CN111260448A - Artificial intelligence-based medicine recommendation method and related equipment - Google Patents
Artificial intelligence-based medicine recommendation method and related equipment Download PDFInfo
- Publication number
- CN111260448A CN111260448A CN202010090681.0A CN202010090681A CN111260448A CN 111260448 A CN111260448 A CN 111260448A CN 202010090681 A CN202010090681 A CN 202010090681A CN 111260448 A CN111260448 A CN 111260448A
- Authority
- CN
- China
- Prior art keywords
- medicine
- recommendation
- target
- probability value
- data
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Game Theory and Decision Science (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Chemical & Material Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Medicinal Chemistry (AREA)
- Epidemiology (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Medical Treatment And Welfare Office Work (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention discloses a medicine recommending method, a medicine recommending system, computer equipment and a computer readable storage medium based on artificial intelligence, wherein the method comprises the following steps: acquiring historical medical data to construct an original portrait; preprocessing an original image and outputting a characteristic data set; the preprocessed feature data set is used as training data and is respectively trained in advance to obtain an XGB OST recommendation model and a deep neural network recommendation model; obtaining a first recommendation strength probability value and a second recommendation strength probability value through an XGBOOST recommendation model and a deep neural network recommendation model; then linear addition is carried out to obtain a target recommendation strength probability value; and screening target medicine information of a preset number before ranking according to the target recommendation strength probability value, and pushing the target medicine information to the user. The invention linearly adds the model results of the two, and can effectively improve the accuracy of drug recommendation.
Description
Technical Field
The invention relates to the field of medicines, in particular to a medicine recommendation method and system based on artificial intelligence, computer equipment and a computer readable storage medium.
Background
When a traditional patient purchases a medicine, the patient usually chooses to go to an entity medicine retail store such as a chain pharmacy to purchase the medicine, and particularly the patient describes the symptoms of the patient to a pharmacist, and the pharmacist chooses the corresponding medicine to the patient, but the mode has the following defects that the levels of medical practitioners in the pharmacy may be uneven, so that the recommended medicine is not necessarily suitable for the patient, and in addition, each entity store needs to be equipped with a plurality of pharmacists to further cause the labor cost to rise rapidly, so a plurality of systems capable of recommending the medicine to the user are developed, most of the existing medicine recommending systems basically match with medicine indications based on the medicine indications, theoretically, the medicine recommending systems are feasible, and a large amount of specialized calendar data are ignored. The completeness and effectiveness of the calendar data will determine the accuracy of drug recommendation, and therefore, how to improve the accuracy of drug recommendation is a problem to be solved at present.
Disclosure of Invention
The invention aims to provide a medicine recommending method, a medicine recommending system, computer equipment and a computer readable storage medium based on artificial intelligence, which can effectively improve the accuracy of medicine recommendation.
In order to achieve the purpose, the invention provides a medicine recommendation method based on artificial intelligence, which comprises the following steps:
acquiring historical medical data to construct an original portrait;
preprocessing an original image and outputting a characteristic data set;
the preprocessed feature data set is used as training data and is respectively trained in advance to obtain an XGB OST recommendation model and a deep neural network recommendation model;
acquiring medicine request information of a user and inputting the medicine request information into an XGB OST recommendation model and a deep neural network recommendation model;
the XGB recommendation model calculates and obtains a first recommendation strength probability value of a target drug corresponding to the drug request information according to the drug request information;
the deep neural network recommendation model calculates and obtains a second recommendation strength probability value of the target medicine corresponding to the medicine request information according to the medicine request information;
linearly adding the first recommended strength probability value and the second recommended strength probability value of each target medicine to obtain a target recommended strength probability value of each target medicine;
and sorting the target medicine information of each target medicine from high to low according to the corresponding target recommendation strength probability value, screening out the preset number of target medicine information before ranking, and pushing the target medicine information to the user.
In particular, the original image includes basic information and medical record information, the data source of the original image includes a plurality of different data sources, and the multidimensional characteristics included in the plurality of different data sources are inconsistent.
Specifically, when historical medical data are collected, a table corresponding relation is built between data in the historical medical data and data in a correlation database through a correlation extraction method, and correlation data in the correlation database are collected to construct an original portrait.
In particular, the preprocessing includes feature expansion, saturation exploration screening, missing value padding, dataset heterogeneity and expansion, feedback selection, and relevance screening.
In particular, the raw portrait includes features of an n × m dimensional matrix, and the pre-processing the raw portrait includes: simplifying the characteristics of the n x m dimensional matrix into the transverse splicing of a plurality of n x 1 dimensional vectors by using a tsfresh algorithm, and then performing particle scanning, wherein the particle scanning comprises single-dimensional particle scanning and multi-dimensional particle scanning.
In particular, the pre-processing of the original representation comprises: the single-dimensional particle scanning is performed with sliding scanning to derive a single-dimensional extended feature set L1, the multi-dimensional particle scanning is performed with sliding scanning to derive a multi-dimensional extended feature set L2, and then the single-dimensional extended feature set, the multi-dimensional extended feature set and the original image are spliced to obtain a feature data set containing more information.
Particularly, when the historical medical data is collected, the classification of the non-standard medicine names comprises the following steps of respectively carrying out word segmentation processing on the non-standard medicine names and the standard medicine names;
based on a Word2Vec shallow neural network and based on a skip _ gram algorithm, respectively converting Word vectors of words in the non-standard medicine name and the standard medicine name;
longitudinally splicing the vectors by using a pooling rule to obtain word vector splicing of the non-standard medicine name and the standard medicine name;
calculating the vector distance between the non-standard medicine name and the standard medicine name;
finding out the standard medicine name which is closest to the vector distance of the non-standard medicine name;
and classifying the non-standard medicine names according to the classification standard of the standard medicine names with the closest vector distance.
The invention also provides a medicine recommendation system based on artificial intelligence, which comprises:
a collection module for collecting historical medical data to construct an original representation;
the preprocessing module is used for preprocessing the original image and outputting a characteristic data set;
the model training module is used for respectively training the preprocessed feature data set as training data in advance to obtain an XGB OST recommendation model and a deep neural network recommendation model;
the input module is used for acquiring the drug request information of the user and inputting the drug request information into the XGB OST recommendation model and the deep neural network recommendation model;
the processing module is used for calculating and obtaining a first recommended strength probability value of a target medicine corresponding to the medicine request information through an XGBOOST recommendation model and calculating and obtaining a second recommended strength probability value of the target medicine corresponding to the medicine request information through a deep neural network recommendation model; linearly adding the first recommended strength probability value and the second recommended strength probability value of each target medicine to obtain a target recommended strength probability value of each target medicine;
and the output module is used for sorting the target medicine information of each target medicine from high to low according to the corresponding target recommendation strength probability value, screening out the preset number of target medicine information before ranking, and pushing the target medicine information to the user.
The invention also provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method when executing the computer program.
The invention further provides a computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method.
According to the method, the XGB OST recommendation model and the deep neural network recommendation model are combined, model results of the XGB recommendation model and the deep neural network recommendation model are linearly added, the recommendation accuracy is improved by 200% compared with that of a non-model, and the correct medicine recommendation rate of the first three ranked medicines is improved from 75% to 95%. The invention extracts more hidden information from the historical medical data in the original portrait by characteristic expansion to ensure the integrity of characteristic values, and utilizes saturation to explore, screen and delete characteristic values without training significance, thereby forming a complete and effective characteristic data set. In addition, the XGB OST recommendation model and the deep neural network recommendation model are combined, so that the degree of automation is high, self-adaptive training and prediction can be realized, and the medicine recommendation efficiency is improved.
Drawings
FIG. 1 is a flow chart of an artificial intelligence based drug recommendation method of the present invention;
FIG. 2 is a flowchart of the categorization process when the historical medical data of step S10 of FIG. 1 relates to a non-standard drug name;
FIG. 3 is a block diagram of an artificial intelligence based drug recommendation system in accordance with the present invention;
fig. 4 is a schematic diagram of a hardware structure of a computer device of the artificial intelligence-based drug recommendation method according to the present invention.
Reference numerals:
1. artificial intelligence-based medicine recommendation system 10, acquisition module 20 and preprocessing module
30. Model training module 40, input module 50, processing module
60. Output module 2, computer device 21, memory 22, processor
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, the artificial intelligence based drug recommendation method of the present invention includes a process of forming a drug recommendation model by pre-training and a process of recommending drugs according to the drug recommendation model, and includes the following steps:
step S10: acquiring historical medical data to construct an original portrait;
step S20: preprocessing an original image and outputting a characteristic data set;
step S30: the preprocessed feature data set is used as training data and is respectively trained in advance to obtain an XGB OST recommendation model and a deep neural network recommendation model;
step S40: acquiring medicine request information of a user and inputting the medicine request information into an XGB OST recommendation model and a deep neural network recommendation model;
step S50: the XGB recommendation model calculates and obtains a first recommendation strength probability value of a target drug corresponding to the drug request information according to the drug request information;
step S60: the deep neural network recommendation model calculates and obtains a second recommendation strength probability value of the target medicine corresponding to the medicine request information according to the medicine request information;
step S70: linearly adding the first recommended strength probability value and the second recommended strength probability value of each target medicine to obtain a target recommended strength probability value of each target medicine;
step S80: and sorting the target medicine information of each target medicine from high to low according to the corresponding target recommendation strength probability value, screening out the preset number of target medicine information before ranking, and pushing the target medicine information to the user.
Referring to fig. 2, fig. 2 is a flowchart illustrating a classification process performed when the historical medical data of step S10 in fig. 1 relates to a non-standard drug name, and if the drug name in the historical medical data is not the strict national standard ICD10 code, the drug cannot be accurately located and classified. The method positions and matches the non-standard medicine names in the historical medical data through the text extraction matching method of the readable medium, so that the non-standard medicine names can be accurately positioned and classified.
The classification of non-standard drug names comprises the following steps:
step S110: and respectively carrying out word segmentation processing on the non-standard medicine name and the standard medicine name.
Step S120: based on a Word2Vec shallow neural network and based on a skip _ gram algorithm, respectively converting Word vectors of words in the non-standard medicine name and the standard medicine name; for example, through the network training of Word2vec, the Word "feel" is converted into a vector [1.90334,2.9874, …,0.988], dimension 5 × 1.
Step S130: longitudinally splicing the vectors by using a pooling rule to obtain word vector splicing of the non-standard medicine name and the standard medicine name; for example, cold drugs can be converted into vectors [1.0334,10.51184, …,1.8115], dimension 15 × 1.
Step S140: and calculating the vector distance between the non-standard medicine name and the standard medicine name.
Step S150: and finding out the standard medicine name which is closest to the vector distance of the non-standard medicine name.
Step S160: and classifying the non-standard medicine names according to the classification standard of the standard medicine names with the closest vector distance.
In this embodiment, the original image in step S10 includes basic information and medical record information, the data source of the original image includes a plurality of different data sources, and the multidimensional features included in the plurality of different data sources are inconsistent.
The basic information comprises field information of sex, age, blood type, marriage, mobile phone number and the like of a medicine purchaser or a medical staff;
the medical record information includes medical record summaries such as current symptoms, duration of symptoms, severity of symptoms, surgical history, medical cost, drugs, drug dosage form, and days of hospital stay, orders for hospital stay, etc. Medical record information of acute diseases such as influenza, chicken pox, hand, foot and mouth, corresponding disease onset time, medication and the like; follow-up information of chronic disease information such as hypertension, diabetes, chronic obstructive pulmonary disease and the like comprises duration of disease, medication and the like.
The multidimensional characteristics of different data sources are inconsistent, for example, the multidimensional characteristics of a sentry point hospital are rich and sufficient, and generally include sex, age, blood type, marital, diastolic pressure, systolic pressure, various physical examinations, and check indexes, and medium and small outpatients, community hospitals and the like may only have basic information; there are fewer brand pharmacies, and there may be only user mobile phone numbers, medicines, medicine formulations, etc.
In step S10, while acquiring the historical medical data, a table correspondence relationship is established between data in the historical medical data and data in an association database by an association extraction method, and the association data in the association database is acquired to construct an original portrait, wherein the association database is a database outside the medical field, and may be a database in multiple fields such as insurance, finance, science and technology, and the association database may be an internal database or an external free database (e.g., MySQL database), and information in the original portrait is subjected to multi-dimensional expansion and complementation by a multi-table main key association extraction method, so as to construct a more complete original portrait. The primary key is an ID, such as a social security account/identity number of a person after data desensitization.
Taking the associated data as the data of the financial field as an example, establishing the consumption capability characteristic dimension of the user by analyzing and mining the data of the financial field associated database so as to expand the consumption capability characteristic dimension of the personal portrait, and establishing the consumption capability characteristic dimension of the user if the data of the financial field associated database comprises financing and unsaying insurance which do not participate in the financial derivatives, namely the consumption capability is low; and if the data of the financial field association database comprises various insurance for managing and purchasing lots of financial derivatives, establishing the consumption capability characteristic dimension of the user, namely high consumption capability, wherein the consumption capability characteristic dimension of the user is used as training data for training to obtain an XGB OST recommendation model and a deep neural network recommendation model. Medicines recommended by the XGBOST recommendation model and the deep neural network recommendation model which are obtained by training of training data with low consumption capability correspond to the consumption capability of the medicines, namely cheap medicines with slow effect are recommended; the XGBOST recommendation model and the deep neural network recommendation model which are obtained by training the training data with high consumption capability recommend medicines corresponding to the consumption capability, namely, the medicines with high cost and remarkable effect are recommended.
In step S10, other related data are collected while collecting the historical medical data, and information in the original portrait is subjected to multi-dimensional expansion and supplementation by a multi-table main key related extraction method.
In the present embodiment, the preprocessing in step S20 includes feature expansion, saturation exploration screening, missing value padding, data set heterogeneity and expansion, feedback selection, and correlation screening.
The feature expansion is performed as follows, and a plurality of intermediate statistical variables such as variance, standard deviation, extreme value, various types of mean values, and the like are calculated by using the tsfresh algorithm for expansion. For example, each two columns of features f1, f2 are calculated as statistical indexes such as variance, standard deviation, extreme value, various mean values, etc. After the expansion, some columns of data are missing, or the characteristic column with unobvious distribution, such as 0 or-1 in the whole column, is indistinguishable, and meaningless to input the medicine recommendation model, and the data columns are deleted.
In this embodiment, the original image includes features of an n × m dimensional matrix, and the features of the n × m dimensional matrix are reduced to horizontal concatenation of n × 1 dimensional vectors, such as { H _1, H _2, …, H _ m } by using tsfresh algorithm when the original image is preprocessed, and then particle scanning is performed, where the particle scanning includes single-dimensional particle scanning and multi-dimensional particle scanning.
In this embodiment, when the original image is preprocessed, the single-dimensional particle scan is performed by sliding scan to derive a single-dimensional extended feature set L1, the multi-dimensional particle scan is performed by sliding scan to derive a multi-dimensional extended feature set L2, and then the single-dimensional extended feature set, the multi-dimensional extended feature set, and the original image are merged to obtain a feature data set with more information content.
Taking a single-dimensional particle scan as an example, each H is a vector of n × 1 dimensions.
First, a scan particle of a × 1 with window _ size k is defined for scanning. When scanning in a single dimension, a sliding scan is performed using a particle a × 1 with window _ size k. And when the particles slide to the corresponding single dimension, performing function calculation on the scanning data a-x 1 obtained by the particles on the single dimension, wherein the function can be various statistical indexes such as functions of statistical indexes such as variance, standard deviation, extreme values, mean values and the like, and can also be nonlinear functions such as tanh, relu and the like. And calculating to obtain a particle scanning value. Then, window sliding is carried out, scanning is carried out, then window sliding scanning is carried out, and the process is circulated until the whole single-column scanning is finished, and multi-column variance, standard deviation, extreme value and other multi-dimensional features can be derived from each single-column feature. And (5) converting the derived feature concat into a single-dimension extended feature set L1. The features in the original feature image are combined with the derived single-dimensional extended feature set to form a preprocessed feature data set.
Taking multi-dimensional particle scanning as an example, C may be taken as a list as an alternative dimension. Then, C1, C2 and the like are cyclically taken out from C for operation. Assuming that c2 is taken out this time, the cycle scans c2 column features with particle morphology d × c2, and a new window _ size _2 is defined in the same way, and the scan is slid on the c2 dimensional vector. The data obtained by scanning is also subjected to function calculation.
Finally, each set of c2 features may be derived a set of derivative features. And the c2 scan may be performed cyclically for all features. In addition, different small C of C, such as C1, C3, …, can be taken for scanning. Finally, the image concat obtained by multi-particle scanning is converted into a multi-dimensional extended feature set L2. Finally, L1, L2 are spliced with the original image to obtain an image W containing more information. The features in the original feature image are combined with the derived multidimensional expansion feature set to form a preprocessed feature data set.
The saturation exploration screening is implemented as follows, and exploration screening is carried out on certain dimension characteristics so as to delete the dimension characteristics with small saturation from the original portrait. For example, if the age is screened by probing, assuming that 100 persons in total, 70 persons have age data characteristic records, and 30 persons do not have age data characteristic records, the saturation of the age data is 70%, and the characteristic with small saturation is deleted as appropriate, because the loss is too serious to retain valid information.
The specific implementation mode of missing value filling is as follows, the missing value filling is performed in a classified manner based on a nonlinear interpolation method and a rpart method, and a missing value filling scheme is selected through result backtracking.
The specific implementation mode of data set isomerism and expansion is as follows, taking the readable medium-based text extraction for data set isomerism and expansion as an example, assuming that the disease history of 100 ten thousand people has diseases such as diabetes and hypertension in the list of characteristics, so much text information is accumulated in the list of characteristics and the model cannot directly obtain an effective text medium. Therefore, the extracted medium with diabetes in the list of characteristics is changed into a new list of characteristics, and the 100 ten thousands of people with diabetes are marked as 1 in the new list of characteristics, and the people without diabetes are marked as 0 in the list of characteristics; and so on for other disease species and other conditions.
In this embodiment, the specific implementation manners of the XGBOOST recommendation model and the deep neural network recommendation model are obtained by pre-training in step S30: firstly, an initial deep neural network recommendation model and an initial XGBOST recommendation model are obtained according to the characteristic data set training of the step S20, and the optimized deep neural network recommendation model and the optimized XGBOST recommendation model are obtained after the test data set is tested and optimized.
The output data of the XGB OST recommendation model and the deep neural network recommendation model are the medicines to be recommended and the corresponding probability values.
The optimization is to set up an XGB OST recommendation model and a deep neural network recommendation model after the preprocessing of the step S20 is completed, adjust parameters and hyper-parameters of the two models according to computing resources and time, set thresholds for the parameters and the hyper-parameters, limit upper and lower limits of the parameters and the hyper-parameters, and prevent the parameters from entering a dead cycle.
And (5) taking the feature data set of the step S20 as training data, and training to obtain an initial deep neural network recommendation model. The initial deep neural network recommendation model is a 5-layer neural network model which comprises a fully-connected layer, wherein the fully-connected layer comprises 256 neurons, 128 neurons and 64 neurons, and a Dropout layer is added to the fully-connected layer. The initial deep neural network recommendation model calculates the partial derivatives through a back propagation algorithm, and the calculation of the partial derivatives follows a chain rule; and continuous network training and parameter adjustment are carried out through a TensorFlow deep learning framework, and finally the optimized deep neural network recommendation model is obtained. And (5) training the feature data set in the step (S20) to obtain an initial XGBOOST recommendation model, further ranking the importance of feature factors through a feature opportunity in the initial XGBOOST recommendation model, screening out factors with lower weight, and finally obtaining the optimized XGBOOST recommendation model.
In this embodiment, the medicine request information of the user in step S40 may be obtained in multiple ways, and the text information or the code in the standard format is converted after voice input, the text information or the code in the standard format is converted after medical record scanning identification, or the text information or the code in the standard format is converted after key field information is manually input, and the converted text information or the code in the standard format is used as the medicine request information. The medicine request information input by the user through the system comprises age, symptoms, medical history and the like, and the medicine request information input by different users is not identical. For example, the medication request information of the first user includes an age of 3 years, a first symptom of a fever of 39 degrees for four days, a second symptom of a dry cough for two days, and a third symptom of a thick nasal discharge for two days. The second user's drug request information includes age 65 years, symptoms of dizziness, a medical history of ten years of hypertension, and ten years of taking a drug.
In this embodiment, in step S50, the XGBOOST recommendation model calculates different first recommendation strength probability values for different pieces of medicine request information. For example, the XGBOOST recommendation model calculates and obtains a first recommendation strength probability value of each drug in the drug library according to the drug request information of the user, where the first recommendation strength probability value is a first drug and 0.19; second drug, 0.55; third drug, 0.97; … … are provided.
In this embodiment, in step S60, the different medicine request information is calculated by the deep neural network recommendation model to obtain different second recommendation strength probability values. For example, the deep neural network recommendation model calculates and obtains a second recommendation strength probability value of each medicine in the medicine library according to the medicine request information of the user, wherein the second recommendation strength probability values are respectively a first medicine and 0.89; second drug, 0.15; third drug, 0.07; … … are provided.
In this embodiment, the first recommended strength probability value and the second recommended strength probability value in step S70 are linearly added according to a weight ratio, where the weight ratio is referred to through Linear Regression (Linear Regression), and finally the target recommended strength probability value of the combined model is obtained. It is assumed that the weight ratio of the XGBOOST recommendation model is a first weight ratio, the weight ratio of the deep neural network recommendation model is a second weight ratio, and the first weight ratio + the second weight ratio is 1.
The target recommended strength probability value of the first medicine is equal to a first recommended strength probability value of the first medicine multiplied by a first weight ratio + a second recommended strength probability value of the first medicine multiplied by a second weight ratio.
The target recommended strength probability value of the second medicine is equal to the first recommended strength probability value of the second medicine multiplied by the first weight ratio and the second recommended strength probability value of the second medicine multiplied by the second weight ratio.
And by analogy, calculating to obtain target recommended strength probability values of all the medicines in the medicine library.
In this embodiment, the specific screening manner of step S80 may be: and selecting and storing the medicines ranked in the first three after the probability values of the recommended target strength are arranged in a reverse order. The information stored in the medicine comprises the rank information of the medicine after screening and the name of the medicine after screening, and the probability value of the recommended strength of the target after screening can also be selectively stored, wherein the rank information of the medicine, the name of the medicine and the probability value of the recommended strength of the target are in one-to-one correspondence. For example, the drug recommendation results of the first user are respectively a first ranking-a third drug-0.97; second rank-seventh drug-0.92; third rank-fiftieth drug-0.89. The medicine recommendation results of the second user are respectively a first ranking-a first medicine-0.89; second rank-second drug-0.83; third rank-third medicine oral product-0.81.
According to the method, the XGB recommendation model and the deep neural network recommendation model are combined, model results of the XGB recommendation model and the deep neural network recommendation model are linearly added, the recommendation accuracy is improved by 200% compared with that of a non-learning model, and the correct medicine recommendation rate of the first three ranked medicines is improved from 75% to 95%. The invention extracts more hidden information from the historical medical data in the original portrait by characteristic expansion to ensure the integrity of characteristic values, and utilizes saturation to explore, screen and delete characteristic values without training significance, thereby forming a complete and effective characteristic data set. In addition, the XGB OST recommendation model and the deep neural network recommendation model are combined, so that the degree of automation is high, self-adaptive training and prediction can be realized, and the medicine recommendation efficiency is improved.
Referring to fig. 3, the present invention further provides an artificial intelligence based drug recommendation system, which includes:
an acquisition module 10 for acquiring historical medical data to construct an original representation;
a pre-processing module 20 for pre-processing the original image and outputting a feature data set;
the model training module 30 is configured to respectively pre-train the preprocessed feature data sets as training data to obtain an XGBOOST recommendation model and a deep neural network recommendation model;
the input module 40 is used for acquiring the medicine request information of the user and inputting the medicine request information into the XGB OST recommendation model and the deep neural network recommendation model;
the processing module 50 is configured to calculate a first recommended strength probability value of the target drug corresponding to the obtained drug request information through an XGBOOST recommendation model, and calculate a second recommended strength probability value of the target drug corresponding to the obtained drug request information through a deep neural network recommendation model; linearly adding the first recommended strength probability value and the second recommended strength probability value of each target medicine to obtain a target recommended strength probability value of each target medicine;
and the output module 60 is configured to sort the target medicine information of each target medicine from high to low according to the corresponding target recommendation strength probability value, screen out a preset number of target medicine information before ranking, and push the target medicine information to the user.
Referring to fig. 4, the present invention further provides a computer device 2, where the computer device 2 includes:
a memory 21 for storing executable program code; and
a processor 22 for calling said executable program code in said memory 21, the execution steps including the artificial intelligence based drug recommendation method described above.
One processor 22 is illustrated in fig. 4.
The memory 21, which is a non-volatile computer-readable storage medium, may be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as program instructions/modules corresponding to the artificial intelligence based drug recommendation method in the embodiments of the present invention. The processor 22 executes various functional applications and data processing of the computer device 2 by executing the non-volatile software programs, instructions and modules stored in the memory 21, i.e. implements the artificial intelligence based drug recommendation method in any of the above-described method embodiments.
The memory 21 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store historical medical data of the user at the computer device 2. Further, the memory 21 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, the memory 21 may optionally include a memory 21 remotely located from the processor 22, and these remote memories 21 may be connected to the artificial intelligence based drug recommendation system 1 via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The one or more modules are stored in the memory 21 and, when executed by the one or more processors 22, perform artificial intelligence based drug recommendation methods in any of the method embodiments described above, e.g., the programs of fig. 1-2 described above.
The product can execute the method provided by the embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.
The computer device 2 of the present embodiment exists in a variety of forms, including but not limited to:
(1) a mobile communication device: such devices are characterized by mobile communications capabilities and are primarily targeted at providing voice, data communications. Such terminals include: smart phones (e.g., iphones), multimedia phones, functional phones, and low-end phones, among others.
(2) Ultra mobile personal computer device: the equipment belongs to the category of personal computers, has calculation and processing functions and generally has the characteristic of mobile internet access. Such terminals include: PDA, MID, and UMPC devices, etc., such as ipads.
(3) A portable entertainment device: such devices can display and play multimedia content. This type of device comprises: audio, video players (e.g., ipods), handheld game consoles, electronic books, and smart toys and portable car navigation devices.
(4) A server: the device for providing the computing service comprises a processor, a hard disk, a memory, a system bus and the like, and the server is similar to a general computer architecture, but has higher requirements on processing capacity, stability, reliability, safety, expandability, manageability and the like because of the need of providing high-reliability service.
(5) And other electronic devices with data interaction functions.
Still another embodiment of the present application provides a non-transitory computer-readable storage medium storing computer-executable instructions for execution by one or more processors, such as the one processor 22 in fig. 4, to cause the one or more processors 22 to perform the artificial intelligence based drug recommendation method in any of the method embodiments described above, such as executing the programs of fig. 1-2 described above.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on at least two network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present application. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a general hardware platform, and certainly can also be implemented by hardware. It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware related to instructions of a computer program, which can be stored in a computer readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-only memory (ROM), a Random Access Memory (RAM), or the like.
The sequence numbers of the embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation method.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A medicine recommendation method based on artificial intelligence is characterized by comprising the following steps:
acquiring historical medical data to construct an original portrait;
preprocessing an original image and outputting a characteristic data set;
the preprocessed feature data set is used as training data and is respectively trained in advance to obtain an XGB OST recommendation model and a deep neural network recommendation model;
acquiring medicine request information of a user and inputting the medicine request information into an XGB OST recommendation model and a deep neural network recommendation model;
the XGB recommendation model calculates and obtains a first recommendation strength probability value of a target drug corresponding to the drug request information according to the drug request information;
the deep neural network recommendation model calculates and obtains a second recommendation strength probability value of the target medicine corresponding to the medicine request information according to the medicine request information;
linearly adding the first recommended strength probability value and the second recommended strength probability value of each target medicine to obtain a target recommended strength probability value of each target medicine;
and sorting the target medicine information of each target medicine from high to low according to the corresponding target recommendation strength probability value, screening out the preset number of target medicine information before ranking, and pushing the target medicine information to the user.
2. The artificial intelligence based drug recommendation method of claim 1, wherein: the original image comprises basic information and medical record information, the data source of the original image comprises a plurality of different data sources, and the multidimensional characteristics of the plurality of different data sources are inconsistent.
3. The artificial intelligence based drug recommendation method of claim 1, wherein: and establishing a table corresponding relation between data in the historical medical data and data in an association database by an association extraction method while acquiring the historical medical data, and acquiring the association data in the association database to construct an original portrait.
4. The artificial intelligence based drug recommendation method of claim 1, wherein: the preprocessing comprises feature expansion, saturation exploration screening, missing value filling, data set isomerism and expansion, feedback selection and correlation screening.
5. The artificial intelligence based drug recommendation method of claim 1, wherein: the original representation includes features of an n × m dimensional matrix, and the pre-processing the original representation includes:
simplifying the characteristics of the n x m dimensional matrix into the transverse splicing of a plurality of n x 1 dimensional vectors by using a tsfresh algorithm, and then performing particle scanning, wherein the particle scanning comprises single-dimensional particle scanning and multi-dimensional particle scanning.
6. The artificial intelligence based drug recommendation method of claim 5, wherein: the pre-processing of the original portrait comprises: the single-dimensional particle scanning is performed with sliding scanning to derive a single-dimensional extended feature set L1, the multi-dimensional particle scanning is performed with sliding scanning to derive a multi-dimensional extended feature set L2, and then the single-dimensional extended feature set, the multi-dimensional extended feature set and the original image are spliced to obtain a feature data set containing more information.
7. The artificial intelligence based drug recommendation method of claim 1, wherein: the classification of non-standard drug names when collecting historical medical data includes the following steps,
respectively carrying out word segmentation processing on the non-standard medicine name and the standard medicine name;
based on a Word2Vec shallow neural network and based on a skip _ gram algorithm, respectively converting Word vectors of words in the non-standard medicine name and the standard medicine name;
longitudinally splicing the vectors by using a pooling rule to obtain word vector splicing of the non-standard medicine name and the standard medicine name;
calculating the vector distance between the non-standard medicine name and the standard medicine name;
finding out the standard medicine name which is closest to the vector distance of the non-standard medicine name;
and classifying the non-standard medicine names according to the classification standard of the standard medicine names with the closest vector distance.
8. A medicine recommendation system based on artificial intelligence is characterized by comprising:
a collection module for collecting historical medical data to construct an original representation;
the preprocessing module is used for preprocessing the original image and outputting a characteristic data set;
the model training module is used for respectively training the preprocessed feature data set as training data in advance to obtain an XGB OST recommendation model and a deep neural network recommendation model;
the input module is used for acquiring the drug request information of the user and inputting the drug request information into the XGB OST recommendation model and the deep neural network recommendation model;
the processing module is used for calculating and obtaining a first recommended strength probability value of a target medicine corresponding to the medicine request information through an XGBOOST recommendation model and calculating and obtaining a second recommended strength probability value of the target medicine corresponding to the medicine request information through a deep neural network recommendation model; linearly adding the first recommended strength probability value and the second recommended strength probability value of each target medicine to obtain a target recommended strength probability value of each target medicine;
and the output module is used for sorting the target medicine information of each target medicine from high to low according to the corresponding target recommendation strength probability value, screening out the preset number of target medicine information before ranking, and pushing the target medicine information to the user.
9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that: the processor, when executing the computer program, implements the artificial intelligence based drug recommendation method of any of claims 1 to 7.
10. A computer-readable storage medium having stored thereon a computer program, characterized in that: the computer program when executed by a processor implements the artificial intelligence based drug recommendation method of any of claims 1 to 7.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010090681.0A CN111260448B (en) | 2020-02-13 | 2020-02-13 | Medicine recommendation method based on artificial intelligence and related equipment |
| PCT/CN2020/093297 WO2021159640A1 (en) | 2020-02-13 | 2020-05-29 | Drug recommendation method based on artificial intelligence, and related device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010090681.0A CN111260448B (en) | 2020-02-13 | 2020-02-13 | Medicine recommendation method based on artificial intelligence and related equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111260448A true CN111260448A (en) | 2020-06-09 |
| CN111260448B CN111260448B (en) | 2023-10-20 |
Family
ID=70954445
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010090681.0A Active CN111260448B (en) | 2020-02-13 | 2020-02-13 | Medicine recommendation method based on artificial intelligence and related equipment |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN111260448B (en) |
| WO (1) | WO2021159640A1 (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111914164A (en) * | 2020-06-20 | 2020-11-10 | 武汉海云健康科技股份有限公司 | Medication prediction method and system based on medical big data |
| CN111949888A (en) * | 2020-09-02 | 2020-11-17 | 上海优扬新媒信息技术有限公司 | Data recommendation method and device |
| CN112035757A (en) * | 2020-08-28 | 2020-12-04 | 康键信息技术（深圳）有限公司 | Medical waterfall flow pushing method, device, equipment and storage medium |
| CN112131788A (en) * | 2020-09-18 | 2020-12-25 | 江西兰叶科技有限公司 | Motor design method and system for teaching |
| CN112562854A (en) * | 2020-12-17 | 2021-03-26 | 山东大学 | Accurate medical care service recommendation method and system for elderly people |
| CN113344415A (en) * | 2021-06-23 | 2021-09-03 | 中国平安财产保险股份有限公司 | Deep neural network-based service distribution method, device, equipment and medium |
| CN113393295A (en) * | 2021-06-15 | 2021-09-14 | 北方健康医疗大数据科技有限公司 | Service data pushing method and device, electronic equipment and storage medium |
| CN113657970A (en) * | 2021-08-30 | 2021-11-16 | 平安医疗健康管理股份有限公司 | Artificial intelligence based medicine recommendation method, device, equipment and storage medium |
| CN113722371A (en) * | 2021-08-31 | 2021-11-30 | 平安国际智慧城市科技股份有限公司 | Medicine recommendation method, device, equipment and storage medium based on decision tree |
| CN113764067A (en) * | 2021-09-08 | 2021-12-07 | 平安国际智慧城市科技股份有限公司 | Medicine recommendation method, system, equipment and storage medium |
| WO2022105003A1 (en) * | 2020-11-23 | 2022-05-27 | 深圳市鹰硕教育服务有限公司 | Medical information processing method and apparatus, and electronic device |
| CN114722976A (en) * | 2022-06-09 | 2022-07-08 | 青岛美迪康数字工程有限公司 | Medicine recommendation system and construction method |
| US20230091973A1 (en) * | 2020-03-09 | 2023-03-23 | Amar Chowdry | Medzone. |
| CN116504354A (en) * | 2023-06-28 | 2023-07-28 | 合肥工业大学 | Intelligent service recommendation method and system based on intelligent medical treatment |
| CN117252664A (en) * | 2023-11-10 | 2023-12-19 | 浙江口碑网络技术有限公司 | Medicine recommendation reason generation method, device, medium and equipment |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113988949A (en) * | 2021-11-15 | 2022-01-28 | 北京有竹居网络技术有限公司 | Promotion information processing method, device, equipment, medium and program product |
| CN117198513B (en) * | 2023-11-07 | 2024-03-29 | 北京烔凡科技有限公司 | Health guidance method and system for hypertension patient |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107423442A (en) * | 2017-08-07 | 2017-12-01 | 火烈鸟网络(广州)股份有限公司 | Method and system, storage medium and computer equipment are recommended in application based on user's portrait behavioural analysis |
| CN109087691A (en) * | 2018-08-02 | 2018-12-25 | 科大智能机器人技术有限公司 | A kind of OTC drugs recommender system and recommended method based on deep learning |
| CN109871464A (en) * | 2019-01-17 | 2019-06-11 | 东南大学 | A kind of video recommendation method and device based on UCL Semantic Indexing |
| CN109933729A (en) * | 2019-03-28 | 2019-06-25 | 广州麦迪森在线医疗科技有限公司 | A kind of academic information recommended method of medical treatment based on user preference and system |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6733886B2 (en) * | 2016-07-12 | 2020-08-05 | グロースメディスン株式会社 | Recommended drug presentation system for osteoporosis |
| CN107945847B (en) * | 2017-12-12 | 2021-05-28 | 科大智能机器人技术有限公司 | Recommendation system and method for non-prescription drugs |
| CN109102855A (en) * | 2018-07-03 | 2018-12-28 | 北京康夫子科技有限公司 | Drug recommended method |
| CN109711887B (en) * | 2018-12-28 | 2021-02-09 | 拉扎斯网络科技（上海）有限公司 | Generation method and device of mall recommendation list, electronic equipment and computer medium |
| CN110085292B (en) * | 2019-04-28 | 2022-07-26 | 广东技术师范大学 | Medicine recommendation method and device and computer-readable storage medium |
-
2020
- 2020-02-13 CN CN202010090681.0A patent/CN111260448B/en active Active
- 2020-05-29 WO PCT/CN2020/093297 patent/WO2021159640A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107423442A (en) * | 2017-08-07 | 2017-12-01 | 火烈鸟网络(广州)股份有限公司 | Method and system, storage medium and computer equipment are recommended in application based on user's portrait behavioural analysis |
| CN109087691A (en) * | 2018-08-02 | 2018-12-25 | 科大智能机器人技术有限公司 | A kind of OTC drugs recommender system and recommended method based on deep learning |
| CN109871464A (en) * | 2019-01-17 | 2019-06-11 | 东南大学 | A kind of video recommendation method and device based on UCL Semantic Indexing |
| CN109933729A (en) * | 2019-03-28 | 2019-06-25 | 广州麦迪森在线医疗科技有限公司 | A kind of academic information recommended method of medical treatment based on user preference and system |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230091973A1 (en) * | 2020-03-09 | 2023-03-23 | Amar Chowdry | Medzone. |
| CN111914164A (en) * | 2020-06-20 | 2020-11-10 | 武汉海云健康科技股份有限公司 | Medication prediction method and system based on medical big data |
| CN111914164B (en) * | 2020-06-20 | 2024-04-26 | 武汉海云健康科技股份有限公司 | Medication prediction method and system based on medical big data |
| CN112035757A (en) * | 2020-08-28 | 2020-12-04 | 康键信息技术（深圳）有限公司 | Medical waterfall flow pushing method, device, equipment and storage medium |
| CN111949888A (en) * | 2020-09-02 | 2020-11-17 | 上海优扬新媒信息技术有限公司 | Data recommendation method and device |
| CN111949888B (en) * | 2020-09-02 | 2023-10-10 | 度小满科技(北京)有限公司 | Data recommendation method and device |
| CN112131788A (en) * | 2020-09-18 | 2020-12-25 | 江西兰叶科技有限公司 | Motor design method and system for teaching |
| WO2022105003A1 (en) * | 2020-11-23 | 2022-05-27 | 深圳市鹰硕教育服务有限公司 | Medical information processing method and apparatus, and electronic device |
| CN112562854A (en) * | 2020-12-17 | 2021-03-26 | 山东大学 | Accurate medical care service recommendation method and system for elderly people |
| CN113393295A (en) * | 2021-06-15 | 2021-09-14 | 北方健康医疗大数据科技有限公司 | Service data pushing method and device, electronic equipment and storage medium |
| CN113344415A (en) * | 2021-06-23 | 2021-09-03 | 中国平安财产保险股份有限公司 | Deep neural network-based service distribution method, device, equipment and medium |
| CN113657970A (en) * | 2021-08-30 | 2021-11-16 | 平安医疗健康管理股份有限公司 | Artificial intelligence based medicine recommendation method, device, equipment and storage medium |
| CN113722371A (en) * | 2021-08-31 | 2021-11-30 | 平安国际智慧城市科技股份有限公司 | Medicine recommendation method, device, equipment and storage medium based on decision tree |
| CN113722371B (en) * | 2021-08-31 | 2024-04-12 | 深圳平安智慧医健科技有限公司 | Medicine recommendation method, device, equipment and storage medium based on decision tree |
| CN113764067A (en) * | 2021-09-08 | 2021-12-07 | 平安国际智慧城市科技股份有限公司 | Medicine recommendation method, system, equipment and storage medium |
| CN113764067B (en) * | 2021-09-08 | 2023-09-08 | 深圳平安智慧医健科技有限公司 | Drug recommendation method, system, device and storage medium |
| CN114722976A (en) * | 2022-06-09 | 2022-07-08 | 青岛美迪康数字工程有限公司 | Medicine recommendation system and construction method |
| CN116504354A (en) * | 2023-06-28 | 2023-07-28 | 合肥工业大学 | Intelligent service recommendation method and system based on intelligent medical treatment |
| CN117252664A (en) * | 2023-11-10 | 2023-12-19 | 浙江口碑网络技术有限公司 | Medicine recommendation reason generation method, device, medium and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2021159640A1 (en) | 2021-08-19 |
| CN111260448B (en) | 2023-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111260448A (en) | Artificial intelligence-based medicine recommendation method and related equipment | |
| CN108766512B (en) | Health data management method and device, computer equipment and storage medium | |
| CN108109700B (en) | Method and device for evaluating curative effect of chronic disease | |
| CN107016438B (en) | System based on traditional Chinese medicine syndrome differentiation artificial neural network algorithm model | |
| CN111191020A (en) | Prescription recommendation method and system based on machine learning and knowledge graph | |
| CN111341456B (en) | Method and device for generating diabetic foot knowledge graph and readable storage medium | |
| CN110289068A (en) | Drug recommended method and equipment | |
| CN110504031B (en) | Cloud management database establishment method and system for health behavior intervention | |
| CN111048167A (en) | Hierarchical case structuring method and system | |
| CN111785366B (en) | Patient treatment scheme determination method and device and computer equipment | |
| CN109935337B (en) | Medical record searching method and system based on similarity measurement | |
| WO2024008043A1 (en) | Automated clinical data generation method and system based on causal relationship mining | |
| Bianchini et al. | Deep learning in science | |
| CN114765075A (en) | Medicine recommendation method, device and system, electronic equipment and storage medium | |
| CN115050442B (en) | Disease category data reporting method and device based on mining clustering algorithm and storage medium | |
| Hu et al. | Variable selection with missing data in both covariates and outcomes: Imputation and machine learning | |
| US10832822B1 (en) | Methods and systems for locating therapeutic remedies | |
| CN116910172A (en) | Follow-up table generation method and system based on artificial intelligence | |
| KR20110098286A (en) | Self health diagnosis system of oriental medicine using fuzzy inference method | |
| Niyirora et al. | Network analysis of medical care services | |
| US20130253892A1 (en) | Creating synthetic events using genetic surprisal data representing a genetic sequence of an organism with an addition of context | |
| CN110729052A (en) | Old health data knowledge analysis method and system | |
| CN111816276B (en) | Method and device for recommending education courses, computer equipment and storage medium | |
| CN110033862B (en) | Traditional Chinese medicine quantitative diagnosis system based on weighted directed graph and storage medium | |
| TWI780608B (en) | Drug recommendation system and method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 40020788 Country of ref document: HK |
|
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |