CN117352178A - Big data-based drug risk assessment system and method - Google Patents
Big data-based drug risk assessment system and method Download PDFInfo
- Publication number
- CN117352178A CN117352178A CN202311492271.9A CN202311492271A CN117352178A CN 117352178 A CN117352178 A CN 117352178A CN 202311492271 A CN202311492271 A CN 202311492271A CN 117352178 A CN117352178 A CN 117352178A
- Authority
- CN
- China
- Prior art keywords
- physiological index
- patient
- data
- historical
- drug
- 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.)
- Pending
Links
- 239000003814 drug Substances 0.000 title claims abstract description 295
- 229940079593 drug Drugs 0.000 title claims abstract description 218
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000012502 risk assessment Methods 0.000 title claims abstract description 45
- 238000012549 training Methods 0.000 claims abstract description 87
- 230000002159 abnormal effect Effects 0.000 claims abstract description 62
- 230000006870 function Effects 0.000 claims description 15
- 230000015654 memory Effects 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000012216 screening Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 5
- 230000005856 abnormality Effects 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 abstract description 2
- 238000005065 mining Methods 0.000 description 10
- 238000013138 pruning Methods 0.000 description 6
- 230000009286 beneficial effect Effects 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 201000010099 disease Diseases 0.000 description 5
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 5
- 206010067484 Adverse reaction Diseases 0.000 description 3
- 230000006838 adverse reaction Effects 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 230000003908 liver function Effects 0.000 description 3
- 241001465754 Metazoa Species 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000003907 kidney function Effects 0.000 description 2
- 230000004199 lung function Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 208000030453 Drug-Related Side Effects and Adverse reaction Diseases 0.000 description 1
- 206010070863 Toxicity to various agents Diseases 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 229940121657 clinical drug Drugs 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000009509 drug development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 229940043263 traditional drug Drugs 0.000 description 1
- 238000011282 treatment Methods 0.000 description 1
- 238000011269 treatment regimen Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/10—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- 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
- G16H70/00—ICT specially adapted for the handling or processing of medical references
- G16H70/40—ICT specially adapted for the handling or processing of medical references relating to drugs, e.g. their side effects or intended usage
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Chemical & Material Sciences (AREA)
- Molecular Biology (AREA)
- Medicinal Chemistry (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Pathology (AREA)
- Databases & Information Systems (AREA)
- Toxicology (AREA)
- Pharmacology & Pharmacy (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The application provides a big data-based drug risk assessment system and a big data-based drug risk assessment method, which relate to the technical field of data analysis, by collecting historical illness state data, historical drug dosage and physiological index data of a patient, training a dosage prediction model for predicting drug dosage Y by utilizing the historical illness state data of the patient, training a physiological index prediction model for predicting future physiological index data Q by utilizing the historical drug dosage and physiological index data of the patient, predicting the future physiological index data Q of the patient by utilizing the physiological index prediction model, predicting the drug dosage Y of the patient by utilizing the dosage prediction model, comparing whether the future physiological index data Q of the patient is larger than a physiological index threshold Q0, if so, setting the patient as an abnormal patient, calculating a drug risk coefficient R of the abnormal patient by utilizing the future physiological index data of the abnormal patient, and sending drug risk early warning to doctors, thereby improving the accuracy and pertinence of drug risk assessment.
Description
Technical Field
The application relates to the technical field of data analysis, in particular to a drug risk assessment system and method based on big data.
Background
The drug risk assessment mainly aims at adverse reactions caused by normal doses of drugs in human bodies, and the drug adverse reactions not only cause burdens on clinic and economy, but also are one of main reasons for causing difficulty in drug development, and bring a plurality of problems to the development of medical industry. The current assessment of drug risk is mainly in three ways: preclinical screening, clinical trials, and post-market supervision. Preclinical screening requires a lot of time and money to study animal experiments, but the animal response to drug toxicity is not equivalent to the human clinical drug response; in the aspect of clinical test, the number of people using the medicine is limited, and the time and financial resources are limited, so that a comprehensive medicine risk assessment result cannot be obtained; in post-market regulatory aspects, risk assessment is limited because of the lengthy period of time and risk of drug use by patients, and once uncontrolled drug risk occurs, the impact on patients is not predictable.
According to the method, reasonable medication records are extracted from an electronic medical record database, according to physiological characteristics of a patient, disease diagnosis results and corresponding reasonable medication doses of the patient, a deep learning model is utilized to model the relation of the medication doses of the medicines with respect to the physiological characteristics of the patient and the disease diagnosis results, the medication dose risk probability of various medicines under different individual conditions is estimated, a medication dose risk assessment model is established, the corresponding physiological characteristics of the patient and the disease diagnosis results are extracted for a new electronic medical record, the medicine dose risk assessment model is used for predicting the reasonable medication doses, the predicted reasonable medication doses are compared with the medication doses set in the electronic medical record, and a medication dose warning method provides decision basis for clinical reasonable medication practice if the medication dose risk coefficients are larger than a set threshold value, but is generally only suitable for internal medication management of hospitals, and personalized and comprehensive patient health management and long-term patient condition monitoring cannot be provided.
Therefore, the risk assessment of the medicine by using the computer technology becomes a necessary trend of the future medical community development, and the medicine adverse reaction assessment by combining the data-driven method is a practical assessment method, so that great social benefits can be generated for society, and the method is economically feasible.
The traditional drug risk assessment method generally depends on limited clinical data and experimental data, and data of medical institutions and pharmaceutical companies are often not shared independently, so that the data of drug risk assessment lacks accuracy, and risks generated by the same drug on different patients are different due to different physiological index data and different historical conditions of the patients, so that potential risks cannot be captured, and the drug risk assessment lacks accuracy and pertinence.
Disclosure of Invention
The present application aims to solve, at least to some extent, one of the technical problems in the related art. Therefore, an object of the present application is to provide a drug risk assessment system and method based on big data, which improves the accuracy and pertinence of drug risk assessment.
