Anti-radiation drug dose data acquisition self-adaptive dose control system

Abstract

The invention discloses an anti-radiation drug dose data acquisition self-adaptive dose control system. The system comprises a physiological parameter real-time acquisition module, a transfer environment data capture module, a multi-scale pharmacokinetic modeling module, a group Bayesian optimization calculation module, an anti-radiation drug dose adjustment module and a system parameter dynamic updating module. All the modules work cooperatively, the wounded physiological data and the transfer environment data are collected and input into the multi-scale pharmacokinetic modeling module to construct a cross-scale drug metabolism model, the group Bayesian optimization calculation module combines group data optimization calculation, and the anti-radiation drug dosage adjusting module adjusts the dosage according to the optimization calculation. And the system parameter dynamic updating module continuously corrects the parameters. According to the system, cross-scale analysis, group and individual data fusion and dynamic adaptation are achieved, it is ensured that the medicine dosage is accurate and appropriate in the transfer process, and a guarantee is provided for wounded treatment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of anti-radiation drug dose data acquisition, in particular to an anti-radiation drug dose data acquisition adaptive dose control system. BACKGROUND

[0002] In nuclear accidents, radiation leaks and other emergencies, radiation-injured patients need to be transported to professional medical institutions for treatment in a timely manner, and the anti-radiation drug dose control during transportation is crucial to ensure treatment effectiveness and reduce adverse reactions. The transportation environment is complex and variable, and the physiological state of the injured person changes dynamically over time. The traditional fixed-dose administration method is difficult to adapt to individual differences and environmental fluctuations, which may result in insufficient or excessive drug. Therefore, a self-adaptive control system that can collect physiological and environmental data in real time and dynamically adjust the drug dose is needed to achieve precise regulation of anti-radiation drug dose during transportation through multi-scale modeling and group optimization algorithm.

[0003] The existing technology has two obvious shortcomings: first, the lack of cross-scale integration capability, most systems only adjust the dose based on blood or tissue single-scale drug concentration data, without considering the interaction between drugs and targets at the cellular level and the differences in tissue metabolism, resulting in dose calculation ignoring the microscopic drug effect mechanism and difficulty in reflecting the true drug efficacy; second, the group data and individual dynamic characteristics are not effectively integrated, traditional methods either rely on fixed group parameters leading to poor individual adaptability or only adjust based on single patient historical data, unable to use group pharmacokinetic rules to correct individual differences, and prone to dose adjustment lag or deviation when the transportation environment changes suddenly. SUMMARY

[0004] In order to overcome the shortcomings and deficiencies of the prior art, the present application provides an anti-radiation drug dose data acquisition adaptive dose control system.

[0005] The technical solution adopted by the present application is an anti-radiation drug dose data acquisition adaptive dose control system, comprising: a physiological parameter real-time acquisition module, a transportation environment data capture module, a multi-scale pharmacokinetic modeling module, a group Bayesian optimization calculation module, an anti-radiation drug dose adjustment module, and a system parameter dynamic updating module. The output end of the physiological parameter real-time acquisition module is connected with the first input end of the multi-scale pharmacokinetics modeling module, the output end of the transport environment data capture module is connected with the second input end of the multi-scale pharmacokinetics modeling module, the output end of the multi-scale pharmacokinetics modeling module is connected with the input end of the population Bayesian optimization calculation module, the output end of the population Bayesian optimization calculation module is connected with the input end of the anti-radiation drug dose adjustment module, the feedback end of the anti-radiation drug dose adjustment module is connected with the input end of the system parameter dynamic updating module, and the output end of the system parameter dynamic updating module is connected with the parameter configuration end of the physiological parameter real-time acquisition module, the transport environment data capture module, the multi-scale pharmacokinetics modeling module, the population Bayesian optimization calculation module and the anti-radiation drug dose adjustment module respectively; the physiological parameter real-time acquisition module acquires the heart rate, blood pressure, blood drug concentration and radiation exposure dose data of the wounded; the transport environment data capture module acquires the vibration frequency, temperature, humidity and altitude data of the transport carrier; the multi-scale pharmacokinetics modeling module constructs a drug metabolism kinetics model across the tissue and cell scales based on the above physiological parameters and environmental data; the population Bayesian optimization calculation module performs optimization calculation by combining the model output results with population pharmacokinetics data; the anti-radiation drug dose adjustment module adjusts the drug infusion rate and single dose according to the optimization calculation results; and the system parameter dynamic updating module updates the calculation parameters of each module according to the physiological parameter changes after dose adjustment.

[0006] Further, when the multi-scale pharmacokinetics modeling module and the population Bayesian optimization calculation module work cooperatively, the following model formula is used to describe the relationship between the metabolism rate of the drug in the transport process and the dose: wherein, is the concentration of the anti-radiation drug in the blood of the wounded at time t, is the transport time, is the drug absorption rate constant, is the drug elimination rate constant, is the radiation intensity of the transport environment, is the effective radiation protection threshold of the drug, is the internal space volume of the transport carrier, is the environmental humidity, When the population Bayesian optimization calculation module processes the confounding factors, the following model formula is used: wherein, is the drug metabolism parameter of the jth wounded in the ith population, is the mean value of the population parameter, is the population level random effect, which is subject to standard normal distribution, is the standard deviation of the population parameter, is the individual level random effect, which is subject to normal distribution with a mean value of 0.

[0007] Furthermore, the anti-radiation drug dosage adjustment module calculates the real-time drug dosage based on the drug distribution coefficient output by the multi-scale pharmacokinetic modeling module and the dosage optimization value obtained by the population Bayesian optimization calculation module, using the following model formula: ,in, This is a single-dose administration. To achieve the target blood drug concentration, This represents the current blood drug concentration. The apparent volume of distribution of the drug. The altitude influence coefficient is... This is the current altitude. For drug bioavailability, This is the temperature influence coefficient. The ambient temperature during transport; The transport environment data acquisition module incorporates vibration frequency data into the dose adjustment correction, using the following formula: ,in, This is the vibration-corrected dosage. This is the vibration influence coefficient. The vibration frequency of the transport carrier.

[0008] Furthermore, the multi-scale pharmacokinetic modeling module describes the drug-target binding process at the cellular scale, with the following model formula: ,in, This refers to the concentration of the drug-target complex. This refers to the extracellular drug concentration. The total concentration at the target site. The binding rate constant is The dissociation rate constant is Radiation dose influencing factor, The cumulative radiation exposure dose for the injured; The population Bayesian optimization calculation module uses the complex concentration data to estimate population parameters. The estimation formula is as follows: ,in, Parameters for given observation data y The posterior probability density, Let be the likelihood function. For parameters The prior distribution, For hyperparameters The marginal distribution.

[0009] Furthermore, based on the execution results of the anti-radiation drug dosage adjustment module, the system parameter dynamic update module updates the scale transformation coefficients of the multi-scale pharmacokinetics modeling module using the following model formula: ,in, These are the updated scale transformation coefficients. is the initial scale conversion coefficient, is the update rate constant, is the reference blood drug concentration, is the integral of the difference between the blood drug concentration and the reference concentration from 0 to t; The prior distribution parameters of the population Bayesian optimization calculation module are updated by the following formula: wherein, is the updated population parameter mean, is the initial population parameter mean, is the weight coefficient, is the sample size, is the ith observation value, is the ith predicted value.

[0010] Further, the data transmission between the physiological parameter real-time acquisition module and the multi-scale pharmacokinetic modeling module adopts a dynamic sampling interval, and the interval time calculation formula is: wherein, is the current sampling interval, is the basic sampling interval, is the sampling adjustment coefficient, is the absolute value of the blood drug concentration change rate; The infusion rate adjustment formula of the anti-radiation drug dose adjustment module is: wherein, is the current drug infusion rate, is the basic infusion rate, is the maximum safe blood drug concentration, is the minimum effective blood drug concentration, is the humidity influence coefficient, is the environmental humidity.

[0011] Further, the multi-scale pharmacokinetic modeling module comprises: a tissue scale drug distribution calculation unit, a cell scale drug transport simulation unit, and a cross-scale parameter mapping unit; the tissue scale drug distribution calculation unit receives blood flow rate and vascular density data output by the physiological parameter real-time acquisition module, combines temperature data output by the transport environment data capture module, and obtains initial distribution of drugs in each tissue by calculating the product of blood perfusion rate and drug distribution coefficient of different tissues and organs; the cell scale drug transport simulation unit simulates the process of drug molecules entering the cell interior by passive diffusion and active transport and combining with the target based on cell membrane permeability and intracellular target concentration data, and outputs the dynamic change curve of drug-target complex; the cross-scale parameter mapping unit converts the drug concentration data calculated at the tissue scale into the boundary condition of the cell scale model, and feeds back the drug metabolism rate simulated at the cell scale to the tissue scale model for coupled calculation of the two scale models.

[0012] Further, the population Bayesian optimization calculation module comprises: a population data hierarchical processing unit, a prior distribution construction unit, a posterior probability calculation unit, and an optimization parameter extraction unit; the population data hierarchical processing unit stratifies the output data of the multi-scale pharmacokinetic modeling module received according to the age, weight, and radiation exposure of the wounded, and calculates the mean, standard deviation, and coefficient of variation of each stratified data; the prior distribution construction unit sets the prior probability density function of the drug metabolism parameters in accordance with the characteristics of the stratified data, which conforms to the normal-gamma distribution, wherein the mean parameter is associated with the stratified mean, and the variance parameter is associated with the stratified standard deviation; the posterior probability calculation unit calculates the posterior probability distribution of each drug metabolism parameter by using the Markov chain Monte Carlo method in combination with the real-time collected blood concentration observation data; and the optimization parameter extraction unit selects the parameter interval corresponding to the cumulative probability of 95% from the posterior probability distribution, and takes the interval midpoint as the optimal parameter value of the population Bayesian optimization calculation.

[0013] Further, the anti-radiation drug dose adjustment module comprises: a dose requirement calculation unit, a drug administration mode selection unit, and an infusion precision control unit; the dose requirement calculation unit receives the optimal parameters output by the population Bayesian optimization calculation module, combines the current blood concentration and radiation exposure dose data output by the physiological parameter real-time acquisition module, calculates the difference between the target blood concentration and the current blood concentration, and the drug increment required for radiation protection, and obtains the total dose requirement value; the drug administration mode selection unit selects the continuous infusion mode when the vibration frequency is lower than the threshold value according to the vibration frequency and the degree of jolt data output by the transport environment data capture module, and switches to the intermittent pulse drug administration mode when the vibration frequency is higher than the threshold value, while adjusting the pulse interval and the single pulse dose; and the infusion precision control unit compensates for the infusion flow fluctuation caused by the vibration of the transport carrier by real-time monitoring of the pressure change of the infusion pipeline, and corrects the deviation between the actual drug administration amount and the theoretical drug administration amount within every 10 seconds.

[0014] Beneficial Effects: This invention proposes an adaptive dose control system for anti-radiation drug dosage data acquisition. Addressing the lack of cross-scale integration capabilities in existing technologies, the system acquires data such as heart rate and blood pressure from the injured person through a real-time physiological parameter acquisition module, and collects information such as vibration frequency and temperature from a transport environment data capture module. This comprehensive input into a multi-scale pharmacokinetic modeling module constructs a pharmacokinetic model spanning tissue and cell scales, enabling drug metabolism analysis from macroscopic to microscopic levels, more realistically reflecting drug efficacy, and overcoming the limitations of single-scale analysis. Addressing the deficiency in existing technologies that fail to effectively integrate population data with individual dynamic characteristics, the population Bayesian optimization calculation module utilizes multi-scale model results combined with population pharmacokinetic data for optimization calculation. The anti-radiation drug dosage adjustment module adjusts the infusion rate and dosage based on the optimization results. The system parameter dynamic update module continuously corrects the calculation parameters of each module based on the adjusted physiological parameter changes. This approach draws on both population patterns and individual real-time states, avoiding problems of poor individual adaptability or adjustment lag, ensuring that the anti-radiation drug dosage remains accurate and appropriate during transport, providing reliable drug dosage control for the injured person. Attached Figure Description

[0015] Figure 1 This is a system unit composition diagram of the present invention; Figure 2 This is a flowchart of the system operation of the present invention. Detailed Implementation

[0016] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] like Figure 1 As shown, an adaptive dose control system for anti-radiation drug dose data acquisition includes: a real-time physiological parameter acquisition module, a transport environment data capture module, a multi-scale pharmacokinetics modeling module, a population Bayesian optimization calculation module, an anti-radiation drug dose adjustment module, and a system parameter dynamic update module. The output end of the physiological parameter real-time acquisition module is connected with the first input end of the multi-scale pharmacokinetics modeling module, the output end of the transport environment data capturing module is connected with the second input end of the multi-scale pharmacokinetics modeling module, the output end of the multi-scale pharmacokinetics modeling module is connected with the input end of the population Bayesian optimization calculation module, the output end of the population Bayesian optimization calculation module is connected with the input end of the anti-radiation drug dose adjustment module, the feedback end of the anti-radiation drug dose adjustment module is connected with the input end of the system parameter dynamic updating module, and the output end of the system parameter dynamic updating module is respectively connected with the parameter configuration end of the physiological parameter real-time acquisition module, the transport environment data capturing module, the multi-scale pharmacokinetics modeling module, the population Bayesian optimization calculation module and the anti-radiation drug dose adjustment module. The physiological parameter real-time acquisition module acquires the heart rate, blood pressure, blood drug concentration and radiation exposure dose data of the wounded person. The transport environment data capturing module acquires the vibration frequency, temperature, humidity and altitude data of the transport carrier. The multi-scale pharmacokinetics modeling module constructs a drug metabolism dynamics model across the tissue and cell scales based on the above physiological parameters and environmental data. The population Bayesian optimization calculation module performs optimization calculation by combining the model output result with the population pharmacokinetics data. The anti-radiation drug dose adjustment module adjusts the drug infusion rate and single dose according to the optimization calculation result. The system parameter dynamic updating module updates the calculation parameters of each module according to the physiological parameter changes after dose adjustment.

[0018] Specifically, the physiological parameter real-time acquisition module is the core component of the system to obtain the basic physiological information of the wounded, and its technical parameters cover key indicators such as acquisition frequency, data accuracy, and monitoring range. The acquisition frequency is set to 1 time per second, which can ensure the capture of subtle fluctuations in parameters such as heart rate and blood pressure; the data accuracy is controlled within ±2%, ensuring the reliability of data such as blood drug concentration and radiation exposure dose; the monitoring range covers heart rate 60-180 times / minute, blood pressure 60-180 mmHg, blood drug concentration 0-50 mg / L, and radiation exposure dose 0-1000 mSv, which can adapt to the physiological state of wounded with different injuries. The significance of this module is to provide real-time and accurate basic data for subsequent drug metabolism modeling and dose optimization, which is the premise of realizing adaptive dose control. The specific implementation process of this module is as follows: the heart rate signal is collected by pasting electrode pieces on the body surface of the wounded, and the non-invasive blood pressure monitor measures the blood pressure value every 10 seconds and records it; the venous blood sample collection device extracts blood samples every 5 minutes, and the blood drug concentration data is obtained by analyzing the portable blood drug concentration detector; at the same time, the radiation dose meter worn on the chest of the wounded continuously monitors the radiation exposure dose, and updates the data every 30 seconds. All collected data is converted to digital after analog-digital conversion and stored in the cache area built-in the module, the cache area capacity is 10 GB, which can store 72 hours of data continuously. The data is transmitted in real time to the multi-scale pharmacokinetic modeling module through wired transmission, the transmission rate is 10 Mbps, which ensures the real-time and integrity of the data, and encryption algorithm is used to ensure data security and prevent information leakage during transmission.

[0019] Specifically, the transport environment data capture module is mainly used to collect external environmental factor data that affects drug metabolism. The technical parameters include sensor type, measurement range, response time, etc. The temperature and humidity sensor used has a measurement range of -10℃-50℃ for temperature and 20%-95% for humidity, with measurement accuracies of ±0.5℃ and ±3%, respectively. The vibration frequency sensor can monitor vibration signals of 0-50Hz, with a measurement error of not more than ±0.1Hz. The altimeter has a measurement range of -500m-5000m, with an accuracy of ±5m. The significance of this module is to provide environmental impact parameters for the drug metabolism model, so that the model can more accurately reflect the influence of complex environments on drug metabolism during transportation and improve the adaptability of dose adjustment. The specific implementation process of this module is as follows: multiple sensors are installed at different positions of the transport carrier. The temperature and humidity sensors are installed near the patient's head and the drug storage place, and data is collected every 2 minutes. The vibration frequency sensor is fixed at the bottom of the transport carrier, and the vibration frequency value is recorded every 0.5 seconds. The altimeter is installed at the top of the transport carrier, and the altitude data is updated every 1 minute. The analog signals collected by the sensors are processed by the signal conditioning circuit and converted into digital signals, which are sent to the data processing unit through the wireless transmission module. The wireless transmission distance can reach 50 meters, and the transmission rate is 2Mbps. The data processing unit checks and filters the received data, removes abnormal data, and sends the effective data to the multi-scale pharmacokinetic modeling module in a data packet of 1KB each time, ensuring the efficiency and stability of data transmission.

[0020] Specifically, the multi-scale pharmacokinetic modeling module is the key to realize the precise analysis of drug metabolism, and its technical parameters involve model scale range, calculation accuracy, data processing capacity, etc. The model covers tissue scale (such as heart, liver, kidney, etc. main organs) and cell scale, the spatial resolution of tissue scale is 1mm, and the spatial resolution of cell scale is 1μm; the calculation accuracy is controlled within ±5%, which can accurately simulate the absorption, distribution, metabolism and excretion process of drugs in vivo; the data processing capacity is 100MB per second, which can meet the demand of real-time modeling. The significance of this module is to break through the limitations of traditional single-scale model, and comprehensively reflect the metabolism law of drugs from macro and micro levels, and provide scientific theoretical basis for dose optimization. The specific implementation process of this module is as follows: after receiving the data sent by the physiological parameter real-time acquisition module and the transport environment data capture module, first, the data is format converted and standardized processed to meet the input requirements of the model. Then, based on the tissue scale, the drug distribution model is constructed, according to the blood perfusion rate of each organ, drug distribution coefficient and other parameters, the concentration distribution of drug in different tissues is calculated, and the drug concentration data of tissue scale is updated every 5 minutes. At the same time, on the cell scale, according to the cell membrane permeability, intracellular target concentration and other data, the process of drug molecules entering cells and binding with target is simulated, and the drug-target complex concentration data of cell scale is updated every 1 minute. Through the cross-scale parameter mapping, the drug concentration data of tissue scale is converted into the boundary conditions of cell scale model, and the metabolism rate of cell scale is fed back to the tissue scale model, realizing the coupled calculation of two scale models. The calculation results are sent to the population Bayesian optimization calculation module in real time, and the data amount of each sending is 500KB.

[0021] Specifically, the population Bayesian optimization calculation module plays an important role in dose optimization, and its technical parameters include population data scale, optimization iteration number, and calculation efficiency. The population data scale covers more than 1000 cases of pharmacokinetic data of wounded personnel of different genders, ages, and weights, ensuring the representativeness of population regularity. The optimization iteration number is set to 50-100 times, which can ensure the optimization accuracy while improving the calculation efficiency. The calculation efficiency is less than 10 seconds per iteration, which can meet the real-time optimization requirements. The significance of this module lies in making full use of population pharmacokinetic data, combining individual real-time physiological and drug metabolism data, and realizing individualized optimization of the dose to improve the accuracy and adaptability of dose adjustment. The specific implementation process of this module is as follows: after receiving the drug metabolism data output by the multi-scale pharmacokinetic modeling module, the pre-set population pharmacokinetic database is loaded, which contains drug metabolism parameters under different radiation exposure and different physiological conditions. Then, the individual real-time data is matched with the population data, and a population subset similar to the current characteristics of the wounded personnel is selected, with a subset size of 50-100 cases. Based on the Bayesian theorem, the objective function is constructed, which takes the maximization of drug efficacy and the minimization of adverse reactions as the optimization objective, and determines the prior distribution combined with the population subset data and the individual real-time data. The posterior distribution is updated through iterative calculation. In each iteration process, the next dose point to be evaluated is selected according to the current posterior distribution, the objective function value of the dose point is calculated, and the posterior distribution is updated. After 50-100 iterations, the optimal drug dose parameters are obtained, including the single dose and the infusion rate, and the calculation results are sent to the anti-radiation drug dose adjustment module. Each time the data sent contains 10 parameters, and the data volume is 100B.

[0022] Specifically, the anti-radiation drug dose adjustment module is the core component of the dose adjustment, and its technical parameters include the adjustment range, the adjustment accuracy, the response time, etc. The adjustment range of the single dose is 1-50 mg, and the adjustment accuracy is ±0.1 mg; the adjustment range of the infusion rate is 0.1-5 mg / h, and the adjustment accuracy is ±0.01 mg / h; the response time is not more than 2 seconds, which can quickly respond to the optimization calculation results. The significance of the module lies in converting the optimized dose parameters into actual drug administration operations, realizing the precise adjustment of the drug dose, and ensuring that the drug is always at an appropriate concentration level during the transportation. The specific implementation process of the module is as follows: after receiving the optimal dose parameters sent by the group Bayesian optimization calculation module, the parameters are first analyzed and verified to confirm the rationality and effectiveness of the parameters. Then, according to the single dose and the infusion rate parameters, the driving motor of the infusion pump is controlled to work. The infusion pump is driven by a stepping motor, and the stepping angle is 1.8 degrees. The drug dose is accurately controlled by controlling the number of motor rotations, and each step corresponds to a drug dose of 0.001 mg. In the process of drug administration, the pressure and flow of the infusion pipeline are monitored in real time. The pressure monitoring range is 0-50 kPa, and the accuracy is ±0.5 kPa; the flow monitoring range is 0-10 mg / h, and the accuracy is ±0.05 mg / h. When the pressure or flow is abnormal, the drug administration is immediately stopped, and an alarm signal is sent. At the same time, the abnormal information is fed back to the system parameter dynamic updating module. After the drug administration is completed, the actual drug dose and the drug administration time are recorded, and the data are stored in the local storage of the module. The storage capacity is 1 GB, and the drug administration records of 30 days can be stored.

[0023] Specifically, the system parameter dynamic updating module is an important part of ensuring the long-term stable operation of the system. Its technical parameters include parameter updating frequency, updating accuracy, data storage capacity, etc. The parameter updating frequency is once every 30 minutes, which can reflect the changes in the system operation state in a timely manner; the updating accuracy is controlled within ±1%, ensuring the accuracy of parameter adjustment; the data storage capacity is 100 GB, which can store the system parameter updating records for 1 year. The significance of this module lies in continuously updating the calculation parameters of each module, so that the system can adapt to the dynamic changes of the physiological state of the wounded and the transfer environment, and continuously improve the accuracy and reliability of dose control. The specific implementation process of this module is as follows: after receiving the drug administration data and physiological parameter change data feedback from the anti-radiation drug dose adjustment module, first analyze and process the data, and calculate the deviation value between the actual administration effect and the expected effect. According to the deviation value, determine the type and range of parameters that need to be updated, which involves the scale conversion coefficient of the multi-scale pharmacokinetic modeling module, the prior distribution parameter of the population Bayesian optimization calculation module, and the sampling frequency of the physiological parameter real-time acquisition module, etc. Then, use the adaptive updating algorithm to adjust these parameters, and the adjustment amplitude is determined according to the size of the deviation value. The larger the deviation value, the larger the adjustment amplitude, but the maximum adjustment amplitude does not exceed 10% of the original parameter value. After the parameter updating is completed, the new parameter value is sent to the corresponding module, and each module updates its own parameter configuration immediately after receiving the parameter, and restarts the relevant calculation process. At the same time, record the time, content and reason of parameter updating, and store it in the local database for subsequent system maintenance and performance analysis. Each parameter updating process takes no more than 30 seconds, ensuring the continuous operation of the system is not affected.

[0024] Preferably, when the multi-scale pharmacokinetic modeling module and the population Bayesian optimization calculation module work together, the following model formula is used to describe the relationship between the metabolism rate of the drug in the transfer process and the dose: wherein, Ct is the concentration of the anti-radiation drug in the blood of the wounded at time t, T is the transfer time, Ka is the drug absorption rate constant, Kel is the drug elimination rate constant, I is the radiation intensity of the transfer environment, Cmax is the effective radiation protection threshold of the drug, V is the internal space volume of the transfer vehicle, RH is the environmental humidity; When dealing with confounding factors, the population Bayesian optimization calculation module uses the following model formula: wherein, Pi,j is the drug metabolism parameter of the jth wounded in the ith population, μ is the mean value of the population parameter, b is the population-level random effect, which follows a standard normal distribution, is a standard deviation of population parameters, is an individual level random effect, following a normal distribution with mean 0.

[0025] Specifically, the multi-scale pharmacokinetic modeling module and the population Bayesian optimization calculation module work together, and the parameters involve the calculation accuracy of drug metabolism rate, the coverage range of population parameters, etc. The calculation accuracy of drug metabolism rate is controlled within ±4%, which can accurately reflect the metabolism change of the drug under different transport environments; the population parameters cover the data of wounded persons of different ages (18-65 years old) and weights (50-100 kg), ensuring the wide applicability of population regularity. Through collaborative modeling and optimization, the relationship between drug metabolism and dose is described more accurately, while the confounding factors are effectively handled, and the accuracy of dose calculation is improved. The implementation is as follows: after receiving the blood drug concentration, radiation exposure dose data and radiation intensity, spatial volume, humidity data from the physiological parameter real-time acquisition module and the transport environment data capture module, the multi-scale pharmacokinetic modeling module combines the drug absorption and elimination rate constants to calculate the drug metabolism rate at different times, wherein the drug absorption rate constant is in the range of 0.5-2.0 L / h, the elimination rate constant is in the range of , and the radiation protection threshold is set to 50-200 mSv / h. The population Bayesian optimization calculation module loads the population pharmacokinetic database, which contains drug metabolism parameters of more than 1000 different populations. The mean of the population parameters fluctuates between 0.8 and 1.2 according to different population characteristics, the standard deviation is controlled within the range of 0.1-0.3, and the standard deviation of the individual level random effect is not more than 0.05. By matching individual data with population data, 50-80 similar population subsets are selected, the posterior probability is calculated using Bayesian theorem, the drug metabolism parameters of the individual are determined, and more accurate basis is provided for dose adjustment. The calculation result is updated every 10 minutes to ensure real-time adaptation to the environmental changes and physiological state fluctuations of the wounded during transport.

[0026] Preferably, the anti-radiation drug dose adjustment module calculates the real-time drug dose based on the drug distribution coefficient output by the multi-scale pharmacokinetic modeling module and the dose optimization value obtained by the population Bayesian optimization calculation module using the following model formula: , wherein, is a single dose, is a target blood drug concentration, is a current blood drug concentration, is an apparent distribution volume of the drug, is an altitude influence coefficient, is a current altitude, is a drug bioavailability, is a temperature influence coefficient, The ambient temperature during transport; The transport environment data acquisition module incorporates vibration frequency data into the dose adjustment correction, using the following formula: ,in, This is the vibration-corrected dosage. This is the vibration influence coefficient. The vibration frequency of the transport carrier.

[0027] Specifically, the radiation-resistant drug dosage adjustment module and the transport environment data capture module work synergistically, with parameters including the calculation accuracy of the drug dosage and the range of vibration correction coefficients. The drug dosage calculation accuracy can reach ±0.05mg, meeting the dosage adjustment needs under different injury conditions; the vibration correction coefficient varies depending on the type of transport carrier. Adjustments are made to ensure dosage accuracy under vibration. Combining multi-scale modeling and population optimization results, and considering factors such as altitude, temperature, and vibration in the transport environment, precise adjustment and dynamic correction of drug dosage are achieved, improving the adaptability and reliability of dosage control. The implementation method is as follows: The anti-radiation drug dosage adjustment module first receives parameters such as target blood drug concentration, apparent volume of distribution, and bioavailability output by the population Bayesian optimization calculation module. The target blood drug concentration range is 5-30 mg / L, the apparent volume of distribution is calculated based on body weight between 0.1-0.5 L / kg, and the bioavailability is set to 0.8-0.95. The module combines the current blood drug concentration data from the real-time physiological parameter acquisition module with altitude (range -500-5000m) and temperature (range -10-50℃) data from the transport environment data capture module. The altitude influence coefficient is set to... Temperature influence coefficient value The single-dose dosage is calculated. Then, based on the vibration frequency (range 0-50Hz) data from the transport environment data capture module, the dosage is corrected using a vibration correction coefficient to obtain the vibration-corrected dosage. During administration, the infusion pump's infusion rate is adjusted according to the corrected dosage, ranging from 0.1-5 mg / h. The flow rate and pressure of the infusion tubing are monitored every second to ensure that the actual dosage deviates from the calculated value by no more than ±0.01 mg. Dosage data is fed back to the system parameter dynamic update module every 5 minutes for subsequent parameter optimization.

[0028] Preferably, the multi-scale pharmacokinetic modeling module describes the drug-target binding process at the cellular scale, and the model formula is as follows: ,in, This refers to the concentration of the drug-target complex. This refers to the extracellular drug concentration. The total concentration at the target site. The binding rate constant, is the dissociation rate constant, is the radiation dose impact factor, is the cumulative radiation exposure dose of the wounded; The population Bayesian optimization calculation module uses the complex concentration data to estimate the population parameters, and the estimation formula is: wherein, is the posterior probability density of the parameter when given observation data y, is the likelihood function, is the prior distribution of the parameter , and is the marginal distribution of the hyperparameter .

[0029] Specifically, the cell scale analysis of the multi-scale pharmacokinetic modeling module is combined with the parameter estimation of the population Bayesian optimization calculation module, and the parameters involve the calculation accuracy of the drug-target complex concentration, the estimation error of the posterior probability distribution, etc. The calculation accuracy of the drug-target complex concentration is controlled within ±0.01 mol / L, which can accurately reflect the binding state of the drug and the target; the estimation error of the posterior probability distribution is not more than ±5%, which ensures the reliability of the parameter estimation. From the cell level, the drug action mechanism is analyzed in depth, and at the same time, the population data is used for parameter estimation, which provides a more microscopic and more accurate theoretical basis for dose optimization, and improves the scientificity and pertinence of dose adjustment. The implementation mode is that the multi-scale pharmacokinetic modeling module, on the cell scale, collects the extracellular drug concentration and target total concentration data in real time according to the physiological parameters of the real-time acquisition module, wherein the extracellular drug concentration ranges from 1 to 10 mol / L, the target total concentration is 0.1-1 mol / L, combined with the cell membrane permeability and the association and dissociation rate constants (the association rate constant is , and the dissociation rate constant is ), and the radiation dose impact factor (range ) and cumulative radiation exposure dose of the wounded (range 0-1000 mSv), the drug-target complex concentration is calculated, and the calculation result is updated every 30 seconds. After the population Bayesian optimization calculation module receives the complex concentration data at the cell scale, it loads a population observation data set containing more than 1000 cases, which covers drug metabolism parameters under different radiation exposure levels and different physiological states. Based on the likelihood function, prior distribution (dependent on hyperparameters, the range of hyperparameters is adjusted between 0.5-2.0 according to the characteristics of the population), and the marginal distribution of hyperparameters, the Markov chain Monte Carlo method is used for 1000-2000 times of iteration calculation to obtain the posterior probability density of the drug metabolism parameters. The parameter interval with a cumulative probability of 95% is extracted from the posterior probability distribution, the interval width is controlled within ±10% of the parameter mean, and the midpoint of the interval is taken as the optimal parameter value. The optimal parameter is sent to the anti-radiation drug dose adjustment module every 5 minutes to guide dose adjustment and ensure that dose adjustment can accurately adapt to changes in drug action at the cell level.

[0030] Preferably, the system parameter dynamic updating module updates the scale conversion coefficient of the multi-scale pharmacokinetic modeling module according to the execution result of the anti-radiation drug dose adjustment module through the following model formula: wherein, is the updated scale conversion coefficient, is the initial scale conversion coefficient, is the update rate constant, is the reference blood drug concentration, is the integral of the difference between the blood drug concentration at time t and the reference concentration from 0 to t; The prior distribution parameter of the population Bayesian optimization calculation module is updated by the following formula: wherein, is the updated population parameter mean, is the initial population parameter mean, is the weight coefficient, is the sample size, is the ith observation value, is the ith predicted value.

[0031] Specifically, the system parameter dynamic updating module updates the parameters of the multi-scale pharmacokinetic modeling module and the population Bayesian optimization calculation module, including the update precision of the scale conversion coefficient, the adjustment range of the population parameter mean, etc. The update precision of the scale conversion coefficient is ±0.001, which can ensure the coupling accuracy of the cross-scale model; the adjustment range of the population parameter mean is not more than ±10% of the initial value, which ensures the stability of the population law. By continuously updating the model parameters, the system can dynamically adapt to the changes in the drug metabolism process, improving the long-term accuracy and reliability of dose control. The implementation is as follows: after receiving the actual blood drug concentration data fed back by the anti-radiation drug dose adjustment module and the reference blood drug concentration (range 5-25 mg / L) data of the multi-scale pharmacokinetic modeling module, the system parameter dynamic updating module calculates the integral of the difference between the blood drug concentration from 0 to the current time and the reference concentration, and the integral result is used to adjust the scale conversion coefficient. The update rate constant is , the initial scale conversion coefficient is set between 0.5-2.0 according to the tissue and cell type, and the updated scale conversion coefficient is calculated by an exponential function, updated every 30 minutes. For the prior distribution parameters of the population Bayesian optimization calculation module, the system parameter dynamic updating module calculates the updated population parameter mean using a weight coefficient (range 0.05-0.2) according to the sample size (50-100 cases) and the difference between the observed value and the predicted value. The weight coefficient is adjusted according to the similarity of the sample, and the higher the similarity, the larger the weight coefficient. After each parameter update, the new parameter value is sent to the corresponding module, and each module completes the parameter configuration update within 10 seconds and restarts the calculation process, while recording the time of parameter update, the before and after parameter values and the reason, and storing them in a database with a capacity of 100 GB for subsequent analysis and tracing.

[0032] Preferably, the data transmission between the physiological parameter real-time acquisition module and the multi-scale pharmacokinetic modeling module uses a dynamic sampling interval, and the interval time calculation formula is: , wherein, is the current sampling interval, is the basic sampling interval, is the sampling adjustment coefficient, is the absolute value of the blood drug concentration change rate; The infusion rate adjustment formula of the anti-radiation drug dose adjustment module is: , wherein, is the current drug infusion rate, is the basic infusion rate, is the maximum safe blood drug concentration, is the minimum effective blood drug concentration, is the humidity influence coefficient, is the environmental humidity.

[0033] Specifically, the real-time physiological parameter acquisition module employs a dynamic sampling mechanism, while the anti-radiation drug dosage adjustment module utilizes an infusion rate adjustment mechanism. Parameters include the adjustment range of the sampling interval and the adjustment precision of the infusion rate. The sampling interval is adjustable from 1 to 10 seconds, dynamically adapting the sampling frequency based on changes in blood drug concentration. The infusion rate adjustment precision is ±0.001 mg / h, ensuring precise control of the dosing rate. By dynamically adjusting the sampling interval and infusion rate, the system can acquire physiological data more efficiently and execute dosage adjustments more accurately, improving the system's response speed and dosage control accuracy. The implementation method is as follows: the real-time physiological parameter acquisition module adjusts the sampling rate based on the absolute value (range) of the rate of change in blood drug concentration. Adjust the sampling interval; the basic sampling interval is set to 2-5 seconds, and the sampling adjustment coefficient is set to [value missing]. When the rate of change in blood drug concentration is large, the sampling interval is shortened, and vice versa. The current sampling interval is calculated using a formula to ensure dense sampling when the blood drug concentration changes rapidly and reduced sampling frequency when the change is gradual, thereby improving data acquisition efficiency. The anti-radiation drug dosage adjustment module receives the basic infusion rate (range 0.1-4 mg / h), maximum safe blood drug concentration (25-35 mg / L), and minimum effective blood drug concentration (3-8 mg / L) data from the population Bayesian optimization calculation module, and calculates the relative deviation ratio based on the difference between the current blood drug concentration and the target blood drug concentration. Simultaneously, based on the ambient humidity (range 20%-95%) and humidity influence coefficient ( ) from the transport environment data capture module, the module also calculates the relative deviation ratio. The infusion rate is adjusted to obtain the current drug infusion rate. When the ambient humidity is high, the infusion rate is appropriately reduced, and vice versa. The infusion rate is adjusted every 10 seconds, and the adjusted rate is converted into a stepper motor rotation signal via the infusion pump drive circuit. Each step of the stepper motor corresponds to a drug dosage of 0.0001 mg, ensuring that the actual infusion rate matches the calculated value. After each sampling interval adjustment and infusion rate adjustment, relevant parameters are fed back to the system parameter dynamic update module for optimizing subsequent sampling and adjustment strategies.

[0034] Preferably, the multiscale pharmacokinetic modeling module comprises: a tissue scale drug distribution calculation unit, a cell scale drug transport simulation unit, and a cross-scale parameter mapping unit; the tissue scale drug distribution calculation unit receives blood flow rate and vascular density data output by the physiological parameter real-time acquisition module, combines temperature data output by the transport environment data capture module, and obtains initial distribution amounts of drugs in various tissues by calculating products of blood perfusion rates of different tissue organs and drug distribution coefficients; the cell scale drug transport simulation unit simulates processes of drug molecules entering cell interiors through passive diffusion and active transport and binding with target points based on cell membrane permeability and intracellular target point concentration data, and outputs dynamic change curves of drug-target point complexes; and the cross-scale parameter mapping unit converts drug concentration data obtained by the tissue scale calculation into boundary conditions of the cell scale model, and feeds drug metabolism rates simulated by the cell scale simulation back to the tissue scale model, so as to perform coupled calculation of the two scale models.

[0035] Specifically, the multi-scale pharmacokinetic modeling module includes a tissue scale drug distribution calculation unit, a cell scale drug transport simulation unit, and a cross-scale parameter mapping unit. The technical parameters include the spatial resolution of the tissue scale, the simulation accuracy of the cell scale, and the error range of the cross-scale parameter mapping. The spatial resolution of the tissue scale is 0.5-2 mm, which can accurately divide the boundaries of different tissues and organs. The simulation accuracy of the cell scale is controlled within ±3%, which can accurately reflect the interaction between drug molecules and targets. The error of the cross-scale parameter mapping is not more than ±5%, which ensures the consistency of the data of different scale models. Through the cooperative work of the three units, the whole process of drug metabolism from macro tissue to micro cell is realized, breaking through the limitations of single scale model, and providing more comprehensive and accurate theoretical support for dose optimization. The implementation is as follows: the tissue scale drug distribution calculation unit receives the blood flow rate (range 0.5-5 L / min) and vascular density (range 100-500 roots / mm³) data output by the physiological parameter real-time acquisition module, and the temperature (range -10-50°C) data output by the transport environment data capture module, calculates the initial distribution amount of drugs in each tissue according to the blood perfusion rate (heart 300-500 mL / min / 100g, liver 100-300 mL / min / 100g, kidney 200-400 mL / min / 100g) and drug distribution coefficient (fat 0.1-0.5, muscle 0.5-1.5, liver 1.0-3.0) of each tissue, and updates the data every 5 minutes. The cell scale drug transport simulation unit simulates the process of drug molecules entering the cell interior and binding to the target through passive diffusion and active transport based on the cell membrane permeability (range 1e-6 to 1e-4 cm / s) and intracellular target concentration (range 0.1-1 μmol / L) data, wherein the passive diffusion rate is proportional to the concentration difference, and the active transport rate is affected by the number of carriers (range 1e3 to 1e5 per cell) and affinity (range 1e-6 to 1e-4 mol / L). The dynamic change curve of drug-target complex is output every 1 minute. The cross-scale parameter mapping unit converts the drug concentration data calculated by the tissue scale into the boundary conditions of the cell scale model, and adjusts the conversion coefficient according to the tissue type between 0.8-1.2, and feeds back the drug metabolism rate (range 0.01-0.1 μmol / min) simulated by the cell scale to the tissue scale model to correct the metabolism parameters of the tissue scale, realizing the coupled calculation of the two scale models, and completing the cross-scale data interaction every 3 minutes.

[0036] Preferably, the population Bayesian optimization calculation module comprises: a population data hierarchical processing unit, a prior distribution construction unit, a posterior probability calculation unit, and an optimization parameter extraction unit; the population data hierarchical processing unit stratifies the received multi-scale pharmacokinetic modeling module output data according to the age, weight, and radiation exposure of the wounded, and calculates the mean, standard deviation, and coefficient of variation of each stratified data; the prior distribution construction unit sets a prior probability density function conforming to a normal-gamma distribution for the drug metabolism parameters according to the characteristics of the stratified data, wherein the mean parameter is associated with the stratified mean, and the variance parameter is associated with the stratified standard deviation; the posterior probability calculation unit calculates the posterior probability distribution of each drug metabolism parameter by using the Markov chain Monte Carlo method in combination with the real-time collected blood drug concentration observation data; and the optimization parameter extraction unit selects the parameter interval corresponding to the cumulative probability of 95% from the posterior probability distribution, and takes the interval midpoint as the optimal parameter value of the population Bayesian optimization calculation.

[0037] Specifically, the population Bayesian optimization calculation module comprises a population data hierarchical processing unit, a prior distribution construction unit, a posterior probability calculation unit, and an optimization parameter extraction unit. The technical parameters involve the statistical accuracy of hierarchical data, the goodness of fit of the prior distribution, the convergence speed of the posterior probability calculation, and the extraction error of the optimization parameter. The statistical accuracy of the hierarchical data is controlled within ±2%, ensuring the accuracy of the characteristics of each hierarchical data. The goodness of fit of the prior distribution is tested by the chi-square test, and the P value is greater than 0.05, indicating that the distribution model is well adapted to the data. The convergence speed of the posterior probability calculation is 50-100 iterations, which can quickly obtain a stable posterior distribution. The extraction error of the optimization parameter is not more than ±4%, ensuring the reliability of the extracted parameter. Through the collaborative work of the four units, the scientific processing of population data and parameter optimization are realized, the individual differences are fully corrected by the population pharmacokinetic rules, and the accuracy and adaptability of dose adjustment are improved. The implementation mode is as follows: after the population data hierarchical processing unit receives the data output by the multi-scale pharmacokinetic modeling module, the data is layered according to the age of the wounded person (18-30 years old, 31-50 years old, 51-65 years old), the weight (50-60 kg, 61-80 kg, 81-100 kg), and the radiation exposure degree (0-200 mSv, 201-500 mSv, 501-1000 mSv), and the mean, standard deviation, and coefficient of variation (the coefficient of variation is controlled within 10%-30%) of each hierarchical data are calculated. The hierarchical results are updated every 10 minutes. The prior distribution construction unit sets the prior probability density function of the drug metabolism parameter according to the characteristics of the hierarchical data, which conforms to the normal-gamma distribution. The deviation of the mean parameter from the hierarchical mean is not more than 5%, and the ratio of the variance parameter to the hierarchical standard deviation is between 0.8 and 1.2, ensuring that the prior distribution can reflect the statistical characteristics of the population data. The posterior probability calculation unit uses the Markov chain Monte Carlo method and combines the real-time collected blood drug concentration observation data (collected every 5 minutes) to set the chain length to 10000-20000 steps and the burn-in period to 2000-5000 steps, and calculate the posterior probability distribution of each drug metabolism parameter. The calculation is completed every 15 minutes. The optimization parameter extraction unit selects the parameter interval corresponding to the cumulative probability of 95% from the posterior probability distribution, controls the ratio of the difference between the upper and lower limits of the interval to the mean of the parameter within 10%-20%, and takes the midpoint of the interval as the optimal parameter value of the population Bayesian optimization calculation. The optimal parameter is sent to the anti-radiation drug dose adjustment module every 5 minutes to guide the dose adjustment.

[0038] The preferred anti-radiation drug dose adjustment module comprises: a dose requirement calculation unit, a drug administration mode selection unit, and an infusion precision control unit; the dose requirement calculation unit receives the optimal parameters output by the population Bayesian optimization calculation module, combines the current blood drug concentration and radiation exposure dose data output by the physiological parameter real-time acquisition module, calculates the difference between the target blood drug concentration and the current blood drug concentration, and the required drug increment for radiation protection to obtain the total dose requirement value; the drug administration mode selection unit selects the continuous infusion mode when the vibration frequency is lower than the threshold value according to the vibration frequency and the degree of vibration data output by the transportation environment data capture module, and switches to the intermittent pulse drug administration mode when the vibration frequency is higher than the threshold value, while adjusting the pulse interval and the single pulse dose; the infusion precision control unit compensates for the infusion flow fluctuation caused by the vibration of the transportation carrier by monitoring the real-time change of the infusion pipeline pressure, and corrects the deviation between the actual drug administration amount and the theoretical drug administration amount within every 10 seconds.

[0039] Specifically, the anti-radiation drug dose adjustment module includes a dose requirement calculation unit, a drug administration mode selection unit, and an infusion precision control unit. The technical parameters involve the accuracy of dose calculation, the response time of drug administration mode switching, and the control error of infusion precision, etc. The accuracy of dose calculation can reach ±0.03 mg, which can meet the dose requirements under different injury conditions; the response time of drug administration mode switching is not more than 1 second, which ensures the rapid adjustment of the drug administration mode when the environment changes; the control error of infusion precision is not more than ±0.005 mg, which ensures the consistency of the actual drug administration amount and the theoretical value. Through the cooperative work of the three units, the accurate calculation of the drug dose, the dynamic adjustment of the drug administration mode, and the precise control of the infusion process are realized, which ensures that the drug dose is always at an appropriate level during transportation, improves the treatment effect, and reduces adverse reactions. The implementation mode is that the dose requirement calculation unit receives the optimal parameters output by the population Bayesian optimization calculation module, combines the current blood drug concentration (range 0-50 mg / L) and radiation exposure dose (range 0-1000 mSv) data output by the physiological parameter real-time acquisition module, calculates the difference between the target blood drug concentration (range 5-30 mg / L) and the current blood drug concentration, and calculates the drug increment required for radiation protection (0.5-2 mg of drug increment per 100 mSv of radiation exposure) according to the radiation exposure dose. The sum of the two is the total dose requirement value, and the calculation result is updated every 5 minutes. The drug administration mode selection unit sets the vibration frequency threshold to 10 Hz according to the vibration frequency (range 0-50 Hz) and the degree of jolt (measured by vibration acceleration, range 0-10 m / s²) data output by the transportation environment data capture module. When the vibration frequency is lower than the threshold, the continuous infusion mode is selected, and the infusion rate is adjusted between 0.1-5 mg / h according to the dose requirement; when the vibration frequency is higher than the threshold, the intermittent pulse drug administration mode is switched, the pulse interval is set to 1-5 minutes, and the single pulse dose is 5%-20% of the total dose requirement, which is dynamically adjusted according to the degree of jolt. The infusion precision control unit monitors the pressure change of the infusion pipeline in real time through the pressure sensor (measurement range 0-50 kPa, accuracy ±0.2 kPa), and collects pressure data every 0.1 seconds. When the pressure fluctuation exceeds ±5%, the driving voltage (range 5-12 V) of the infusion pump is adjusted to compensate for the flow fluctuation, correct the deviation between the actual drug administration amount and the theoretical drug administration amount within every 10 seconds, ensure that the deviation is not more than ±0.005 mg, and feedback the correction data to the system parameter dynamic updating module for optimizing the subsequent infusion control strategy.

[0040] The space-time multi-scale physiological pharmacokinetic model in the application is a model capable of describing the metabolism process of an anti-radiation drug in a wounded person from multiple dimensions of time and space. The implementation thereof involves a tissue scale, a cell scale and a cross-scale synergistic effect. A tissue scale drug distribution calculation unit receives blood flow rate and vascular density data of a physiological parameter real-time acquisition module and temperature data of a transport environment data capturing module, and calculates and updates an initial distribution amount of the drug in different tissues according to blood perfusion rates and drug distribution coefficients of each tissue organ, with a spatial resolution of 0.5-2 mm and an accurate division of tissue organ boundaries. A cell scale drug transport simulation unit simulates the process of drug molecules entering cells and binding with target points through passive diffusion and active transport based on cell membrane permeability and intracellular target point concentration data, with a simulation accuracy controlled within ±3%, and outputs a drug-target point compound dynamic change curve every 1 minute. A cross-scale parameter mapping unit performs cross-scale data interaction every 3 minutes, converts the tissue scale drug concentration data into boundary conditions of the cell scale model, and feeds back the cell scale drug metabolism rate to the tissue scale model to correct the metabolism parameters, with a mapping error of not more than ±5%. The model is used to comprehensively reflect the metabolism law of the drug in the body from a macroscopic tissue to a microscopic cell, breaks through the limitation of a single scale model, and provides a multi-dimensional and accurate theoretical basis for subsequent dose optimization. The significance thereof lies in that the metabolism process of the drug under a complex transport environment and a dynamic physiological state of a wounded person can be more truly presented, a scientific foundation is laid for precise control of the anti-radiation drug dose, and the pertinence and effectiveness of dose adjustment are improved.

[0041] The hybrid factor adjusted population Bayesian optimization model is a model that combines population pharmacokinetic data and individual real-time data to adjust the hybrid factors affecting the dose of anti-radiation drugs to achieve dose optimization. Its implementation process is completed by four units: population data hierarchical processing, prior distribution construction, posterior probability calculation, and optimized parameter extraction. The population data hierarchical processing unit processes the data output by the multi-scale pharmacokinetic modeling module according to the age, weight, and radiation exposure of the wounded, and calculates the mean, standard deviation, and coefficient of variation of each hierarchical data. The coefficient of variation is controlled within 10%-30%, and the hierarchical results are updated every 10 minutes with a statistical accuracy of ±2%. The prior distribution construction unit sets the prior probability density function of the drug metabolism parameters according to the characteristics of the hierarchical data, ensuring that the mean parameter deviation is not more than 5% from the hierarchical mean, and the variance parameter ratio is between 0.8-1.2 of the hierarchical standard deviation. The distribution model is adapted through chi-square test (P value>0.05). The posterior probability calculation unit uses the Markov chain Monte Carlo method to combine the blood concentration observation data collected every 5 minutes, sets the chain length to 10000-20000 steps and the burn-in period to 2000-5000 steps, and completes the posterior probability distribution calculation every 15 minutes with a convergence speed of 50-100 iterations. The optimized parameter extraction unit selects the parameter interval with a cumulative probability of 95% from the posterior probability distribution, controls the ratio of the interval upper and lower limits to the parameter mean within 10%-20%, takes the midpoint as the optimal parameter, and extracts the error of ±4% every 5 minutes to send to the dose adjustment module. The role of this model is to fully utilize the population pharmacokinetic rules to correct individual differences and improve the accuracy and adaptability of dose adjustment. Its significance lies in solving the problem of insufficient integration of population data and individual dynamic characteristics in traditional methods, enabling dose optimization to both learn from population common rules and adapt to individual real-time state, achieving precise adjustment of anti-radiation drug dose in complex transport environment, ensuring treatment effect and reducing adverse reactions.

[0042] As shown in Figure 2 An anti-radiation drug dose data acquisition adaptive dose control system, the system runs includes: Step S1, the physiological parameter real-time acquisition module acquires the heart rate, blood pressure, blood drug concentration, and radiation exposure dose data of the wounded through implantable sensors and body surface monitoring equipment, and stores them in the cache area at a frequency of 1 minute / time and marks the acquisition time stamp; Step S2, the transport environment data capture module uses temperature and humidity sensors, vibration sensors, and altimeters to collect temperature, humidity, vibration frequency, and altitude data inside the transport carrier, and synchronizes the physiological parameter data in time; Step S3, the multi-scale pharmacokinetic modeling module calls the preset cross-tissue-cell scale model, takes the synchronized physiological and environmental data as input variables, and calculates the distribution concentration of the drug in the blood, liver and kidney and the generation rate of the metabolite; Step S4, the population Bayesian optimization calculation module loads the historical population pharmacokinetic data, constructs a target function in combination with the multi-scale model output result, and obtains the optimized value of the drug dose suitable for the current transport environment through iterative calculation; Step S5, the anti-radiation drug dose adjustment module drives the infusion pump actuator according to the optimized value to adjust the drug infusion rate and single dose, and collects the adjusted blood drug concentration data; Step S6, the system parameter dynamic updating module compares the adjusted blood drug concentration data with the target concentration, corrects the parameter coefficients of the multi-scale model and the prior distribution parameters of the population Bayesian optimization according to the deviation value, and completes a self-adaptive control cycle.

[0043] An anti-radiation drug dose data acquisition self-adaptive dose control system, which has significant advantages in cross-scale analysis and can effectively overcome the problem of lack of cross-scale integration capability in the prior art. The physiological parameter real-time acquisition module accurately acquires data such as the heart rate, blood pressure and blood drug concentration of the wounded, the transport environment data capture module comprehensively collects information such as the vibration frequency, temperature and humidity of the transport carrier, and both of them cooperatively input data into the multi-scale pharmacokinetic modeling module. The module builds a drug metabolism dynamics model covering the tissue and cell scales with the aid of these data, realizes the analysis of the whole process of drug metabolism from the macro to the micro, changes the limitation of traditional technology which only relies on single scale data, and more truly reflects the mechanism and effect of the drug in the body.

[0044] In the fusion of population and individual data, the system has outstanding advantages and can solve the defect that the prior art does not effectively fuse population data and individual dynamic characteristics. The population Bayesian optimization calculation module fully utilizes the output result of the multi-scale pharmacokinetic modeling module, and performs optimization calculation in combination with rich population pharmacokinetic data, which not only absorbs the common characteristics of population rules, but also pays attention to individual differences. The anti-radiation drug dose adjustment module accurately adjusts the drug infusion rate and single dose according to the optimization result, and the system parameter dynamic updating module continuously corrects the calculation parameters of each module according to the changes of the adjusted physiological parameters, realizing the organic combination of population data and individual real-time state and avoiding the problems of poor individual adaptability or adjustment lag.

[0045] In addition, the system has strong dynamic adaptability and can maintain stable and efficient operation in a complex transport environment. The modules are closely connected and work in collaboration. The physiological parameter real-time acquisition module and the transport environment data capture module continuously provide dynamic data. The multi-scale pharmacokinetic modeling module and the population Bayesian optimization calculation module quickly process and analyze. The anti-radiation drug dose adjustment module responds in a timely manner. The system parameter dynamic updating module continuously optimizes parameter configuration. This efficient collaborative mechanism enables the system to quickly adapt to various environmental changes and fluctuations in the physiological state of the wounded during transport, ensuring that the anti-radiation drug dose remains at a precise and appropriate level, providing a reliable guarantee for the treatment of the wounded.

[0046] In the description of the present application, it should be noted that unless otherwise explicitly specified and limited, the terms "arrangement", "installation", "connection", "link", "fixing" should be understood broadly, for example, can be fixedly connected, can also be detachably connected, or integrally connected; can be mechanically connected, can also be electrically connected; can be directly connected, can also be indirectly connected through an intermediate medium, can be internal communication of two elements. For those skilled in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0047] Although the embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, replacements and variations of these embodiments can be made without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalent scope.

Claims

1. An adaptive dose control system for acquiring radiation-resistant drug dose data, characterized in that, include: The system includes a real-time physiological parameter acquisition module, a transport environment data capture module, a multi-scale pharmacokinetics modeling module, a population Bayesian optimization calculation module, an anti-radiation drug dosage adjustment module, and a system parameter dynamic update module. The output of the real-time physiological parameter acquisition module is connected to the first input of the multi-scale pharmacokinetic modeling module; the output of the transport environment data capture module is connected to the second input of the multi-scale pharmacokinetic modeling module; the output of the multi-scale pharmacokinetic modeling module is connected to the input of the population Bayesian optimization calculation module; the output of the population Bayesian optimization calculation module is connected to the input of the anti-radiation drug dosage adjustment module; the feedback of the anti-radiation drug dosage adjustment module is connected to the input of the system parameter dynamic update module; and the output of the system parameter dynamic update module is connected to the real-time physiological parameter acquisition module, the transport environment data capture module, and the multi-scale pharmacokinetic modeling module, respectively. The module connects to the parameter configuration terminals of the population Bayesian optimization calculation module and the anti-radiation drug dosage adjustment module; the physiological parameter real-time acquisition module acquires data on the patient's heart rate, blood pressure, blood drug concentration, and radiation exposure dose; the transport environment data capture module acquires data on the vibration frequency, temperature, humidity, and altitude of the transport vehicle; the multi-scale pharmacokinetic modeling module constructs a cross-tissue and cellular pharmacokinetic model based on the above physiological parameters and environmental data; the population Bayesian optimization calculation module uses the model output results combined with population pharmacokinetic data to perform optimization calculations; and the anti-radiation drug dosage adjustment module adjusts the drug infusion rate and single dose according to the optimization calculation results. The system parameter dynamic update module updates the calculation parameters of each module based on the changes in physiological parameters after dose adjustment.

2. The adaptive dose control system for acquiring radiation-resistant drug dose data according to claim 1, characterized in that, When the multi-scale pharmacokinetic modeling module and the population Bayesian optimization calculation module work together, the following model formula is used to describe the relationship between the metabolic rate and dose of the drug during transport: ,in, Let t be the concentration of anti-radiation drug in the blood of the wounded. For transit time, This is the drug absorption rate constant. This is the drug elimination rate constant. To transport environmental radiation intensity, The effective radiation protection threshold for drugs, The internal volume of the transport carrier. For ambient humidity; When dealing with confounding factors, the population Bayesian optimization calculation module uses the following model formula: ,in, Let be the drug metabolism parameters of the j-th injured person in the i-th group. The mean of the group parameters, This is a group-level random effect that follows a standard normal distribution. For the standard deviation of the population parameter, It represents a random effect at the individual level, following a normal distribution with a mean of 0.

3. The adaptive dose control system for acquiring radiation-resistant drug dose data according to claim 1, characterized in that, The anti-radiation drug dosage adjustment module calculates the real-time dosage based on the drug distribution coefficient output by the multi-scale pharmacokinetic modeling module and the dosage optimization value obtained by the population Bayesian optimization calculation module, using the following model formula: ,in, This is a single-dose dose. To achieve the target blood drug concentration, This represents the current blood drug concentration. The apparent volume of distribution of the drug. The altitude influence coefficient is... This is the current altitude. For drug bioavailability, This is the temperature influence coefficient. The ambient temperature during transport; The transport environment data acquisition module incorporates vibration frequency data into the dose adjustment correction, using the following formula: ,in, This is the vibration-corrected dosage. This is the vibration influence coefficient. The vibration frequency of the transport carrier.

4. The adaptive dose control system for acquiring radiation-resistant drug dose data according to claim 1, characterized in that, The multi-scale pharmacokinetic modeling module describes the drug-target binding process at the cellular scale. The model formula is as follows: ,in, This refers to the concentration of the drug-target complex. This refers to the extracellular drug concentration. The total concentration at the target site. The binding rate constant, The dissociation rate constant is Radiation dose influencing factor, The cumulative radiation exposure dose for the injured; The population Bayesian optimization calculation module uses the complex concentration data to estimate population parameters. The estimation formula is as follows: ,in, Parameters for given observation data y The posterior probability density, Let be the likelihood function. For parameters The prior distribution, For hyperparameters The marginal distribution.

5. The adaptive dose control system for acquiring radiation-resistant drug dose data according to claim 1, characterized in that, The system parameter dynamic update module updates the scale transformation coefficients of the multi-scale pharmacokinetics modeling module based on the execution results of the anti-radiation drug dosage adjustment module using the following model formula: ,in, These are the updated scale transformation coefficients. These are the initial scale transformation coefficients. The update rate constant is... For reference blood drug concentration, The integral of the difference between the blood drug concentration and the reference concentration from time 0 to time t; The prior distribution parameters of the population Bayesian optimization computation module are updated using the following formula: ,in, The updated population parameter mean. The initial population parameter mean, These are the weighting coefficients. For sample size, For the i-th observation, Let be the i-th predicted value.

6. The adaptive dose control system for acquiring radiation-resistant drug dose data according to claim 1, characterized in that, Data transmission between the real-time physiological parameter acquisition module and the multi-scale pharmacokinetic modeling module adopts a dynamic sampling interval, and the interval time is calculated using the following formula: ,in, This is the current sampling interval. Based on the sampling interval, This is the sampling adjustment coefficient. This represents the absolute value of the rate of change in blood drug concentration. The infusion rate adjustment formula for the anti-radiation drug dosage adjustment module is as follows: ,in, This is the current drug infusion rate. Based on the infusion rate, To achieve the maximum safe blood drug concentration, To achieve the minimum effective blood drug concentration, Humidity influence coefficient This refers to ambient humidity.

7. The adaptive dose control system for acquiring radiation-resistant drug dose data according to claim 1, characterized in that, The multi-scale pharmacokinetic modeling module includes: a tissue-scale drug distribution calculation unit, a cell-scale drug transport simulation unit, and a cross-scale parameter mapping unit. The tissue-scale drug distribution calculation unit receives blood flow velocity and vascular density data output from the real-time physiological parameter acquisition module, and combines it with temperature data output from the transport environment data capture module. By calculating the product of blood perfusion rate and drug distribution coefficient in different tissues and organs, it obtains the initial distribution amount of drug in each tissue. The cell-scale drug transport simulation unit simulates the process of drug molecules entering the cell and binding to the target through passive diffusion and active transport based on cell membrane permeability and intracellular target concentration data, and outputs the dynamic change curve of the drug-target complex. The cross-scale parameter mapping unit converts the drug concentration data calculated at the tissue scale into the boundary conditions of the cell-scale model, and feeds back the drug metabolism rate simulated at the cell scale to the tissue-scale model, performing coupled calculations of the two scale models.

8. The adaptive dose control system for acquiring radiation-resistant drug dose data according to claim 1, characterized in that, The population Bayesian optimization computation module includes: a population data stratification processing unit, a prior distribution construction unit, a posterior probability calculation unit, and an optimization parameter extraction unit. The population data stratification processing unit stratifies the received multi-scale pharmacokinetic modeling module output data according to the injured person's age, weight, and radiation exposure level, and calculates the mean, standard deviation, and coefficient of variation for each stratum. The prior distribution construction unit sets a prior probability density function conforming to a normal-gamma distribution for the drug metabolism parameters based on the stratified data characteristics, where the mean parameter is correlated with the stratified mean, and the variance parameter is correlated with the stratified standard deviation. The posterior probability calculation unit uses the Markov chain Monte Carlo method, combined with real-time collected blood drug concentration observation data, to calculate the posterior probability distribution of each drug metabolism parameter. The optimization parameter extraction unit selects the parameter interval corresponding to a cumulative probability of 95% from the posterior probability distribution, and takes the midpoint of the interval as the optimal parameter value for the population Bayesian optimization computation.

9. The adaptive dose control system for acquiring radiation-resistant drug dose data according to claim 1, characterized in that, The anti-radiation drug dosage adjustment module includes: a dosage requirement calculation unit, a dosing method selection unit, and an infusion accuracy control unit. The dosage requirement calculation unit receives the optimal parameters output by the population Bayesian optimization calculation module, and combines them with the current blood drug concentration and radiation exposure dose data output by the physiological parameter acquisition module in real time. By calculating the difference between the target blood drug concentration and the current blood drug concentration, and the drug increment required for radiation protection, the total dosage requirement value is obtained. The dosing method selection unit captures the vibration frequency and bump level data output by the transport environment data acquisition module. When the vibration frequency is below the threshold, the continuous infusion mode is selected. When the vibration frequency is above the threshold, the intermittent pulse dosing mode is switched, and the pulse interval and single pulse dose are adjusted at the same time. The infusion accuracy control unit compensates for the infusion flow fluctuation caused by the vibration of the transport carrier by monitoring the pressure change of the infusion pipeline in real time, and corrects the deviation between the actual dosage and the theoretical dosage every 10 seconds.

Images