Untoward drug reaction monitoring and early warning system

Abstract

The invention discloses an adverse drug reaction monitoring and early warning system, which relates to the technical field of drug monitoring and early warning, and comprises a physiological data acquisition module, a drug risk grade evaluation module, a comprehensive monitoring module and a comprehensive early warning module, the data acquisition module is used for acquiring multi-dimensional patient physiological data including organ monitoring data, allergy monitoring data, metabolism monitoring data and drug heat monitoring data, and preprocessing the physiological data; according to the system, a multi-module cooperation technology, a multivariate data acquisition technology and a modern information technology are closely combined, real-time and comprehensive monitoring of the adverse drug reaction is achieved, through cooperation of all the modules, the risk level of the adverse drug reaction is evaluated, and the risk level of the adverse drug reaction is evaluated. Evaluating the risk level of the adverse drug reaction, and sending out a corresponding early warning signal according to an evaluation result.

Description

Technical Field

[0001] The present invention relates to the technical field of drug monitoring and early warning, and particularly relates to a drug adverse reaction monitoring and early warning system. Background Art

[0002] With the rapid development of modern medicine, drugs play a key role in disease treatment. However, the problem of drug adverse reactions has become increasingly prominent, threatening the lives and health of patients. Traditional drug adverse reaction monitoring mainly relies on the empirical judgments of doctors and pharmacists and the spontaneous reports of patients, which has obvious lag. Many ADR events are difficult to detect and warn in a timely manner, resulting in ineffective treatment. With the booming development of big data and artificial intelligence technologies, new opportunities have been brought to solve the problems of drug adverse reaction monitoring and early warning. However, in the process of monitoring drug adverse reactions in the existing technology, the data source is single, and it is difficult to comprehensively integrate the multi-dimensional physiological information of patients, resulting in inaccurate monitoring results. Moreover, in terms of the early warning mechanism, false negatives and false positives are likely to occur.

[0003] Although the existing technology has made great progress in the monitoring of drug adverse reactions, there are still some problems to be optimized. In the existing technology, it is difficult to analyze the multi-dimensional physiological data of patients, monitor and evaluate the adverse reactions of drugs in real time, classify the levels of drug adverse reactions, and issue corresponding warning signals according to the evaluated drug adverse reactions. Therefore, the accuracy of drug adverse reaction monitoring and early warning is affected. For this reason, a drug adverse reaction monitoring and early warning system is proposed. Summary of the Invention

[0004] To achieve the above objectives, the present invention is realized through the following technical solutions: A drug adverse reaction monitoring and early warning system includes a physiological data acquisition module, a drug risk level assessment module, a comprehensive monitoring module, and a comprehensive early warning module. Among them, each module is communicatively connected;

[0005] The physiological data acquisition module is used to collect multi-dimensional physiological data of patients including organ monitoring data, allergy monitoring data, metabolism monitoring data, and drug fever monitoring data, and preprocess the physiological data, providing data support for the realization of the functions of subsequent modules;

[0006] The drug risk level assessment module combines the preprocessed multi-dimensional physiological data of patients to evaluate the risk level of drug adverse reactions, providing a guarantee for constructing a drug adverse reaction monitoring model;

[0007] The comprehensive monitoring module constructs a drug adverse reaction monitoring model based on the evaluation results of the drug adverse reaction risk level, analyzes the drug adverse reactions of patients, realizes the analysis of the multi-dimensional physiological data of patients, and achieves the monitoring and evaluation of the adverse reactions of drugs;

[0008] The comprehensive early warning module combines the output results of the adverse drug reaction monitoring model and issues corresponding early warning signals, solving the problem of low early warning accuracy for adverse drug reactions in the prior art.

[0009] A further improvement of the technical solution of the present invention lies in that: the drug risk level assessment module includes an organ damage assessment unit, a drug allergy assessment unit, a drug metabolism assessment unit, and a drug fever assessment unit. Among them, the risk levels of adverse drug reactions include the risk level of drugs causing organ damage to patients, the risk level of drugs causing allergies to patients, the risk level of drugs affecting the metabolic function of patients, and the risk level of drugs causing drug fever to patients;

[0010] The organ damage assessment unit uses the preprocessed organ monitoring data to construct an organ damage risk monitoring model, obtains the percentage of organ damage of the patient, and further evaluates the risk level of drugs causing organ damage to the patient;

[0011] The drug allergy assessment unit analyzes the preprocessed allergy monitoring data, obtains the percentage of drug allergy of the patient, and further evaluates the risk level of drugs causing allergies to the patient;

[0012] The drug metabolism assessment unit calculates the drug metabolism index of the patient through the preprocessed metabolism monitoring data and evaluates the risk level of drugs affecting the metabolic function of the patient;

[0013] The drug fever assessment unit uses the preprocessed drug fever monitoring data to construct a drug fever risk monitoring model, obtains the percentage of drug fever of the patient, and further evaluates the risk level of drugs causing drug fever to the patient.

[0014] A further improvement of the technical solution of the present invention lies in that: in the process of collecting multi-dimensional physiological data of patients by the physiological data collection module:

[0015] Deploy different collection devices to collect organ monitoring data, allergy monitoring data, metabolism monitoring data, and drug fever monitoring data of patients. Among them, the collection devices include a multi-functional liquid phase chip analyzer, an oxidative stress free radical detection analyzer, a histamine detector, an ELISA kit, an enzyme label instrument, an incubator, a C-reactive protein analyzer, a blood cell analyzer, a bilirubin analyzer, and a radionuclide imaging device;

[0016] The organ monitoring data includes inflammatory factor concentration and oxidative stress index; the allergy monitoring data includes histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count; the metabolism monitoring data includes bilirubin concentration and glomerular filtration rate; the drug fever monitoring data includes eosinophil count and C-reactive protein concentration;

[0017] Using a multi-functional liquid-phase chip analyzer, collect the concentrations of inflammatory factors; through an oxidative stress free radical detection analyzer, collect oxidative stress indicators; using a histamine detector, collect histamine concentrations; in combination with an ELISA kit, an enzyme-labeled instrument, and an incubator, collect tryptase concentrations; using a C-reactive protein analyzer, collect C-reactive protein concentrations; through a blood cell analyzer, collect eosinophil counts; through a bilirubin analyzer, collect bilirubin concentrations; using a radionuclide imaging device, collect glomerular filtration rate;

[0018] Perform data cleaning and data standardization processing on the collected organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data, remove outliers and duplicate values, add timestamps to the organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data respectively, and synchronize the collection times of the organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data by adjusting the timestamps;

[0019] Integrate the preprocessed organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data into a patient physiological data set, and divide the patient physiological data set into a training set and a test set, where the ratio of the training set to the test set is 8:2.

[0020] A further improvement of the technical solution of the present invention lies in: for the organ damage assessment unit, the process of obtaining the percentage of patient organ damage includes:

[0021] Extract the organ monitoring data of the patient from the patient physiological data set, use the organ monitoring data in the training set, in combination with the multiple linear regression algorithm, take the organ monitoring data as the input and the percentage of patient organ damage as the output, learn the linear relationship between the organ monitoring data and the percentage of patient organ damage, and train the organ damage risk monitoring model;

[0022] Input the organ monitoring data in the test set into the organ damage monitoring model, optimize the organ damage monitoring model by adjusting the intercept term and regression coefficient of the organ damage monitoring model, and obtain the final organ damage risk monitoring model;

[0023] The expression of this organ damage risk monitoring model is as follows:

[0024]

[0025] Wherein, is the percentage of patient organ damage, and are the concentration of inflammatory factors and oxidative stress indicators respectively, is the intercept term of the organ damage risk monitoring model, is the error term of the organ damage risk monitoring model, and They are the regression coefficients of the inflammatory factor concentration and the oxidative stress index respectively;

[0026] Input the organ monitoring data into the organ damage risk monitoring model, output the percentage of the patient's organ damage through the organ damage risk monitoring model, and integrate the percentage of the patient's organ damage into the patient's physiological data set.

[0027] A further improvement of the technical solution of the present invention lies in that: the process of the organ damage assessment unit for assessing the risk level of drug-induced organ damage in a patient includes:

[0028] The risk levels of drug-induced organ damage in the patient include low organ damage risk, medium organ damage risk, and high organ damage risk;

[0029] According to the percentage of the patient's organ damage output by the organ damage monitoring model, when the percentage of the patient's organ damage is less than 20%, the risk level of drug-induced organ damage in the patient corresponds to low organ damage risk;

[0030] When the percentage of the patient's organ damage is between 20% and 50%, the risk level of drug-induced organ damage in the patient corresponds to medium organ damage risk;

[0031] When the percentage of the patient's organ damage is greater than 50%, the risk level of drug-induced organ damage in the patient corresponds to high organ damage risk.

[0032] A further improvement of the technical solution of the present invention lies in that: the process of the drug allergy assessment unit for assessing the risk level of drug-induced allergy in a patient includes:

[0033] The risk levels of drug-induced allergy in the patient include low allergy risk, medium allergy risk, and high allergy risk. Baseline ranges are set for the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count respectively;

[0034] Among them, the baseline range of the histamine concentration is less than 10 ng / ml, the baseline range of the tryptase concentration is less than 15 ng / ml, the baseline range of the C-reactive protein concentration is greater than 10 mg / L, and the baseline range of the eosinophil count is less than 0.5 per L;

[0035] When the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count are all within the corresponding set baseline ranges, assign a drug allergy percentage of 10% to the patient, corresponding to low allergy risk;

[0036] When one of the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count data is not within the corresponding set baseline range, assign a drug allergy percentage of 35% to the patient, corresponding to a medium allergy risk;

[0037] When more than one of the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count data is not within the corresponding set baseline range, assign a drug allergy percentage of 75% to the patient, corresponding to a high allergy risk;

[0038] By analyzing the preprocessed allergy monitoring data, obtain the drug allergy percentage of the patient and integrate the drug allergy percentage of the patient into the patient's physiological dataset.

[0039] A further improvement of the technical solution of the present invention lies in that: in the process of the drug metabolism evaluation unit evaluating the risk level of the drug affecting the patient's metabolic function, it includes:

[0040] The risk levels of the drug affecting the patient's metabolic function include low metabolic risk, medium metabolic risk, and high metabolic risk;

[0041] Divide the bilirubin risk assessment range, which includes a low bilirubin risk assessment range, a medium bilirubin risk assessment range, and a high bilirubin risk assessment range. Among them, the low bilirubin risk assessment range is less than , the medium bilirubin risk assessment range is , and the high bilirubin risk assessment range is greater than ;

[0042] Assign bilirubin risk percentages of 10%, 30%, and 65% to the low bilirubin risk assessment range, the medium bilirubin risk assessment range, and the high bilirubin risk assessment range respectively;

[0043] Divide the glomerular filtration rate risk assessment range, which consists of a low glomerular filtration rate risk assessment range, a medium glomerular filtration rate risk assessment range, and a high glomerular filtration rate risk assessment range. Among them, the low glomerular filtration rate risk assessment range is greater than , the medium glomerular filtration rate risk assessment range is , and the high glomerular filtration rate risk assessment range is less than ;

[0044] Assign glomerular filtration rate risk percentages of 10%, 45%, and 90% to the low glomerular filtration rate risk assessment range, the medium glomerular filtration rate risk assessment range, and the high glomerular filtration rate risk assessment range respectively;

[0045] Weights are assigned to the corresponding bilirubin risk percentage and glomerular filtration rate risk percentage respectively. Using the assigned weights, the patient's drug metabolism index is calculated, and the calculated patient's drug metabolism index is integrated into the patient's physiological dataset. The calculation formula is as follows:

[0046]

[0047] Among them, is the patient's drug metabolism index, and are the weights of the bilirubin risk percentage and the glomerular filtration rate risk percentage respectively, and are the bilirubin risk percentage and the glomerular filtration rate risk percentage respectively;

[0048] According to the patient's drug metabolism index, when the patient's drug metabolism index is lower than 0.2, the risk level of the corresponding drug affecting the patient's metabolic function is low metabolism risk; when the patient's drug metabolism index is between 0.2 and 0.5, the risk level of the corresponding drug affecting the patient's metabolic function is medium metabolism risk; when the patient's drug metabolism index is higher than 0.5, the risk level of the corresponding drug affecting the patient's metabolic function is high metabolism risk.

[0049] A further improvement of the technical solution of the present invention lies in that: in the process of the drug heat evaluation unit evaluating the risk level of drug-induced drug heat in patients, it includes:

[0050] The risk levels of drug-induced drug heat in patients include low drug heat risk, medium drug heat risk and high drug heat risk;

[0051] Extract the drug heat monitoring data in the patient's physiological dataset. Using the drug heat monitoring data in the training set and combining with the multiple linear regression algorithm, taking the drug heat monitoring data as the input and the patient's drug heat percentage as the output, learn the linear relationship between the drug heat monitoring data and the patient's drug heat percentage, and train the drug heat risk monitoring model;

[0052] Input the drug heat monitoring data in the test set into the drug heat risk monitoring model. By adjusting the intercept term and regression coefficient of the drug heat risk monitoring model, optimize the drug heat risk monitoring model to obtain the final drug heat risk monitoring model. The expression of this drug heat risk monitoring model is:

[0053]

[0054] Among them, is the patient's drug heat percentage, is the intercept term of the drug heat risk monitoring model, is the error term of the drug heat risk monitoring model, and the eosinophil count and the C-reactive protein concentration respectively, and the regression coefficients of the eosinophil count and the C-reactive protein concentration respectively;

[0055] Input the drug fever monitoring data into the drug fever risk monitoring model, output the percentage of patients with drug fever through the drug fever risk monitoring model, and integrate the percentage of patients with drug fever into the patient physiological dataset;

[0056] When the percentage of patients with drug fever is lower than 25%, it corresponds to a low risk of drug fever; when the percentage of patients with drug fever is between 25% and 65%, it corresponds to a medium risk of drug fever; when the percentage of patients with drug fever is greater than 65%, it corresponds to a high risk of drug fever.

[0057] A further improvement of the technical solution of the present invention lies in that: the process of constructing the adverse drug reaction monitoring model by the comprehensive monitoring module based on the evaluation result of the risk level of adverse drug reactions includes:

[0058] Using the data in the training set, combined with the neural network model, taking the organ monitoring data, allergy monitoring data, metabolic monitoring data and drug fever monitoring data as inputs, and taking the percentage of organ damage of the patient, the percentage of drug allergy of the patient, the drug metabolism index of the patient and the percentage of drug fever of the patient as outputs, respectively learn the non-linear relationship between the organ monitoring data and the percentage of organ damage of the patient, the non-linear relationship between the allergy monitoring data and the percentage of drug allergy of the patient, the non-linear relationship between the metabolic monitoring data and the drug metabolism index of the patient, and the non-linear relationship between the drug fever monitoring data and the percentage of drug fever of the patient, and train the adverse drug reaction monitoring model;

[0059] Input the organ monitoring data, allergy monitoring data, metabolic monitoring data and drug fever monitoring data in the test set into the adverse drug reaction monitoring model respectively, compare the error between the output result of the adverse drug reaction monitoring model and the actual percentage of organ damage of the patient, the percentage of drug allergy of the patient, the drug metabolism index of the patient and the percentage of drug fever of the patient respectively, evaluate the performance of the adverse drug reaction monitoring model, adjust the parameters of the adverse drug reaction monitoring model, optimize the adverse drug reaction monitoring model, and obtain the final adverse drug reaction monitoring model.

[0060] A further improvement of the technical solution of the present invention lies in that: the process of the comprehensive warning module combining the output result of the adverse drug reaction monitoring model and sending out corresponding warning signals includes:

[0061] Input organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data into the adverse drug reaction monitoring model respectively. Through the adverse drug reaction monitoring model, output the percentage of organ damage of the patient, the percentage of drug allergy of the patient, the drug metabolism index of the patient, and the percentage of drug fever of the patient respectively;

[0062] Set up a warning display screen. According to the risk levels of adverse drug reactions corresponding to the percentage of organ damage of the patient, the percentage of drug allergy of the patient, the drug metabolism index of the patient, and the percentage of drug fever of the patient respectively, the warning display screen sends out corresponding warning signals in the form of text display.

[0063] The beneficial effects of the present invention are as follows: In the adverse drug reaction monitoring and warning system of the present invention, compared with the traditional adverse drug reaction monitoring and warning system, the multi-module cooperation technology, multi-source data acquisition technology, and modern information technology in the system of the present invention are closely combined to accurately capture the organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data of the patient, and then obtain the percentage of organ damage of the patient, the percentage of drug allergy, the drug metabolism index, and the percentage of drug fever, achieving real-time and comprehensive monitoring of adverse drug reactions. Through the collaborative work of each module, the risk level of adverse drug reactions is evaluated, and corresponding warning signals are sent according to the evaluation results, solving the problem in the prior art that it is difficult to analyze the multi-dimensional physiological data of patients in adverse drug reaction monitoring, real-time monitor and evaluate adverse drug reactions, classify the levels of adverse drug reactions, and send corresponding warning signals according to the evaluated adverse drug reactions, which affects the monitoring and warning accuracy of adverse drug reactions, ensuring that the present invention can refine the dynamic monitoring standard for an adverse drug reaction monitoring and warning system within a more accurate range, making the monitored data more accurate indicators under the same conditions. The research and application of this method significantly enhance the degree of intelligence in the process of adverse drug reaction monitoring and warning. Description of the Drawings

[0064] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0065] Figure 1 It is a block diagram of an adverse drug reaction monitoring and warning system of the present invention. Detailed Embodiments

[0066] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0067] As Figure 1 shown, the present invention provides a drug adverse reaction monitoring and early warning system, including a physiological data acquisition module, a drug risk level assessment module, a comprehensive monitoring module, and a comprehensive early warning module. Among them, each module is communicatively connected;

[0068] The physiological data acquisition module is used to collect multi-dimensional patient physiological data including organ monitoring data, allergy monitoring data, metabolism monitoring data, and drug fever monitoring data, and preprocess the physiological data, providing data support for the implementation of subsequent module functions;

[0069] The drug risk level assessment module combines the preprocessed multi-dimensional patient physiological data to evaluate the risk level of drug adverse reactions, providing a guarantee for constructing a drug adverse reaction monitoring model;

[0070] The comprehensive monitoring module constructs a drug adverse reaction monitoring model based on the evaluation result of the drug adverse reaction risk level, analyzes the drug adverse reactions of the patient, realizes the analysis of the multi-dimensional physiological data of the patient, and achieves the monitoring and evaluation of the drug adverse reactions;

[0071] The comprehensive early warning module combines the output result of the drug adverse reaction monitoring model and issues a corresponding early warning signal, solving the problem of low early warning accuracy of drug adverse reactions in the prior art.

[0072] Preferably, the drug risk level assessment module includes an organ damage assessment unit, a drug allergy assessment unit, a drug metabolism assessment unit, and a drug fever assessment unit. Among them, the risk levels of drug adverse reactions include the risk level of drug-induced organ damage to the patient, the risk level of drug-induced allergy to the patient, the risk level of drug affecting the patient's metabolic function, and the risk level of drug-induced drug fever in the patient;

[0073] The organ damage assessment unit uses the preprocessed organ monitoring data to construct an organ damage risk monitoring model, obtains the percentage of organ damage of the patient, and then evaluates the risk level of drug-induced organ damage to the patient;

[0074] The drug allergy assessment unit analyzes the preprocessed allergy monitoring data, obtains the percentage of drug allergy of the patient, and then evaluates the risk level of drug-induced allergy to the patient;

[0075] A drug metabolism evaluation unit calculates a patient's drug metabolism index based on the preprocessed metabolism monitoring data and evaluates the risk level of the drug's impact on the patient's metabolic function.

[0076] A drug fever evaluation unit constructs a drug fever risk monitoring model using the preprocessed drug fever monitoring data, obtains the percentage of drug fever in the patient, and further evaluates the risk level of the drug causing drug fever in the patient.

[0077] Preferably, the physiological data acquisition module. The process of collecting multi-dimensional patient physiological data includes:

[0078] Deploy different acquisition devices to collect the patient's organ monitoring data, allergy monitoring data, metabolism monitoring data, and drug fever monitoring data. Among them, the acquisition devices include a multi-functional liquid phase chip analyzer, an oxidative stress free radical detection analyzer, a histamine detector, an ELISA kit, an enzyme labeler, an incubator, a C-reactive protein analyzer, a blood cell analyzer, a bilirubin analyzer, and a radionuclide imaging device.

[0079] Among them, the organ monitoring data includes the concentration of inflammatory factors and oxidative stress indicators; the allergy monitoring data includes the concentration of histamine, the concentration of tryptase, the concentration of C-reactive protein, and the eosinophil count; the metabolism monitoring data includes the concentration of bilirubin and the glomerular filtration rate; the drug fever monitoring data includes the eosinophil count and the concentration of C-reactive protein.

[0080] Use a multi-functional liquid phase chip analyzer to collect the concentration of inflammatory factors; use an oxidative stress free radical detection analyzer to collect oxidative stress indicators; use a histamine detector to collect the concentration of histamine; combine an ELISA kit, an enzyme labeler, and an incubator to collect the concentration of tryptase; use a C-reactive protein analyzer to collect the concentration of C-reactive protein; use a blood cell analyzer to collect the eosinophil count; use a bilirubin analyzer to collect the concentration of bilirubin; use a radionuclide imaging device to collect the glomerular filtration rate.

[0081] Perform data cleaning and data standardization processing on the collected organ monitoring data, allergy monitoring data, metabolism monitoring data, and drug fever monitoring data, remove outliers and duplicate values, add time stamps to the organ monitoring data, allergy monitoring data, metabolism monitoring data, and drug fever monitoring data respectively, and synchronize the collection times of the organ monitoring data, allergy monitoring data, metabolism monitoring data, and drug fever monitoring data by adjusting the time stamps.

[0082] Integrate the preprocessed organ monitoring data, allergy monitoring data, metabolism monitoring data, and drug fever monitoring data into a patient physiological data set, and divide the patient physiological data set into a training set and a test set. Among them, the ratio of the training set to the test set is 8:2.

[0083] Preferably, for the organ damage assessment unit, the process of obtaining the percentage of organ damage of the patient includes:

[0084] Extract the organ monitoring data of the patient from the patient's physiological dataset. Using the organ monitoring data in the training set and combining with the multiple linear regression algorithm, take the organ monitoring data as the input and the percentage of organ damage of the patient as the output to learn the linear relationship between the organ monitoring data and the percentage of organ damage of the patient, and train the organ damage risk monitoring model;

[0085] Input the organ monitoring data in the test set into the organ damage monitoring model, and optimize the organ damage monitoring model by adjusting the intercept term and regression coefficient of the organ damage monitoring model to obtain the final organ damage risk monitoring model;

[0086] The expression of the organ damage risk monitoring model is as follows:

[0087]

[0088] Where, is the percentage of organ damage of the patient, and are the concentration of inflammatory factors and oxidative stress index respectively, is the intercept term of the organ damage risk monitoring model, is the error term of the organ damage risk monitoring model, and are the regression coefficients of the concentration of inflammatory factors and oxidative stress index respectively;

[0089] Input the organ monitoring data into the organ damage risk monitoring model, output the percentage of organ damage of the patient through the organ damage risk monitoring model, and integrate the percentage of organ damage of the patient into the patient's physiological dataset.

[0090] Preferably, for the organ damage assessment unit, the process of evaluating the risk level of drug-induced organ damage in the patient includes:

[0091] Among them, the risk levels of drug-induced organ damage in the patient include low organ damage risk, medium organ damage risk, and high organ damage risk;

[0092] According to the percentage of organ damage of the patient output by the organ damage monitoring model, when the percentage of organ damage of the patient is less than 20%, the corresponding risk level of drug-induced organ damage in the patient is low organ damage risk;

[0093] When the percentage of organ damage of the patient is between 20% and 50%, the corresponding risk level of drug-induced organ damage in the patient is medium organ damage risk;

[0094] When the percentage of organ damage in the patient is greater than 50%, the risk level of organ damage caused by the corresponding drug is a high organ damage risk.

[0095] Preferably, the drug allergy assessment unit, the process of assessing the risk level of drug-induced allergy in the patient includes:

[0096] Among them, the risk levels of drug-induced allergy in the patient include low allergy risk, medium allergy risk, and high allergy risk. Baseline ranges are set for histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count respectively;

[0097] Among them, the baseline range of histamine concentration is less than 10 ng / ml, the baseline range of tryptase concentration is less than 15 ng / ml, the baseline range of C-reactive protein concentration is greater than 10 mg / L, and the baseline range of eosinophil count is less than 0.5 per L;

[0098] When the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count are all within the corresponding set baseline ranges, the drug allergy percentage of the patient is given as 10%, corresponding to a low allergy risk;

[0099] When there is one item of data among the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count that is not within the corresponding set baseline range, the drug allergy percentage of the patient is given as 35%, corresponding to a medium allergy risk;

[0100] When there is more than one item of data among the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count that is not within the corresponding set baseline range, the drug allergy percentage of the patient is given as 75%, corresponding to a high allergy risk;

[0101] By analyzing the preprocessed allergy monitoring data, the drug allergy percentage of the patient is obtained, and the drug allergy percentage of the patient is integrated into the patient's physiological dataset.

[0102] Preferably, the drug metabolism assessment unit, the process of assessing the risk level of drug affecting the patient's metabolic function includes:

[0103] Among them, the risk levels of drug affecting the patient's metabolic function include low metabolism risk, medium metabolism risk, and high metabolism risk;

[0104] Divide the bilirubin risk assessment range, which includes bilirubin low risk assessment range, bilirubin medium risk assessment range, and bilirubin high risk assessment range. Among them, the bilirubin low risk assessment range is less than , the bilirubin medium risk assessment range is , the bilirubin high risk assessment range is greater than ;

[0105] Assign bilirubin risk percentages of 10%, 30%, and 65% to the low bilirubin risk assessment range, the medium bilirubin risk assessment range, and the high bilirubin risk assessment range, respectively;

[0106] Divide the glomerular filtration rate risk assessment range, which consists of the low glomerular filtration rate risk assessment range, the medium glomerular filtration rate risk assessment range, and the high glomerular filtration rate risk assessment range. Among them, the low glomerular filtration rate risk assessment range is greater than , the medium glomerular filtration rate risk assessment range is , and the high glomerular filtration rate risk assessment range is less than ;

[0107] Assign glomerular filtration rate risk percentages of 10%, 45%, and 90% to the low glomerular filtration rate risk assessment range, the medium glomerular filtration rate risk assessment range, and the high glomerular filtration rate risk assessment range, respectively;

[0108] Assign weights to the corresponding bilirubin risk percentage and glomerular filtration rate risk percentage respectively. Using the assigned weights, calculate the patient's drug metabolism index and integrate the calculated patient's drug metabolism index into the patient's physiological dataset. The calculation formula is as follows:

[0109]

[0110] Wherein, is the patient's drug metabolism index, and are the weights of the bilirubin risk percentage and the glomerular filtration rate risk percentage respectively, and are the bilirubin risk percentage and the glomerular filtration rate risk percentage respectively;

[0111] According to the patient's drug metabolism index, when the patient's drug metabolism index is lower than 0.2, the risk level of the corresponding drug affecting the patient's metabolic function is low metabolic risk; when the patient's drug metabolism index is between 0.2 and 0.5, the risk level of the corresponding drug affecting the patient's metabolic function is medium metabolic risk; when the patient's drug metabolism index is higher than 0.5, the risk level of the corresponding drug affecting the patient's metabolic function is high metabolic risk.

[0112] Preferably, the drug fever assessment unit, the process of assessing the risk level of drug-induced drug fever in patients includes:

[0113] Among them, the risk levels of drug-induced drug fever in patients include low drug fever risk, medium drug fever risk, and high drug fever risk;

[0114] Extract the drug fever monitoring data from the patient's physiological dataset. Using the drug fever monitoring data in the training set and combining with the multiple linear regression algorithm, take the drug fever monitoring data as the input and the patient's drug fever percentage as the output, learn the linear relationship between the drug fever monitoring data and the patient's drug fever percentage, and train the drug fever risk monitoring model;

[0115] Input the drug fever monitoring data in the test set into the drug fever risk monitoring model. By adjusting the intercept term and regression coefficients of the drug fever risk monitoring model, optimize the drug fever risk monitoring model to obtain the final drug fever risk monitoring model. The expression of this drug fever risk monitoring model is:

[0116]

[0117] Where, is the patient's drug fever percentage, is the intercept term of the drug fever risk monitoring model, is the error term of the drug fever risk monitoring model, and are the eosinophil count and C-reactive protein concentration respectively, and are the regression coefficients of the eosinophil count and C-reactive protein concentration respectively;

[0118] Input the drug fever monitoring data into the drug fever risk monitoring model, output the patient's drug fever percentage through the drug fever risk monitoring model, and integrate the patient's drug fever percentage into the patient's physiological dataset;

[0119] When the patient's drug fever percentage is less than 25%, it corresponds to a low drug fever risk; when the patient's drug fever percentage is between 25% and 65%, it corresponds to a medium drug fever risk; when the patient's drug fever percentage is greater than 65%, it corresponds to a high drug fever risk.

[0120] Preferably, for the comprehensive monitoring module, the process of constructing the drug adverse reaction monitoring model based on the evaluation result of the drug adverse reaction risk level includes:

[0121] Using the data in the training set and combining with the neural network model, take the organ monitoring data, allergy monitoring data, metabolism monitoring data and drug fever monitoring data as the input, and the patient's organ damage percentage, patient's drug allergy percentage, patient's drug metabolism index and patient's drug fever percentage as the output. Respectively learn the non-linear relationship between the organ monitoring data and the patient's organ damage percentage, the non-linear relationship between the allergy monitoring data and the patient's drug allergy percentage, the non-linear relationship between the metabolism monitoring data and the patient's drug metabolism index, and the non-linear relationship between the drug fever monitoring data and the patient's drug fever percentage, and train the drug adverse reaction monitoring model;

[0122] Input the organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data in the test set into the adverse drug reaction monitoring model respectively. Compare the errors between the output results of the adverse drug reaction monitoring model and the actual organ damage percentage of the patient, the drug allergy percentage of the patient, the drug metabolism index of the patient, and the drug fever percentage of the patient, evaluate the performance of the adverse drug reaction monitoring model, adjust the parameters of the adverse drug reaction monitoring model, optimize the adverse drug reaction monitoring model, and obtain the final adverse drug reaction monitoring model.

[0123] Preferably, the process of the comprehensive warning module combining the output results of the adverse drug reaction monitoring model and sending out corresponding warning signals includes:

[0124] Input the organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data into the adverse drug reaction monitoring model respectively. Through the adverse drug reaction monitoring model, output the organ damage percentage of the patient, the drug allergy percentage of the patient, the drug metabolism index of the patient, and the drug fever percentage of the patient respectively;

[0125] Set up a warning display screen. According to the risk levels of adverse drug reactions corresponding to the organ damage percentage of the patient, the drug allergy percentage of the patient, the drug metabolism index of the patient, and the drug fever percentage of the patient respectively, the warning display screen sends out corresponding warning signals in the form of text display.

[0126] First, through the acquisition device, collect the organ monitoring data, allergy monitoring data, metabolic monitoring data, and drug fever monitoring data of the patient, and preprocess the collected data; secondly, use the preprocessed organ monitoring data of the patient, combine with the multiple linear regression algorithm, obtain the organ damage percentage of the patient, and divide the risk level of drug-induced organ damage in the patient according to the organ damage percentage of the patient; then, analyze the preprocessed allergy monitoring data, assign the corresponding drug allergy percentage of the patient to different allergy monitoring data, and then evaluate the risk level of drug-induced allergy in the patient; then, through the preprocessed metabolic monitoring data, combine with the weight to calculate the drug metabolism index of the patient, and evaluate the risk level of drug affecting the metabolic function of the patient; then, use the preprocessed drug fever monitoring data to construct a drug fever risk monitoring model, obtain the drug fever percentage of the patient, and then evaluate the risk level of drug-induced drug fever in the patient; then, based on the evaluation results of the risk levels of adverse drug reactions, use the neural network algorithm to construct an adverse drug reaction monitoring model; finally, combine the output results of the adverse drug reaction monitoring model and send out corresponding warning signals.

[0127] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims described above.

Claims

1. A drug adverse reaction monitoring and early warning system, comprising a physiological data acquisition module, a drug risk level assessment module, a comprehensive monitoring module and a comprehensive early warning module, wherein: Each module is connected in communication, characterized by: The physiological data acquisition module is used to collect multi-dimensional patient physiological data including organ monitoring data, allergy monitoring data, metabolic monitoring data and drug thermal monitoring data, and pre-process the physiological data; The drug risk level assessment module assesses the risk level of adverse drug reactions by combining pre-processed multi-dimensional patient physiological data; The comprehensive monitoring module constructs an adverse drug reaction monitoring model based on the evaluation results of the adverse drug reaction risk level and analyzes the adverse drug reactions of patients; The comprehensive early warning module combines the output results of the adverse drug reaction monitoring model and issues corresponding early warning signals.

2. A drug adverse reaction monitoring and early warning system according to claim 1, characterized in that: The drug risk level assessment module includes an organ damage assessment unit, a drug allergy assessment unit, a drug metabolism assessment unit and a drug fever assessment unit, wherein the risk level of the adverse drug reaction includes the risk level of the drug causing organ damage to the patient, the risk level of the drug causing allergy to the patient, the risk level of the drug affecting the patient's metabolic function and the risk level of the drug causing drug fever to the patient; The organ damage assessment unit uses the preprocessed organ monitoring data to construct an organ damage risk monitoring model, obtains the patient's organ damage percentage, and then assesses the risk level of the patient's organ damage caused by the drug; The drug allergy assessment unit analyzes the pre-processed allergy monitoring data to obtain the patient's drug allergy percentage, and then assesses the risk level of drug-induced allergy in the patient; The drug metabolism assessment unit calculates the patient's drug metabolism index through the pre-processed metabolic monitoring data, and assesses the risk level of the drug affecting the patient's metabolic function; The drug heat assessment unit uses the preprocessed drug heat monitoring data to construct a drug heat risk monitoring model, obtains the patient's drug heat percentage, and then assesses the risk level of the drug causing the patient's drug heat.

3. A drug adverse reaction monitoring and early warning system according to claim 2, characterized in that: The physiological data acquisition module, the process of collecting multi-dimensional patient physiological data includes: Deploy different collection equipment to collect the patient's organ monitoring data, allergy monitoring data, metabolic monitoring data and drug thermal monitoring data, wherein the collection equipment includes a multifunctional liquid phase chip analyzer, an oxidative stress free radical detection analyzer, a histamine detector, an ELISA kit, an enzyme labeler, a constant temperature box, a C-reactive protein analyzer, a blood cell analyzer, a bilirubin analyzer and a radionuclide imaging device; The organ monitoring data include inflammatory factor concentrations and oxidative stress indicators; the allergy monitoring data include histamine concentrations, tryptase concentrations, C-reactive protein concentrations and eosinophil counts; the metabolic monitoring data include bilirubin concentrations and glomerular filtration rate; the drug thermal monitoring data include eosinophil counts and C-reactive protein concentrations; The collected organ monitoring data, allergy monitoring data, metabolic monitoring data and drug thermal monitoring data are cleaned and standardized, and timestamps are added to the organ monitoring data, allergy monitoring data, metabolic monitoring data and drug thermal monitoring data, and the collection time of the organ monitoring data, allergy monitoring data, metabolic monitoring data and drug thermal monitoring data is synchronized by adjusting the timestamps; The preprocessed organ monitoring data, allergy monitoring data, metabolic monitoring data and drug thermal monitoring data are integrated into a patient physiological data set, and the patient physiological data set is divided into a training set and a test set.

4. A drug adverse reaction monitoring and early warning system according to claim 3, characterized in that: The process of obtaining the patient's organ damage percentage in the organ damage assessment unit includes: Extract the patient's organ monitoring data from the patient's physiological data set, use the organ monitoring data in the training set, combine it with the multivariate linear regression algorithm, take the organ monitoring data as input, take the patient's organ damage percentage as output, learn the linear relationship between the organ monitoring data and the patient's organ damage percentage, and train the organ damage risk monitoring model; Input the organ monitoring data in the test set into the organ damage monitoring model, optimize the organ damage monitoring model by adjusting the intercept term and regression coefficient of the organ damage monitoring model, and obtain the final organ damage risk monitoring model; The organ monitoring data is input into the organ damage risk monitoring model, the organ damage risk monitoring model outputs the patient's organ damage percentage, and the patient's organ damage percentage is integrated into the patient's physiological data set.

5. A drug adverse reaction monitoring and early warning system according to claim 4, characterized in that: The process of the organ damage assessment unit assessing the risk level of organ damage caused by drugs includes: The risk levels of organ damage caused by the drug include low organ damage risk, medium organ damage risk and high organ damage risk; According to the patient's organ damage percentage output by the organ damage monitoring model, when the patient's organ damage percentage is less than 20%, the risk level of the corresponding drug causing the patient's organ damage is low organ damage risk; When the percentage of organ damage in patients is between 20% and 50%, the risk level of organ damage caused by the corresponding drug is medium organ damage risk; When the percentage of organ damage in a patient is greater than 50%, the risk level of organ damage caused by the corresponding drug is high risk of organ damage.

6. A drug adverse reaction monitoring and early warning system according to claim 5, characterized in that: The process of evaluating the risk level of drug allergy in patients by the drug allergy assessment unit includes: The risk levels of allergies caused by the drug include low allergy risk, moderate allergy risk and high allergy risk, and baseline ranges are set for histamine concentration, tryptase concentration, C-reactive protein concentration and eosinophil count respectively; When the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count were all within the corresponding set baseline ranges, the patient was assigned a drug allergy percentage of 10%, corresponding to a low risk of allergy; When one of the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count is not within the corresponding baseline range, the patient's drug allergy percentage is assigned to 35%, corresponding to a medium risk of allergy; When one or more of the histamine concentration, tryptase concentration, C-reactive protein concentration, and eosinophil count data is not within the corresponding baseline range, the patient's drug allergy percentage is assigned to 75%, corresponding to a high risk of allergy; By analyzing the pre-processed allergy monitoring data, the drug allergy percentage of the patient is obtained, and the drug allergy percentage of the patient is integrated into the patient's physiological data set.

7. A drug adverse reaction monitoring and early warning system according to claim 6, characterized in that: The process of the drug metabolism assessment unit assessing the risk level of a drug affecting a patient's metabolic function includes: The risk levels of the drug affecting the patient's metabolic function include low metabolic risk, medium metabolic risk and high metabolic risk; Dividing the bilirubin risk assessment range, which includes a bilirubin low risk assessment range, a bilirubin medium risk assessment range and a bilirubin high risk assessment range; assigning bilirubin risk percentages of 10%, 30% and 65% to the bilirubin low risk assessment range, the bilirubin medium risk assessment range and the bilirubin high risk assessment range, respectively; Dividing the GFR risk assessment range, which consists of a GFR low risk assessment range, a GFR medium risk assessment range, and a GFR high risk assessment range; The low GFR risk assessment range, the medium GFR risk assessment range, and the high GFR risk assessment range were assigned GFR risk percentages of 10%, 45%, and 90%, respectively; Assign weights to the corresponding bilirubin risk percentage and glomerular filtration rate risk percentage, calculate the patient's drug metabolism index using the assigned weights, and integrate the calculated patient's drug metabolism index into the patient's physiological data set; According to the patient's drug metabolism index, when the patient's drug metabolism index is lower than 0.2, the risk level of the corresponding drug affecting the patient's metabolic function is low metabolic risk; when the patient's drug metabolism index is between 0.2 and 0.5, the risk level of the corresponding drug affecting the patient's metabolic function is medium metabolic risk; when the patient's drug metabolism index is higher than 0.5, the risk level of the corresponding drug affecting the patient's metabolic function is high metabolic risk.

8. A drug adverse reaction monitoring and early warning system according to claim 7, characterized in that: The process of the drug heat assessment unit assessing the risk level of drug-induced drug fever in patients includes: The risk levels of drug fever caused by the drug include low risk of drug fever, medium risk of drug fever and high risk of drug fever; Extract the drug thermal monitoring data from the patient's physiological data set, use the drug thermal monitoring data in the training set, combine it with the multivariate linear regression algorithm, take the drug thermal monitoring data as input, take the patient's drug thermal percentage as output, learn the linear relationship between the drug thermal monitoring data and the patient's drug thermal percentage, and train the drug thermal risk monitoring model; The drug heat monitoring data in the test set is input into the drug heat risk monitoring model, and the drug heat risk monitoring model is optimized by adjusting the intercept term and regression coefficient of the drug heat risk monitoring model to obtain the final drug heat risk monitoring model; Inputting drug thermal monitoring data into a drug thermal risk monitoring model, outputting the patient drug thermal percentage through the drug thermal risk monitoring model, and integrating the patient drug thermal percentage into the patient physiological data set; When the patient's drug fever percentage is lower than 25%, it corresponds to a low risk of drug fever; when the patient's drug fever percentage is between 25% and 65%, it corresponds to a medium risk of drug fever; when the patient's drug fever percentage is greater than 65%, it corresponds to a high risk of drug fever.

9. A drug adverse reaction monitoring and early warning system according to claim 8, characterized in that: The comprehensive monitoring module, based on the evaluation result of the adverse drug reaction risk level, constructs a process of the adverse drug reaction monitoring model, including: Using the data in the training set, combined with the neural network model, taking the organ monitoring data, allergy monitoring data, metabolic monitoring data and drug heat monitoring data as input, taking the patient's organ damage percentage, the patient's drug allergy percentage, the patient's drug metabolism index and the patient's drug heat percentage as output, respectively learning the nonlinear relationship between the organ monitoring data and the patient's organ damage percentage, the nonlinear relationship between the allergy monitoring data and the patient's drug allergy percentage, the nonlinear relationship between the metabolic monitoring data and the patient's drug metabolism index, and the nonlinear relationship between the drug heat monitoring data and the patient's drug heat percentage, to train the drug adverse reaction monitoring model; The organ monitoring data, allergy monitoring data, metabolic monitoring data and drug thermal monitoring data in the test set are respectively input into the adverse drug reaction monitoring model, and the errors between the output results of the adverse drug reaction monitoring model and the actual patient organ damage percentage, patient drug allergy percentage, patient drug metabolic index and patient drug thermal percentage are compared to evaluate the performance of the adverse drug reaction monitoring model, adjust the adverse drug reaction monitoring model parameters, optimize the adverse drug reaction monitoring model, and obtain the final adverse drug reaction monitoring model.

10. A drug adverse reaction monitoring and early warning system according to claim 9, characterized in that: The process of the comprehensive early warning module combining the output results of the adverse drug reaction monitoring model and issuing corresponding early warning signals includes: The organ monitoring data, allergy monitoring data, metabolic monitoring data and drug heat monitoring data are respectively input into the adverse drug reaction monitoring model, and the adverse drug reaction monitoring model outputs the patient's organ damage percentage, the patient's drug allergy percentage, the patient's drug metabolism index and the patient's drug heat percentage; An early warning display screen is set up to send out corresponding early warning signals through text display according to the risk levels of adverse drug reactions corresponding to the patient's organ damage percentage, drug allergy percentage, drug metabolism index and drug heat percentage.

Images