, and the plan is p1 = {h1,h2,...,h x}, replace y options in the treatment plan based on the patient's physical condition in the test case and optimize the treatment plan. The value range of y is 0≤y≤x.

[0060] Specifically, the replacement formula is For the replaced options, From the treatment options p2, p3, ..., p o Each treatment option is examined in turn. ψ The yth θ Options if Satisfy the conditions That is, this option is not related to the patient's medical history E i , Patient's physical factors i If there is a conflict, select this option as the replacement option, otherwise check the next treatment option p ψ+1 . min means selecting the option with the smallest index ψ among all the options that meet the conditions, that is, the option that meets the conditions first.

[0061] The above replacement formula is looped until y options are replaced, and the optimized treatment plan p'1, ​​that is, the customized treatment plan, is obtained.

[0062] S36. Calculate the predicted trend of treatment effect based on treatment influencing factors.

[0063] The treatment impact factor includes the communication eigenvector C i , medical history feature vector E i , physical factor feature vector D i The specific formula is Among them, P actual represents the prediction trend of the treatment effect of the optimized treatment plan, p'1 represents the optimized treatment plan, φ1 represents the weight of the communication feature vector, C iq Represents the communication feature vector C i The qth feature in , φ2 represents the weight of the medical history feature vector, E iq Represents the medical history feature vector E i The qth feature in φ3 represents the weight of the body factor feature vector, D iq Represents the physical factor feature vector D i The qth feature in Indicates the impact of the patient's emotional state on communication, Indicates the impact of long-term drug use on medical history, Indicates the impact of occupational environment on the body.

[0064] S40. Adjust the customized treatment plan based on the predicted trend of treatment effect and the actual treatment trend.

[0065] The patient's actual treatment trend reflects his or her response to the current customized treatment plan. The customized treatment plan is adjusted according to the actual treatment trend. The specific formula for the actual treatment trend is T i ={t i1 , t i2 , t i3}, T i represents the actual treatment trend feature vector of the i-th patient, t i1 Indicates the changes in the patient's symptoms during treatment, t i2 Indicates the change of inspection index, t i3 Indicates changes in patients' quality of life.

[0066] The customized treatment plan is adjusted according to the predicted trend of treatment effect and the actual treatment trend. The specific formula is: Where P adjust For the adjusted customized treatment plan, L phase Indicates the treatment stage, including the initial treatment stage and the stable treatment stage. i ,P actual ) is used to compare the actual treatment trend T i P actual The result value of the deviation is different. Depending on the treatment stage, the threshold of the deviation between the actual treatment trend and the predicted trend of the treatment effect is different. If the deviation exceeds the threshold V threshold , you need to use the replacement formula Replace and adjust the treatment plan to obtain better treatment effect.

[0067] Embodiment 2

[0068] like Figure 2 As shown, Embodiment 2 of the present application provides a customized treatment plan recommendation system based on big data mining, including:

[0069] Database module: used to collect medical treatment data samples, process the samples using data cleaning and standardization techniques, and form a structured medical database.

[0070] Medical treatment data samples are collected from multiple channels such as electronic health record systems, medical literature databases, and public health data platforms. Data cleaning and standardization techniques are used to remove noise and inconsistencies in the samples, ensure data quality, and form a structured medical database.

[0071] Extraction module: used to extract feature description keywords of samples in the database and extract feature keywords of test cases.

[0072] Extract the characteristic description keywords of samples in the database:

[0073] Assume that the number of samples in the medical database is n. For each sample i, extract sample feature description keywords based on the given features. Each sample has m features. The m feature attributes include disease category, disease cause, disease symptoms, allergy history, diagnosis result, treatment plan, treatment cost, reason for successful cure, reason for failed cure, etc. Each feature has j feature description keywords extracted. The formula for extracting feature description keywords is: It means extracting j descriptive keywords of m features in the i-th sample. Represents selection and feature t a The first j related keywords, Represents the extracted feature t a Keywords, TF(t a ,D i ) represents the feature t a In the sample document D i The frequency of occurrence in Represents the feature t a The inverse document frequency, N is the total number of words in the document, DF(t) is the number of words containing feature t a The number of documents.

[0074] The above extraction process is repeated to extract all the feature description keywords of n samples and store them in set S.

[0075] Extract test case feature keywords:

[0076] Use the formula extracted in the above steps to extract the features of the test case. Suppose the number of test case features that are not empty after extraction is k, k≤m, and the feature set of the test case is f k , the keyword set of the test case is

[0077] Training module: includes summary submodule, feature group submodule, treatment plan set submodule, sorting submodule, optimization submodule, and prediction trend submodule.

[0078] The summary submodule is used to communicate with patients and summarize test cases and treatment influencing factors.

[0079] Before building a customized treatment plan model, you first need to communicate with the patient and record the communication content to form a test case.

[0080] Treatment influencing factors include communication feature vectors, medical history feature vectors, and physical factor feature vectors.

[0081] Communication between patients and doctors. Good communication can improve treatment compliance and effectiveness. The characteristics of communication can be obtained through the frequency and content of communication. The specific formula is C i ={c i1 , c i2 , c i3}, where C i Indicates i The communication feature vector of each patient, c i1 represents the communication frequency, c i2 Indicates the communication method. Different communication methods will affect the effect of information transmission. i3 Indicates the content of communication.

[0082] The patient's medical history is crucial to the choice of treatment plan. The specific formula is E i ={e i1 , e i2 , e i3 , e i4}, where E i represents the medical history feature vector of the i-th patient, e i1 Indicates previous illness, the illness the patient had in the past and its treatment, e i2 Indicates family medical history, including hereditary diseases, infectious diseases, i3 Indicates surgical history, the patient's past surgical experience and its results. Surgical history can help assess the patient's physical recovery ability and potential risks. i4 Indicates allergic condition.

[0083] The patient's physical factors will also affect the effect of treatment. The specific formula is D i ={d i1 , d i2 , d i3 , d i4}, where D i represents the physical factor feature vector of the i-th patient, d i1 represents the patient's age, d i2 represents the patient's gender, d i3 Indicates the patient's living habits, i4 Indicates the patient's physical condition.

[0084] The feature group submodule is used to match the test case features with the features of the sample set in the database to form a feature group.

[0085] The feature set of the test case is matched with the feature set of the sample. The specific formula is: R represents the feature group set, k represents the test case has k features to match, represents the indicator function, f b represents the bth feature of the test case, Indicates the qth feature of the ith sample among n samples. If any feature of the test case does not exist in a certain sample, the indicator function δ=0, and the sample does not enter the feature population set R; if any feature of the test case exists in a certain sample, the indicator function δ=1, and the sample enters the feature population set R.

[0086] The treatment plan set submodule is used to match the samples with the most similar features to the test case in the feature population set to form a historical sample set, so as to find the corresponding treatment plan set.

[0087] Match samples with all the features of the test case in the feature group set and enter the historical sample set.

[0088] Assume that there are z samples in the feature population set R, define the feature matching indicator function I, and if a sample contains all the features of the test case, it will enter the historical sample set. Specifically, the matching formula is W represents the historical sample set, f c is the cth feature of the test case, is the qth feature in the i-th sample in the feature population set R.

[0089] For samples that enter the historical sample set, their treatment plans enter the treatment plan set P.

[0090] The sorting submodule prioritizes the treatment plans in the treatment plan set.

[0091] Perform keyword matching on samples in the historical sample set. The higher the keyword matching score of the sample, the higher the priority of its treatment plan. Define the matching score function Score to calculate the sample keyword matching score.

[0092] Specifically, the formula is f represents the test case, W z represents the sample in the historical sample set, β represents the proportion score of the test case feature in the sample feature, α is the parameter for adjusting the weight of features and keywords, k represents the number of features of the test case, is a set of keywords that characterize the test case. It is a set of keywords of features in the sample that correspond to the features of the test case.

[0093] The optimization submodule is used to select the treatment plan with the highest priority for optimization and obtain a customized treatment plan.

[0094] Assume that the treatment plan set P = {p1, p2, ..., po}, select the treatment plan p1 with the highest priority in the treatment plan set P, which contains x options, and the plan is p1 = {h1,h2,...,h x}, replace y options in the treatment plan based on the patient's physical condition in the test case and optimize the treatment plan. The value range of y is 0≤y≤x.

[0095] Specifically, the replacement formula is For the replaced options, From the treatment options p2, p3, ..., p o Each treatment option is examined in turn. ψ The yth θ Options if Satisfy the conditions That is, this option is not related to the patient's medical history E i , Patient's physical factors i If there is a conflict, select this option as the replacement option, otherwise check the next treatment option p ψ+1 . min means selecting the option with the smallest index ψ among all the options that meet the conditions, that is, the option that meets the conditions first.

[0096] The above replacement formula is looped until y options are replaced, and the optimized treatment plan p'1, ​​that is, the customized treatment plan, is obtained.

[0097] The prediction trend submodule is used to calculate the prediction trend of treatment effect based on the influencing factors.

[0098] The treatment impact factor includes the communication eigenvector C i , medical history feature vector E i , physical factor feature vector D i The specific formula is Among them, P actual represents the prediction trend of the treatment effect of the optimized treatment plan, p'1 represents the optimized treatment plan, φ1 represents the weight of the communication feature vector, C iq Represents the communication feature vector C i The qth feature in , φ2 represents the weight of the medical history feature vector, E iq Represents the medical history feature vector E i The qth feature in φ3 represents the weight of the body factor feature vector, D iq Represents the physical factor feature vector D i The qth feature in Indicates the impact of the patient's emotional state on communication, Indicates the impact of long-term drug use on medical history, Indicates the impact of occupational environment on the body.

[0099] The adjustment module is used to adjust the customized treatment plan according to the predicted trend of treatment effect and the actual treatment trend.

[0100] The patient's actual treatment trend reflects his or her response to the current customized treatment plan. The customized treatment plan is adjusted according to the actual treatment trend. The specific formula for the actual treatment trend is T i ={t i1 , t i2 , t i3}, T i represents the actual treatment trend feature vector of the i-th patient, t i1 Indicates the changes in the patient's symptoms during treatment, t i2 Indicates the change of inspection index, t i3 Indicates changes in patients' quality of life.

[0101] The customized treatment plan is adjusted according to the predicted trend of treatment effect and the actual treatment trend. The specific formula is: Where P adjust For the adjusted customized treatment plan, L phase Indicates the treatment stage, including the initial treatment stage and the stable treatment stage. i ,P actual ) is used to compare the actual treatment trend T i P actual The result value of the deviation is different. Depending on the treatment stage, the threshold of the deviation between the actual treatment trend and the predicted trend of the treatment effect is different. If the deviation exceeds the threshold V threshold , you need to use the replacement formula Replace and adjust the treatment plan to obtain better treatment effect.

[0102] Corresponding to the above embodiment, an embodiment of the present invention provides a computer storage medium, including: at least one memory and at least one processor;

[0103] The memory is used to store one or more program instructions;

[0104] A processor is used to run one or more program instructions to execute a customized treatment plan recommendation method based on big data mining.

[0105] Corresponding to the above-mentioned embodiments, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer storage medium contains one or more program instructions, and the one or more program instructions are used by a processor to execute a customized treatment plan recommendation method based on big data mining.

[0106] The embodiments disclosed in the present invention provide a computer-readable storage medium, in which computer program instructions are stored. When the computer program instructions are executed on a computer, the computer executes the above-mentioned customized treatment plan recommendation method based on big data mining.

[0107] In the embodiment of the present invention, the processor may be an integrated circuit chip having the ability to process signals. The processor may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0108] The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present invention can be directly embodied as a hardware decoding processor for execution, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The processor reads the information in the storage medium and completes the steps of the above method in combination with its hardware.

[0109] The storage medium may be a memory, which may be, for example, a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memory.

[0110] Among them, the non-volatile memory can be a read-only memory (ROM), a programmable ROM (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory.

[0111] The volatile memory may be a random access memory (RAM) which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDRSDRAM), enhanced synchronous DRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus DRAM (DRRAM).

[0112] The storage media described in the embodiments of the present invention are intended to include, but are not limited to, these and any other suitable types of memory.

[0113] Those skilled in the art will appreciate that in one or more of the above examples, the functions described in the present invention can be implemented using a combination of hardware and software. When software is used, the corresponding functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any media that facilitates the transmission of computer programs from one place to another. Storage media can be any available media that can be accessed by general or special-purpose computers.

[0114] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A customized treatment plan recommendation method based on big data mining, characterized in that: include: S10, collect medical treatment data samples and build a medical database; S20, extracting sample feature keywords from the medical database; S30. Train a customized treatment plan model based on sample characteristics and test cases. The model outputs a customized treatment plan and gives a prediction trend of treatment effect. S40. Adjust the customized treatment plan based on the predicted trend of treatment effect and the actual treatment trend.

2. A customized treatment plan recommendation method based on big data mining as claimed in claim 1, characterized in that: The collected medical treatment data samples need to use data cleaning and standardization techniques to remove noise and inconsistencies in the samples to ensure data quality.

3. The method for recommending customized treatment plans based on big data mining according to claim 1, characterized in that: Extracting sample feature keywords is done based on given features. Based on the given features, descriptions related to the features are extracted from relevant documents of the sample as keywords.

4. The method for recommending customized treatment plans based on big data mining according to claim 1, characterized in that: Training a customized treatment plan model is divided into the following sub-steps: Communicate with patients to summarize test cases and treatment influencing factors; Extract test case features, match the test case features with the features of the sample set in the database to form a feature group; Match the samples with the most similar features to the test cases in the feature population set to form a historical sample set, so as to find the corresponding treatment plan set; Prioritizing treatment options in a treatment option set; Select the treatment plan with the highest priority for optimization and get a customized treatment plan; Calculate treatment effect prediction trends based on impact factors.

5. The method for recommending customized treatment plans based on big data mining according to claim 4, characterized in that: The condition for a sample to enter the feature population set is that the sample's features match the test case's features. If any feature of the test case exists in the sample's features, the sample enters the feature population set.

6. A customized treatment plan recommendation method based on big data mining as claimed in claim 4, characterized in that: The condition for a sample to enter the historical sample set is that a sample contains all the features of the test case, then it enters the historical sample set.

7. The method for recommending customized treatment plans based on big data mining according to claim 4, characterized in that: Optimizing the treatment plan is to replace the options that do not meet the requirements one by one until all options meet the requirements.

8. A customized treatment plan recommendation system based on big data mining, characterized in that: include: Database module: used to collect medical treatment data samples, process the samples using data cleaning and standardization techniques, and form a structured medical database. Extraction module: used to extract feature description keywords of samples in the database and extract feature keywords of test cases. Training module: used to communicate with patients, summarize test cases and treatment influencing factors; match test case features with features of sample sets in the database to form feature groups; match samples in the feature group set that are most similar to test case features to form a historical sample set, thereby finding a corresponding set of treatment plans; prioritize treatment plans in the treatment plan set; select the treatment plan with the highest priority for optimization to obtain a customized treatment plan; calculate treatment effect prediction trends based on influencing factors. Adjustment module: used to adjust customized treatment plans based on the predicted trend of treatment effects and the actual treatment trend.

9. A computer storage medium, characterized in that include: at least one memory and at least one processor; A memory for storing one or more program instructions; A processor, used to run one or more program instructions to execute a method for implementing a customized treatment plan recommendation system based on big data mining as described in any one of claims 1-7.

Images