One aspect of the present application provides a big data based drug risk assessment method, comprising:
Step S100: collecting historical illness state data, historical medicine dosage and physiological index data of a patient;
the collection modes of the historical illness state data, the historical medicine dosage and the physiological index data are as follows: setting a data acquisition period T, and acquiring the historical illness state data, the historical medicine dosage and the physiological index data of a patient from a public medical database once every the time of the data acquisition period T, wherein T times are acquired in total;
the historical illness state data of the patient refer to illness state expression data of the patient in the historical time recorded in a public medical database, and the historical illness state data of t times is collected to be A1, A2, at;
the historical drug dose is drug dose data used under the corresponding historical illness state data of a patient in the historical time, and the historical drug doses for t times are collected to be B1, B2, >
the physiological index data of the patient refer to functional index data of each organ of the body of the patient, and the physiological index data collected for t times are C1, C2, ct;
step S200: training a dose prediction model for predicting a drug dose Y using historical patient condition data of the patient;
the training process of the dose prediction model for predicting the drug dose Y by using the historical illness state data of the patient is as follows: collecting T times according to a data acquisition period T to obtain the historical illness state data as A1, A2, and At, obtaining the historical medicine doses as B1, B2, and Bt, wherein the historical illness state data and the corresponding historical medicine doses are used as a group of training samples, T training samples are respectively (A1, B1), (A2, B2), and (At, bt), i training samples are (Ai, bi), the training samples of the dose prediction model use the historical illness state data as input data, the historical medicine doses required by the corresponding historical illness state data are used as output data, the dose prediction model is trained, bi represents the actual historical medicine doses in the i training samples, Representing the predicted historical drug dose of the ith training sample, taking the accurately predicted historical drug dose as a prediction target of a dose prediction model, taking a loss function between the minimized actual historical drug dose and the predicted historical drug dose as a training target of the training dose prediction model, and obtaining a dose prediction model for predicting the drug dose Y;
the calculation formula of the loss function is as follows:
preferably, the dose prediction model is a regression model;
step S300: training a physiological index prediction model for predicting future physiological index data Q by using the historical drug dosage and the physiological index data of the patient;
the training process of the physiological index prediction model for predicting the future physiological index data Q is as follows: obtaining a historical medicament dose sequence { B1, B2, & gt, bt } from the historical medicament dose of the patient, obtaining a physiological index sequence { C1, C2, & gt, ct } from the physiological index data of the patient, setting a prediction time step, a sliding window length and a sliding step, dividing the historical medicament dose sequence and the physiological index sequence into t historical medicament dose sequence samples and t physiological index sequence samples by utilizing a sliding window method, and training a physiological index prediction model;
The historical medicine dosage sequence sample and the physiological index sequence sample are used as training samples for training a physiological index prediction model to be input, physiological index data in a prediction time step is used as output data of the training physiological index prediction model, cj represents actual physiological index data in a j-th training sample,representing predicted physiological index data in the jth training sample, taking the accurate predicted physiological index data as a prediction target of a physiological index prediction model to minimize the actual physiological index data and a loss function of the predicted physiological index data ≡>As a training target of a physiological index prediction model;
the formula of the loss function of the physiological index prediction model is as follows:
step S400: predicting future physiological index data Q of the patient by using a physiological index prediction model, and predicting the drug dosage Y of the patient by using a dosage prediction model;
the predicting the future physiological index data Q of the patient by using the physiological index prediction model refers to collecting the drug dosage data and the physiological index data of the patient, and predicting the future physiological index data of the patient by using the trained physiological index prediction model;
the step of predicting the drug dosage Y of the patient by using the dosage prediction model is to collect the disease data of the patient and predict the drug dosage of the patient by using the trained dosage prediction model;
The future physiological index data is obtained based on the predicted drug dose, and reflects the influence degree of the predicted drug dose on the body of the patient;
step S500: comparing whether future physiological index data Q of the patient is larger than a physiological index threshold Q0, and if so, setting the patient as an abnormal patient;
the physiological index threshold Q0 is set according to the experience of medical specialists and the actual situation of the patient, when the future physiological index data of the patient is smaller than or equal to the physiological index threshold Q0, the future physiological index data of the patient is normal, and if the future physiological index data of the patient is normal, the future drug dosage does not need to be adjusted; when the future physiological index data of the patient is larger than the physiological index threshold value Q0, the future physiological index data of the patient is abnormal, and the patient is an abnormal patient and needs to adjust the future drug dosage;
step S600: calculating a drug risk coefficient R of the abnormal patient by using future physiological index data of the abnormal patient, and sending a drug risk early warning to a doctor;
the specific method for calculating the drug risk coefficient R of the abnormal patient by using the future physiological index data of the abnormal patient comprises the following steps:
step S610: acquiring an association rule set;
The specific method for acquiring the association rule set comprises the following steps:
step S611: acquiring a data set of medicine, medicine dosage and future physiological index data used by an abnormal patient;
the acquisition mode of the data set is as follows:
counting the medicine doses of all kinds of medicines used by G abnormal patients, and dividing the medicine doses into different dose levels; carrying out data coding on different medicines and future physiological index data;
the future physiological index data are future physiological index data of an abnormal patient, the future physiological index data of the abnormal patient are used for association rule mining, and the future physiological index data used for association rule mining are larger than a physiological index threshold Q0;
step S612: generating a frequent item set by using an association rule mining algorithm based on the data set;
the process of generating the frequent item set is as follows:
setting a support degree threshold W, wherein the support degree threshold is set according to actual conditions and expert suggestions;
setting the initial value of the variable z as 1, scanning the data set, calculating the occurrence frequency of future physiological index data, medicines and medicine doses of all patients, and generating a candidate item set z;
scanning a candidate item set z, counting the occurrence frequency of each item to obtain a support degree, screening the items with the support degree larger than or equal to a support degree threshold W to obtain a frequent item set z, connecting and pruning the frequent item set z, updating the value of j to z+1, generating a candidate item set z+1, scanning the candidate item set z+1, and counting the candidate items with each support degree larger than the support degree threshold W to obtain a frequent item set z+1;
Repeating connection and pruning operations on the frequent item sets to generate candidate item sets, and counting items with the support degree greater than a support degree threshold value to obtain the frequent item sets until the generated candidate item sets are empty;
step S613: generating a set of association rules based on the set of frequent items;
the association rules in the association rule set are association rules with support degree more than or equal to a support degree threshold W, and the association rule set consists of association rules in frequent item sets;
step S620: according to the association rule set, obtaining association rule weight Sx of an association rule item matched with future physiological index data;
the method for obtaining the association rule weight of the association rule item matched with the future physiological index data according to the association rule set comprises the following steps: obtaining association rule items between future physiological index data and medicines and medicine doses from an association rule set, and obtaining the support degree of the association rule items according to the association rule set, wherein the support degree is the association rule weight of the association rule items matched with the future physiological index data, and the association rule items represent medicines and medicine doses influencing the future physiological index data;
the purpose of obtaining the association rule item and the association rule weight thereof is to obtain the influence degree of the drug and the drug dosage on future physiological index data, when the support degree of a certain dosage level of a certain drug and the future physiological index data is greater than a support degree threshold value, the influence degree of the association rule item is the support degree, and the future physiological index data is associated with the drug and the dosage level;
The association rule weight Sx refers to the association rule weight of the x-th association rule item;
step S630: calculating a drug risk coefficient P of the abnormal patient according to the association rule weight;
the calculation method of the drug risk coefficient P comprises the following steps:wherein x refers to an x-th association rule item, sx refers to association rule weight of the x-th association rule item, and Hx refers to condition satisfaction of the x-th association rule item;
the condition satisfaction degree represents the matching degree between future physiological index data of the patient and the association rule item;
another aspect of the present application provides a big data based medication risk assessment system comprising:
the data collection module is used for collecting historical illness state data, historical medicine dosage and physiological index data of a patient;
a dose prediction model training module for training a dose prediction model for predicting a drug dose Y using historical patient condition data of the patient;
the physiological index prediction model training module is used for training a physiological index prediction model for predicting future physiological index data Q by using the historical drug dosage and the physiological index data of the patient;
the model prediction module is used for predicting future physiological index data Q of the patient by using the physiological index prediction model and predicting the drug dosage Y of the patient by using the dosage prediction model;
The abnormality screening module is used for comparing whether future physiological index data Q of the patient is larger than a physiological index threshold Q0 or not, and if so, setting the patient as an abnormal patient;
the risk coefficient calculation module is used for calculating a drug risk coefficient R of the abnormal patient by using future physiological index data of the abnormal patient and sending a drug risk early warning to a doctor.
One aspect of the present application provides an electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, which when executed by the processor, performs steps in a big data based medication risk assessment method.
One aspect of the present application provides a computer readable storage medium storing a computer program adapted to be loaded by a processor for performing steps in a big data based medication risk assessment method.
One or more technical solutions provided in the present application have at least the following technical effects or advantages:
according to the large-data-based drug risk assessment system and method, by collecting historical illness state data, historical drug doses and physiological index data of a patient, training a dose prediction model for predicting drug doses Y by utilizing the historical illness state data of the patient, training a physiological index prediction model for predicting future physiological index data Q by utilizing the historical drug doses and the physiological index data of the patient, predicting the future physiological index data Q of the patient by utilizing the physiological index prediction model, predicting the drug doses Y of the patient by utilizing the dose prediction model, comparing whether the future physiological index data Q of the patient is larger than a physiological index threshold Q0, if so, setting the patient as an abnormal patient, calculating a drug risk coefficient R of the abnormal patient by utilizing the future physiological index data of the abnormal patient, sending a drug risk early warning to a doctor, predicting the drug doses and the future physiological index data needed to be used by the patient by utilizing a machine learning model, accurately assessing the drug risk by utilizing a data driving method according to specific data of each patient, timely improving the accuracy and pertinence of drug risk assessment by combining a machine learning model and an association rule algorithm, and reducing the drug risk treatment strategy.
In order to more clearly illustrate the method of implementation of the invention or the technical solutions in the prior art, the following description will briefly introduce the drawings used in the examples or the description of the prior art, and it will be obvious to those skilled in the art that other drawings can be obtained from these drawings without inventive effort.
Drawings
FIG. 1 is a flow chart of a big data based drug risk assessment method provided by the present application;
FIG. 2 is a functional block diagram of a big data based drug risk assessment system provided herein;
fig. 3 is a schematic structural diagram of an electronic device provided in the present application;
fig. 4 is a schematic structural diagram of a computer readable storage medium provided in the present application.
Detailed Description
For a better understanding of the present application, various aspects of the present application will be described in more detail with reference to the accompanying drawings. It should be understood that these detailed description are merely illustrative of exemplary embodiments of the application and are not intended to limit the scope of the application in any way. The expression "and/or" includes any and all combinations of one or more of the associated listed items.
In the drawings, the size, dimensions and shape of elements have been slightly adjusted for convenience of description. The figures are merely examples and are not drawn to scale. As used herein, the terms "about," "approximately," and the like are used as terms of a table approximation, not as terms of a table degree, and are intended to account for inherent deviations in measured or calculated values that will be recognized by one of ordinary skill in the art. In addition, in this application, the order in which the processes of the steps are described does not necessarily indicate the order in which the processes occur in actual practice, unless explicitly defined otherwise or the context may be inferred.
It will be further understood that terms such as "comprises," "comprising," "includes," "including," "having," "containing," "includes" and/or "including" are open-ended, rather than closed-ended, terms that specify the presence of the stated features, elements, and/or components, but do not preclude the presence or addition of one or more other features, elements, components, and/or groups thereof. Furthermore, when a statement such as "at least one of the following" appears after a list of features listed, it modifies the entire list of features rather than just modifying the individual elements in the list. Furthermore, when describing embodiments of the present application, use of "may" means "one or more embodiments of the present application. Also, the term "exemplary" is intended to refer to an example or illustration.
Unless otherwise defined, all terms (including engineering and technical terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
In addition, embodiments and features of embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
Example 1
In this embodiment, the present application provides a big data-based drug risk assessment method, as shown in fig. 1, including:
step S100: collecting historical illness state data, historical medicine dosage and physiological index data of a patient;
the collection modes of the historical illness state data, the historical medicine dosage and the physiological index data are as follows: setting a data acquisition period T, and acquiring the historical illness state data, the historical medicine dosage and the physiological index data of a patient from a public medical database once every the time of the data acquisition period T, wherein T times are acquired in total;
the historical illness state data of the patient refer to illness state expression data of the patient in the historical time recorded in a public medical database, and the historical illness state data of t times is collected to be A1, A2, at;
the historical drug dose is drug dose data used under the corresponding historical illness state data of a patient in the historical time, and the historical drug doses for t times are collected to be B1, B2, >
The physiological index data of the patient refer to functional index data of each organ of the body of the patient, and the physiological index data collected for t times are C1, C2, ct;
illustratively, the physiological index data includes, but is not limited to: liver function index, kidney function index and lung function index;
this step describes how to collect patient historical condition data, historical medication doses, physiological index data for use in model training and association rule mining in subsequent steps, achieving the beneficial effect of providing data support for subsequent medication risk assessment;
step S200: training a dose prediction model for predicting a drug dose Y using historical patient condition data of the patient;
the training process of the dose prediction model for predicting the drug dose Y by using the historical illness state data of the patient is as follows: collecting T times according to a data acquisition period T to obtain the historical illness state data as A1, A2,at, obtaining the historical medicine doses as B1, B2, & gt, taking the historical illness state data and the corresponding historical medicine doses thereof as a group of training samples, taking t training samples as (A1, B1), (A2, B2) respectively, (At, bt) and (Ai, bi) as the i-th training sample, taking the historical illness state data as input data by the training samples of the dose prediction model, taking the required historical medicine doses corresponding to the historical illness state data as output data, training the dose prediction model, wherein Bi represents the actual historical medicine doses in the i-th training sample, Representing the predicted historical drug dose of the ith training sample, taking the accurately predicted historical drug dose as a prediction target of a dose prediction model, taking a loss function between the minimized actual historical drug dose and the predicted historical drug dose as a training target of the training dose prediction model, and obtaining a dose prediction model for predicting the drug dose Y;
the calculation formula of the loss function is as follows:
preferably, the dose prediction model is a regression model;
the step describes how to train a dose prediction model for predicting the drug dose Y by using the historical illness state data of the patient, so as to obtain a trained dose prediction model, and achieve the beneficial effects of providing a model foundation for the subsequent acquisition of the physiological index data of the patient and the mining of association rules under different drug doses;
step S300: training a physiological index prediction model for predicting future physiological index data Q by using the historical drug dosage and the physiological index data of the patient;
the training process of the physiological index prediction model for predicting the future physiological index data Q is as follows: obtaining a historical medicament dose sequence { B1, B2, & gt, bt } from the historical medicament dose of the patient, obtaining a physiological index sequence { C1, C2, & gt, ct } from the physiological index data of the patient, setting a prediction time step, a sliding window length and a sliding step, dividing the historical medicament dose sequence and the physiological index sequence into t historical medicament dose sequence samples and t physiological index sequence samples by utilizing a sliding window method, and training a physiological index prediction model;
The historical medicine dosage sequence sample and the physiological index sequence sample are used as training samples for training a physiological index prediction model to be input, physiological index data in a prediction time step is used as output data of the training physiological index prediction model, cj represents actual physiological index data in a j-th training sample,representing predicted physiological index data in the jth training sample, taking the accurate predicted physiological index data as a prediction target of a physiological index prediction model to minimize the actual physiological index data and a loss function of the predicted physiological index data ≡>As a training target of a physiological index prediction model;
the formula of the loss function of the physiological index prediction model is as follows:
preferably, the physiological index prediction model adopts a cyclic neural network model RNN or a long-short-time memory network LSTM;
the meaning of the predicted time step, the sliding window length and the sliding step is as follows: assuming that a historical medicine dose sequence { B1, B2, & gt, bt } and a physiological index sequence { C1, C2, & gt, ct } are provided, wherein the two sequences consist of t historical medicine doses and t physiological index data obtained for t times, the sequences are formed according to the obtained time sequence, the predicted time step length refers to the predicted time length of predicting future data according to the historical data, the sliding window length refers to the window defined in the sequence, the size of the window is defined in the sequence, and the sliding step length refers to the length of each sliding of the window;
Illustratively, defining a prediction time step of 1 in the historical drug dose sequence { B1, B2,..2, bt } and the physiological index sequence { C1, C2,..ct }, defining a window of sliding window length of 3, defining a sliding step of 1, then the first historical drug dose sequence sample is { B1, B2, B3}, the first physiological index sequence sample is { C1, C2, C3}, the first training sample is { { B1, B2, B3}, { C1, C2, C3}, and predicting future physiological index data within a prediction time step of { C4}, by the first training sample;
the second historical drug dose sequence sample is { B2, B3, B4}, the second physiological index sequence sample is { C2, C3, C4}, the second training sample is { { B2, B3, B4}, { C2, C3, C4}, and future physiological index data within a predicted time step is predicted to be { C5}, by the second training sample;
the step describes how to train a physiological index prediction model for predicting future physiological index data by using the historical drug dose and the physiological index data of the patient, and the change of the future physiological index data of the patient under different drug doses is known, so that the beneficial effect of providing a model foundation for calculating the drug risk by using the future physiological index data later is achieved;
Step S400: predicting future physiological index data Q of the patient by using a physiological index prediction model, and predicting the drug dosage Y of the patient by using a dosage prediction model;
the predicting the future physiological index data Q of the patient by using the physiological index prediction model refers to collecting the drug dosage data and the physiological index data of the patient, and predicting the future physiological index data of the patient by using the trained physiological index prediction model;
the step of predicting the drug dosage Y of the patient by using the dosage prediction model is to collect the disease data of the patient and predict the drug dosage of the patient by using the trained dosage prediction model;
the future physiological index data is obtained based on the predicted drug dose, and reflects the influence degree of the predicted drug dose on the body of the patient;
this step describes how the trained model is used to predict future physiological index data and drug dosage of the patient, providing data support for subsequent association rule mining using future physiological index data and drug dosage;
step S500: comparing whether future physiological index data Q of the patient is larger than a physiological index threshold Q0, and if so, setting the patient as an abnormal patient;
The physiological index threshold Q0 is set according to the experience of medical specialists and the actual situation of the patient, when the future physiological index data of the patient is smaller than or equal to the physiological index threshold Q0, the future physiological index data of the patient is normal, and if the future physiological index data of the patient is normal, the future drug dosage does not need to be adjusted; when the future physiological index data of the patient is larger than the physiological index threshold value Q0, the future physiological index data of the patient is abnormal, and the patient is an abnormal patient and needs to adjust the future drug dosage;
the method comprises the steps of determining whether future physiological index data of a patient is abnormal according to a physiological index threshold value, and if the future physiological index data of the patient is greater than a physiological index threshold value Q0, indicating that the patient is abnormal, and if the patient is an abnormal patient, taking drug risk assessment to treat abnormal conditions, so that the physiological abnormal conditions of the patient are identified, and the beneficial effect of providing a judgment basis for subsequent drug risk assessment is achieved;
step S600: calculating a drug risk coefficient R of the abnormal patient by using future physiological index data of the abnormal patient, and sending a drug risk early warning to a doctor;
the specific method for calculating the drug risk coefficient R of the abnormal patient by using the future physiological index data of the abnormal patient comprises the following steps:
Step S610: acquiring an association rule set;
the specific method for acquiring the association rule set comprises the following steps:
step S611: acquiring a data set of medicine, medicine dosage and future physiological index data used by an abnormal patient;
the acquisition mode of the data set is as follows:
counting the medicine doses of all kinds of medicines used by G abnormal patients, and dividing the medicine doses into different dose levels; carrying out data coding on different medicines and future physiological index data;
for example, the drug is of the type { E1, E2,., E5}, the dosage levels of the drug doses are { Y1, Y2, Y3, the codes of the future physiological index data are { Q1, Q2, Q3};
the future physiological index data are future physiological index data of an abnormal patient, the future physiological index data of the abnormal patient are used for association rule mining, and the future physiological index data used for association rule mining are larger than a physiological index threshold Q0;
step S612: generating a frequent item set by using an association rule mining algorithm based on the data set;
the process of generating the frequent item set is as follows:
setting a support degree threshold W, wherein the support degree threshold is set according to actual conditions and expert suggestions;
Setting the initial value of the variable z as 1, scanning the data set, calculating the occurrence frequency of future physiological index data, medicines and medicine doses of all patients, and generating a candidate item set z;
scanning a candidate item set z, counting the occurrence frequency of each item to obtain a support degree, screening the items with the support degree larger than or equal to a support degree threshold W to obtain a frequent item set z, connecting and pruning the frequent item set z, updating the value of j to z+1, generating a candidate item set z+1, scanning the candidate item set z+1, and counting the candidate items with each support degree larger than the support degree threshold W to obtain a frequent item set z+1;
repeating connection and pruning operations on the frequent item sets to generate candidate item sets, and counting items with the support degree greater than a support degree threshold value to obtain the frequent item sets until the generated candidate item sets are empty;
it should be noted that, using the Apriori algorithm to mine the frequent item set for the data set as a conventional technical means of data mining, the present application is not illustrated herein schematically, but for the sake of easier implementation of the present application, the present application provides the following examples about generating the frequent item set:
assuming a data set, as shown in table 1, comprising 5 drugs E1, E2, E5, dose level of 3 drug doses Y1, Y2, Y3}, future physiological index data codes are Q1, Q2, Q3 respectively, and the support threshold is set to w=0.3:
Candidate set 1 is obtained as shown in table 2:
screening the items with the support degree greater than or equal to the support degree threshold W to obtain a frequent item set 1, as shown in Table 3:
the connection and pruning operations are performed to obtain candidate set 2, as shown in table 4:
screening the items with the support degree greater than or equal to the support degree threshold value to obtain frequent item set 2, as shown in table 5:
connection and pruning were performed to obtain candidate set 3 as shown in table 6:
screening the items with support greater than the support threshold value to obtain frequent item set 3, as shown in table 7:
the candidate item set cannot be generated again, and the excavation process is finished;
the frequent items in the frequent item set 3 are the association rules among the required medicines, the medicine dosages and the future physiological index data;
step S613: generating a set of association rules based on the set of frequent items;
the association rules in the association rule set are association rules with support degree more than or equal to a support degree threshold W, and the association rule set consists of association rules in frequent item sets;
step S620: according to the association rule set, obtaining association rule weight Sx of an association rule item matched with future physiological index data;
the method for obtaining the association rule weight of the association rule item matched with the future physiological index data according to the association rule set comprises the following steps: obtaining association rule items between future physiological index data and medicines and medicine doses from an association rule set, and obtaining the support degree of the association rule items according to the association rule set, wherein the support degree is the association rule weight of the association rule items matched with the future physiological index data, and the association rule items represent medicines and medicine doses influencing the future physiological index data;
The purpose of obtaining the association rule item and the association rule weight thereof is to obtain the influence degree of the drug and the drug dosage on future physiological index data, when the support degree of a certain dosage level of a certain drug and the future physiological index data is greater than a support degree threshold value, the influence degree of the association rule item is the support degree, and the future physiological index data is associated with the drug and the dosage level;
the association rule weight Sx refers to the association rule weight of the x-th association rule item;
for example, the association rule term and association rule weight S1 of the future physiological index data are denoted as S { E1, Y2, Q3} = S1, and the degree of influence on the future physiological index data of Q3 when the dose level of the drug E1 is Y2 is S1;
for another example, the association rule terms and their association rule weights that match future physiological index data include:
S{E1，Y2，Q3}＝S1；
S{E2，Y3，Q3}＝S2；
S{E4，Y3，Q3}＝S3；
step S630: calculating a drug risk coefficient P of the abnormal patient according to the association rule weight;
the drug risk coefficient P is calculated according to the association rule weight between future physiological index data and the drug and drug dosage, the influence degree on the physiological index of the patient is represented, the drug risk coefficient can be used for knowing the adaptation degree of the drug and the drug dosage to the treatment of the patient, and the doctor adjusts the medication strategy of the patient according to the drug risk coefficient;
The calculation method of the drug risk coefficient P comprises the following steps:wherein x refers to an x-th association rule item, sx refers to association rule weight of the x-th association rule item, and Hx refers to condition satisfaction of the x-th association rule item;
the condition satisfaction degree is set according to the actual situation and the advice of medical specialists, and represents the matching degree between future physiological index data of the patient and the association rule item;
illustratively, a case is given in which the method solves the problem: patient Zhang three, collect its past 5 years history condition data and physiological index data, physiological index data include: the liver function index, the kidney function index and the lung function index are used for collecting historical medicine doses, the medicine doses needed by Zhang Sanning are predicted by using the historical medicine doses of Zhang Sanning and the historical illness state data, the liver function index is abnormal by using the physiological index data of Zhang Sanning and the predicted medicine doses to predict Zhang Sanning future physiological index data, and the medicine risk coefficient is 70, so that medicine risk early warning is sent to doctors, and the doctors adjust the medication strategy of Zhang Sanning according to the medicine risk coefficient;
in the step, the association rule between the predicted drug dose and future physiological index data after the predicted drug dose is mined by using the association rule, and the drug risk coefficients of different patients under the conditions of different drugs and drug doses are calculated according to the association rule weight, so that the accuracy and pertinence of drug risk assessment are improved.
Example 2
As shown in fig. 2, another aspect of the present application provides a big data based drug risk assessment system, comprising:
the data collection module is used for collecting historical illness state data, historical medicine dosage and physiological index data of a patient;
a dose prediction model training module for training a dose prediction model for predicting a drug dose Y using historical patient condition data of the patient;
the physiological index prediction model training module is used for training a physiological index prediction model for predicting future physiological index data Q by using the historical drug dosage and the physiological index data of the patient;
the model prediction module is used for predicting future physiological index data Q of the patient by using the physiological index prediction model and predicting the drug dosage Y of the patient by using the dosage prediction model;
the abnormality screening module is used for comparing whether future physiological index data Q of the patient is larger than a physiological index threshold Q0 or not, and if so, setting the patient as an abnormal patient;
the risk coefficient calculation module is used for calculating a drug risk coefficient R of the abnormal patient by using future physiological index data of the abnormal patient and sending a drug risk early warning to a doctor;
further, the dose prediction model training module further comprises:
A loss function minimization module for taking a loss function between a minimized actual historical medication dose and a predicted historical medication dose as a training target for a training dose prediction model;
further, the physiological index prediction model training module further includes:
a loss function module for minimizing a loss function of the actual physiological index data and the predicted physiological index dataAs a training target of a physiological index prediction model;
further, the model prediction module further includes:
the physiological index prediction module is used for predicting future physiological index data Q of the patient by using the physiological index prediction model;
a dose prediction module for predicting a drug dose Y of the patient using a dose prediction model;
further, the abnormality detection module further includes:
the physiological threshold setting module is used for setting a physiological index threshold Q0 according to the experience of medical specialists and the actual condition of the patient;
further, the risk coefficient calculation module further includes:
the association rule set acquisition module is used for acquiring an association rule set;
the association rule weight calculation module is used for obtaining association rule weights Sx of association rule items matched with future physiological index data according to the association rule set;
The drug risk calculation module is used for calculating a drug risk coefficient P of the abnormal patient according to the association rule weight;
further, the association rule set obtaining module further includes:
the data set acquisition module is used for acquiring data sets of medicines, medicine doses and future physiological index data used by abnormal patients;
the frequent item set generation module is used for generating a frequent item set by using an association rule mining algorithm based on the data set;
the association rule set generation module is used for generating an association rule set based on the frequent item set;
further, the data set acquisition module further includes:
the dose grading module is used for counting the drug doses of all kinds of drugs used by G abnormal patients and dividing the drug doses into different dose grades;
the data coding module is used for data coding of different medicines and future physiological index data;
further, the frequent item set generation module further includes:
the threshold setting module is used for setting a support degree threshold W;
further, the association rule set generation module further includes:
the association rule item acquisition module is used for acquiring association rule items between future physiological index data and medicaments and medicament doses from the association rule set;
And the association rule weight acquisition module is used for obtaining the support degree of the association rule item according to the association rule set, wherein the support degree is the association rule weight of the association rule item matched with the future physiological index data.
Example 3
Fig. 3 is a schematic structural diagram of an electronic device provided in the present application. As shown in fig. 3, an electronic device is also provided according to another aspect of the present application. The electronic device includes one or more processors and one or more memories. Wherein the memory has stored therein computer readable code which, when executed by the one or more processors, may perform the medication risk assessment method as described above.
The method or system according to embodiments of the present application may also be implemented by means of the architecture of the electronic device shown in fig. 3. As shown in fig. 3, the electronic device includes a bus, one or more CPUs, a Read Only Memory (ROM), a Random Access Memory (RAM), a communication port connected to a network, an input/output component, a hard disk, and the like. A storage device, such as a ROM or hard disk, in an electronic device may store the big data based medication risk assessment method provided herein. The big data based drug risk assessment method may for example comprise: collecting historical illness state data, historical medicine dosage and physiological index data of a patient; training a dose prediction model for predicting a drug dose Y using historical patient condition data of the patient; training a physiological index prediction model for predicting future physiological index data Q by using the historical drug dosage and the physiological index data of the patient; predicting future physiological index data Q of the patient by using a physiological index prediction model, and predicting the drug dosage Y of the patient by using a dosage prediction model; comparing whether future physiological index data Q of the patient is larger than a physiological index threshold Q0, and if so, setting the patient as an abnormal patient; and calculating a drug risk coefficient R of the abnormal patient by using future physiological index data of the abnormal patient, and sending a drug risk early warning to a doctor. Further, the electronic device may also include a user interface. Of course, the architecture shown in fig. 3 is merely exemplary, and one or more components of the electronic device shown in fig. 3 may be omitted as may be practical in implementing different devices.
Example 4
Fig. 4 is a schematic structural diagram of a computer readable storage medium according to an embodiment of the present application. As shown in fig. 4, is a computer readable storage medium according to one embodiment of the present application. The computer readable storage medium has computer readable instructions stored thereon. When the computer readable instructions are executed by the processor, the big data based medication risk assessment method according to the embodiments of the present application described with reference to the above figures may be performed. Storage media include, but are not limited to, for example, volatile memory and/or nonvolatile memory. Volatile memory can include, for example, random Access Memory (RAM), cache memory (cache), and the like. The non-volatile memory may include, for example, read Only Memory (ROM), hard disk, flash memory, and the like.
In addition, according to embodiments of the present application, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, the present application provides a non-transitory machine-readable storage medium storing machine-readable instructions executable by a processor to perform instructions corresponding to the method steps provided herein, such as: collecting historical illness state data, historical medicine dosage and physiological index data of a patient; training a dose prediction model for predicting a drug dose Y using historical patient condition data of the patient; training a physiological index prediction model for predicting future physiological index data Q by using the historical drug dosage and the physiological index data of the patient; predicting future physiological index data Q of the patient by using a physiological index prediction model, and predicting the drug dosage Y of the patient by using a dosage prediction model; comparing whether future physiological index data Q of the patient is larger than a physiological index threshold Q0, and if so, setting the patient as an abnormal patient; and calculating a drug risk coefficient R of the abnormal patient by using future physiological index data of the abnormal patient, and sending a drug risk early warning to a doctor. The above-described functions defined in the method of the present application are performed when the computer program is executed by a Central Processing Unit (CPU).
The methods and apparatus, devices, and apparatus of the present application may be implemented in numerous ways. For example, the methods and apparatus, devices of the present application may be implemented by software, hardware, firmware, or any combination of software, hardware, firmware. The above-described sequence of steps for the method is for illustration only, and the steps of the method of the present application are not limited to the sequence specifically described above unless specifically stated otherwise. Furthermore, in some embodiments, the present application may also be implemented as programs recorded in a recording medium, the programs including machine-readable instructions for implementing the methods according to the present application. Thus, the present application also covers a recording medium storing a program for executing the method according to the present application.
In addition, in the foregoing technical solutions provided in the embodiments of the present application, parts consistent with implementation principles of corresponding technical solutions in the prior art are not described in detail, so that redundant descriptions are avoided.
The purpose, technical scheme and beneficial effects of the invention are further described in detail in the detailed description. It is to be understood that the above description is only of specific embodiments of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A big data based medication risk assessment method, comprising:
collecting historical illness state data, historical medicine dosage and physiological index data of a patient;
training a dose prediction model for predicting a drug dose Y using historical patient condition data of the patient;
training a physiological index prediction model for predicting future physiological index data Q by using the historical drug dosage and the physiological index data of the patient;
predicting future physiological index data Q of the patient by using a physiological index prediction model, and predicting the drug dosage Y of the patient by using a dosage prediction model;
and comparing whether the future physiological index data Q of the patient is larger than a physiological index threshold Q0, if so, setting the patient as an abnormal patient, calculating a drug risk coefficient R of the abnormal patient by using the future physiological index data of the abnormal patient, and sending a drug risk early warning to a doctor.
2. The big data based medication risk assessment method according to claim 1, wherein the historical condition data, the historical medication dose, and the physiological index data are collected by the following ways: setting a data acquisition period T, and acquiring the historical illness state data, the historical medicine dosage and the physiological index data of the patient from the public medical database once every the time of the data acquisition period T, wherein the total number of the data acquisition period T is T.
3. The big data based medication risk assessment method as recited in claim 2The training process of the dose prediction model for predicting the drug dose Y by using the historical illness state data of the patient is characterized by comprising the following steps: collecting T times according to a data acquisition period T to obtain the historical illness state data as A1, A2, and At, obtaining the historical medicine doses as B1, B2, and Bt, wherein the historical illness state data and the corresponding historical medicine doses are used as a group of training samples, T training samples are respectively (A1, B1), (A2, B2), and (At, bt), i training samples are (Ai, bi), the training samples of the dose prediction model use the historical illness state data as input data, the historical medicine doses required by the corresponding historical illness state data are used as output data, the dose prediction model is trained, bi represents the actual historical medicine doses in the i training samples,and expressing the predicted historical medicine dosage of the ith training sample, taking the accurately predicted historical medicine dosage as a prediction target of a dosage prediction model, taking a loss function between the minimized actual historical medicine dosage and the predicted historical medicine dosage as a training target of a training dosage prediction model, and obtaining a dosage prediction model for predicting the medicine dosage Y.
4. A big data based drug risk assessment method according to claim 3, wherein the training process of the physiological index prediction model for predicting future physiological index data Q is: obtaining a historical medicament dose sequence { B1, B2, & gt, bt } from the historical medicament dose of the patient, obtaining a physiological index sequence { C1, C2, & gt, ct } from the physiological index data of the patient, setting a prediction time step, a sliding window length and a sliding step, dividing the historical medicament dose sequence and the physiological index sequence into t historical medicament dose sequence samples and t physiological index sequence samples by utilizing a sliding window method, and training a physiological index prediction model; the historical medicine dose sequence sample and the physiological index sequence sample are used as training samples for training a physiological index prediction model to be input, and physiological index data in a prediction time step is used as trainingThe output data of the physical index prediction model is trained, cj represents the actual physical index data in the jth training sample,representing predicted physiological index data in the jth training sample, taking the accurate predicted physiological index data as a prediction target of a physiological index prediction model to minimize the actual physiological index data and a loss function of the predicted physiological index data ≡ >As a training target of a physiological index prediction model.
5. The big data based drug risk assessment method according to claim 4, wherein the physiological index threshold Q0 is set according to the experience of a medical expert and the actual situation of the patient, and when the future physiological index data of the patient is smaller than or equal to the physiological index threshold Q0, the future physiological index data of the patient is normal, and if the future physiological index data of the patient is normal, the future drug dosage does not need to be adjusted; when the future physiological index data of the patient is greater than the physiological index threshold value Q0, the future physiological index data of the patient is abnormal, and the patient is an abnormal patient, so that the future drug dosage needs to be adjusted.
6. The big data based drug risk assessment method according to claim 5, wherein the specific method for calculating the drug risk coefficient R of the abnormal patient using the future physiological index data of the abnormal patient is as follows: acquiring an association rule set; according to the association rule set, obtaining association rule weight Sx of an association rule item matched with future physiological index data; and calculating the drug risk coefficient P of the abnormal patient according to the association rule weight.
7. The big data based medication risk assessment method according to claim 6, wherein the calculation method of the medication risk coefficient P is as follows: Wherein x refers to the x-th association rule item, sx refers to the association rule weight of the x-th association rule item, and Hx refers to the condition satisfaction of the x-th association rule item.
8. A big data based medication risk assessment system, comprising:
the data collection module is used for collecting historical illness state data, historical medicine dosage and physiological index data of a patient;
a dose prediction model training module for training a dose prediction model for predicting a drug dose Y using historical patient condition data of the patient;
the physiological index prediction model training module is used for training a physiological index prediction model for predicting future physiological index data Q by using the historical drug dosage and the physiological index data of the patient;
the model prediction module is used for predicting future physiological index data Q of the patient by using the physiological index prediction model and predicting the drug dosage Y of the patient by using the dosage prediction model;
the abnormality screening module is used for comparing whether future physiological index data Q of the patient is larger than a physiological index threshold Q0 or not, and if so, setting the patient as an abnormal patient;
the risk coefficient calculation module is used for calculating a drug risk coefficient R of the abnormal patient by using future physiological index data of the abnormal patient and sending a drug risk early warning to a doctor.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, when executing the program, to implement the steps in the big data based medication risk assessment method according to any of claims 1-7.
10. A readable storage medium, characterized in that it stores a computer program adapted to be loaded by a processor for performing the steps of the big data based drug risk assessment method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311492271.9A CN117352178A (en) | 2023-11-10 | 2023-11-10 | Big data-based drug risk assessment system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311492271.9A CN117352178A (en) | 2023-11-10 | 2023-11-10 | Big data-based drug risk assessment system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117352178A true CN117352178A (en) | 2024-01-05 |
Family
ID=89359552
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311492271.9A Pending CN117352178A (en) | 2023-11-10 | 2023-11-10 | Big data-based drug risk assessment system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117352178A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046362A (en) * | 2015-07-24 | 2015-11-11 | 河南科技大学 | Real-time prediction method of food safety on the basis of association rule mining |
| CN107341345A (en) * | 2017-06-22 | 2017-11-10 | 厦门大学 | A kind of clinical rational drug use methods of risk assessment based on electronic health record big data |
| CN111863266A (en) * | 2020-01-16 | 2020-10-30 | 上海海事大学 | Method for screening risk factors of sporadic colorectal adenomas based on directional weighted association rule model |
| CN113100719A (en) * | 2021-04-08 | 2021-07-13 | 中国人民解放军陆军特色医学中心 | Renal dialysis patient cardiovascular event prediction system based on machine learning |
| CN115526531A (en) * | 2022-10-21 | 2022-12-27 | 国家食品安全风险评估中心 | Wheat quality safety early warning method and system |
| CN116884615A (en) * | 2023-07-31 | 2023-10-13 | 吾征智能技术(北京)有限公司 | Hypertension risk prediction method, system, electronic equipment and storage medium |
-
2023
- 2023-11-10 CN CN202311492271.9A patent/CN117352178A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046362A (en) * | 2015-07-24 | 2015-11-11 | 河南科技大学 | Real-time prediction method of food safety on the basis of association rule mining |
| CN107341345A (en) * | 2017-06-22 | 2017-11-10 | 厦门大学 | A kind of clinical rational drug use methods of risk assessment based on electronic health record big data |
| CN111863266A (en) * | 2020-01-16 | 2020-10-30 | 上海海事大学 | Method for screening risk factors of sporadic colorectal adenomas based on directional weighted association rule model |
| CN113100719A (en) * | 2021-04-08 | 2021-07-13 | 中国人民解放军陆军特色医学中心 | Renal dialysis patient cardiovascular event prediction system based on machine learning |
| CN115526531A (en) * | 2022-10-21 | 2022-12-27 | 国家食品安全风险评估中心 | Wheat quality safety early warning method and system |
| CN116884615A (en) * | 2023-07-31 | 2023-10-13 | 吾征智能技术(北京)有限公司 | Hypertension risk prediction method, system, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110459324B (en) | Disease prediction method and device based on long-term and short-term memory model and computer equipment | |
| CN110880362B (en) | Large-scale medical data knowledge mining and treatment scheme recommending system | |
| US7844560B2 (en) | Personalized prognosis modeling in medical treatment planning | |
| KR102237449B1 (en) | Method, server and program of learning a patient diagnosis | |
| CN105528529B (en) | The data processing method of tcm clinical practice technical ability appraisement system based on big data analysis | |
| JP2006507875A (en) | System and method for automatically diagnosing patient health | |
| JP2008532104A (en) | A method, system, and computer program product for generating and applying a prediction model capable of predicting a plurality of medical-related outcomes, evaluating an intervention plan, and simultaneously performing biomarker causality verification | |
| CN110310740B (en) | Re-hospitalizing information prediction method and system based on cross attention neural network | |
| CN113270203A (en) | Drug dose prediction method, device, electronic device and storage medium | |
| Krikunov et al. | Complex data-driven predictive modeling in personalized clinical decision support for Acute Coronary Syndrome episodes | |
| Lu et al. | Understanding heart failure patients EHR clinical features via SHAP interpretation of tree-based machine learning model predictions | |
| CN111863238A (en) | Parallel intelligence based chronic disease diagnosis and treatment system and diagnosis and treatment method | |
| CN108231146A (en) | A kind of medical records model building method, system and device based on deep learning | |
| CN116864139A (en) | Disease risk assessment method, device, computer equipment and readable storage medium | |
| CN112542242A (en) | Data transformation/symptom scoring | |
| Chou et al. | Extracting drug utilization knowledge using self-organizing map and rough set theory | |
| WO2022010384A1 (en) | Clinical decision support system | |
| Schlosser et al. | Visual acuity prediction on real-life patient data using a machine learning based multistage system | |
| EP3929939A1 (en) | System and method for peri-anaesthetic risk evaluation | |
| CN116864104A (en) | Chronic thromboembolic pulmonary artery high-pressure risk classification system based on artificial intelligence | |
| CN117352178A (en) | Big data-based drug risk assessment system and method | |
| CN116259396A (en) | Treatment expense prediction method, system, equipment and storage medium based on machine learning | |
| CN113066531A (en) | Risk prediction method and device, computer equipment and storage medium | |
| US20190295701A1 (en) | Drug administration timing | |
| Wang et al. | Prediction Models for Glaucoma in a Multicenter Electronic Health Records Consortium: The Sight Outcomes Research Collaborative |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |