Methods, apparatus, storage media, and electronic equipment for determining the stochastic perturbation term in pharmacokinetic models.

Abstract

This disclosure provides a method, apparatus, computer program product, non-transitory computer-readable storage medium, and electronic device for determining random perturbation terms in a pharmacokinetic model. The method includes: obtaining a range of values ​​for the random perturbation term; randomly selecting multiple initial values ​​for the random perturbation term based on the range of values; and determining a target value for the random perturbation term based on the multiple initial values. Embodiments of this disclosure can improve the calculation accuracy of random perturbation terms in pharmacokinetic models, thereby improving the prediction accuracy of blood drug concentrations.

Description

Technical Field

[0001] This disclosure generally relates to the field of computer technology, and more specifically to a method, apparatus, computer program product, non-transitory computer-readable storage medium, and electronic device for determining the random perturbation term of a pharmacokinetic model. Background Technology

[0002] Personalized dosing regimens are tailored to individual patients. Through therapeutic drug monitoring, utilizing the principles and methods of clinical pharmacokinetics, and combining clinical experience, the optimal dosage is determined for each patient to achieve the best therapeutic effect, reduce or avoid adverse reactions, and improve the safety and effectiveness of clinical drug use. Previously, recommended drug doses were mostly population average doses; only a few low-toxicity drugs achieved optimal efficacy when administered at average doses. Even with the same dosage, some drugs often only produce satisfactory results in a portion of patients. The reasons for this phenomenon are related to both the patient's own condition and the drug itself.

[0003] Developing individualized dosing regimens using population pharmacokinetic models, considering both patient-related and drug-related factors, requires complex computational processes. Existing individualized dosing support systems have shortcomings. Individualized dosing decision support systems employ maximum a posteriori (MAP) Bayesian estimation, which necessitates prior information such as compartment model fitting and estimation of pharmacokinetic parameters and their distributions. The accuracy of individual pharmacokinetic parameter estimation is affected by inter-individual and residual variability, potentially leading to unstable results. Multiple linear regression methods for predicting blood drug concentrations lack universality. Furthermore, multiple linear regression models do not require compartment model fitting, relying only on patient information and pathophysiological indicators to estimate individualized dosages, resulting in relatively low prediction accuracy. Neural networks also have limitations; they may estimate too many parameters, have insufficient model generalization ability, and suffer from overlearning problems, leading to poor prediction performance.

[0004] Therefore, it is necessary to propose a new scheme for determining pharmacokinetic parameters in order to solve at least one of the above-mentioned technical problems. Summary of the Invention

[0005] The purpose of this disclosure is to provide a method, apparatus, computer program product, non-transitory computer-readable storage medium, and electronic device for determining the random perturbation term of a pharmacokinetic model, so as to improve the calculation accuracy of the random perturbation term of the pharmacokinetic model and thereby improve the prediction accuracy of blood drug concentration.

[0006] According to a first aspect of this disclosure, a method for determining a random perturbation term in a pharmacokinetic model is provided, comprising: obtaining a range of values ​​for the random perturbation term; randomly selecting a plurality of initial values ​​for the random perturbation term based on the range of values ​​for the random perturbation term; and determining a target value for the random perturbation term based on the plurality of initial values ​​for the random perturbation term.

[0007] According to a second aspect of this disclosure, an apparatus for determining a random perturbation term in a pharmacokinetic model is provided, comprising: a value range acquisition module for acquiring the value range of the random perturbation term; an initial value determination module for randomly selecting multiple initial values ​​of the random perturbation term based on the value range of the random perturbation term; and a target value determination module for determining a target value of the random perturbation term based on the multiple initial values ​​of the random perturbation term.

[0008] According to a third aspect of this disclosure, a computer program product is provided, including program code instructions that, when executed by a computer, cause the computer to perform the method described according to a first aspect of this disclosure.

[0009] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described according to a first aspect of this disclosure.

[0010] According to a fifth aspect of this disclosure, an electronic device is provided, comprising: a processor; a memory in electronic communication with the processor; and instructions stored in the memory and executable by the processor to cause the electronic device to perform the method according to a first aspect of this disclosure.

[0011] It should be understood that the content described in this section is not intended to identify key or essential features of the claimed invention, nor is it intended to be used alone to determine the scope of the claimed invention. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In all drawings, the same reference numerals refer to similar but not necessarily the same elements.

[0013] Figure 1 A system architecture diagram of an embodiment of a method for determining the stochastic perturbation term of a pharmacokinetic model according to the present disclosure is shown;

[0014] Figure 2 A flowchart illustrating an embodiment of a method for determining the random perturbation term of a pharmacokinetic model according to the present disclosure is shown;

[0015] Figure 3A flowchart illustrating a specific example of a method for determining the stochastic perturbation term of a pharmacokinetic model according to this disclosure is shown.

[0016] Figure 4 An exemplary block diagram of an apparatus for determining random perturbation terms of a pharmacokinetic model according to an embodiment of the present disclosure is shown;

[0017] Figure 5 A schematic diagram of an example electronic device that can be used to implement embodiments of the present disclosure is shown.

[0018] Specific implementation method

[0019] The present disclosure will be described more fully below with reference to the accompanying drawings. However, the present disclosure may be embodied in many alternative forms and should not be construed as limited to the embodiments described herein. Therefore, although the present disclosure is readily adaptable to various modifications and alternatives, specific embodiments thereof are shown by way of example in the accompanying drawings and will be described in detail herein. However, it should be understood that this is not intended to limit the present disclosure to the specific forms disclosed, but rather, the present disclosure covers all modifications, equivalents, and alternatives falling within the spirit and scope of the present disclosure as defined by the claims.

[0020] It should be understood that although various elements may be described herein using terms such as first, second, etc., these elements should not be limited by these terms. These terms are used only to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the teachings of this disclosure.

[0021] This document describes several examples using block diagrams and / or flowcharts, where each block represents a section of circuitry, modular blocks, or code comprising one or more executable instructions for implementing a specified logical function. It should also be noted that in other implementations, the functions described in the blocks may occur in a different order. For example, depending on the function involved, two blocks shown consecutively may actually execute substantially simultaneously, or these blocks may sometimes execute in reverse order.

[0022] The phrases “according to…example” or “in…example” used in this document mean that a particular feature, structure, or characteristic described in connection with the example can be included in at least one implementation of this disclosure. The phrases “according to…example” or “in…example” appearing in different places throughout this document do not necessarily refer to the same example, nor are they necessarily separate or alternative examples that are mutually exclusive with other examples.

[0023] Figure 1 An exemplary system architecture 100 is shown, illustrating embodiments of methods, apparatus, terminal devices, and storage media for determining random perturbation terms for which the pharmacokinetic models of this disclosure can be applied.

[0024] like Figure 1 As shown, system architecture 100 may include terminal devices 101, 102, and 103, a network 104, and a server 105. Network 104 serves as the medium for providing communication links between terminal devices 101, 102, and 103 and server 105. Network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0025] Users can use terminal devices 101, 102, and 103 to interact with server 105 via network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, and 103, such as voice interaction applications, video conferencing applications, short video social applications, web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0026] Terminal devices 101, 102, and 103 can be either hardware or software. When terminal devices 101, 102, and 103 are hardware, they can be various electronic devices with microphones and speakers, including but not limited to smartphones, tablets, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 players (Moving Picture Experts Group Audio Layer IV), portable computers, and desktop computers, etc. When terminal devices 101, 102, and 103 are software, they can be installed in the aforementioned electronic devices. They can be implemented as multiple software programs or software modules, or as a single software program or software module. No specific limitations are made here.

[0027] Server 105 can be a server that provides various services, such as a backend server that processes requests from terminal devices 101, 102, and 103 to determine the random perturbation terms of the pharmacokinetic model.

[0028] In some cases, the method for determining the random perturbation term of the pharmacokinetic model provided in this disclosure can be executed by terminal devices 101, 102, and 103. Correspondingly, the device for determining the random perturbation term of the pharmacokinetic model can also be set in terminal devices 101, 102, and 103. In this case, the system architecture 100 may not include server 105.

[0029] In some cases, the method for determining the random perturbation term of the pharmacokinetic model provided in this disclosure can be executed by the server 105. Accordingly, the device for determining the random perturbation term of the pharmacokinetic model can also be set in the server 105. In this case, the system architecture 100 may not include terminal devices 101, 102, and 103.

[0030] In some cases, the method for determining the random perturbation term of the pharmacokinetic model provided in this disclosure can be jointly executed by terminal devices 101, 102, 103 and server 105. Correspondingly, the device for determining the random perturbation term of the pharmacokinetic model can also be respectively set in terminal devices 101, 102, 103 and server 105.

[0031] It should be noted that server 105 can be either hardware or software. When server 105 is hardware, it can be implemented as a distributed server cluster consisting of multiple servers, or as a single server. When server 105 is software, it can be implemented as multiple software programs or software modules (e.g., used to provide distributed services), or as a single software program or software module. No specific limitations are made here.

[0032] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0033] Figure 2 A flowchart illustrating an embodiment of a method for determining the random perturbation term of a pharmacokinetic model according to the present disclosure is shown. The method in this embodiment can be derived from... Figure 1 Implemented by the terminal equipment in, or by Figure 1 The server in the middle is implemented, or by Figure 1 The terminal devices and servers in the process are implemented together.

[0034] The pharmacokinetic model in this embodiment is, for example, a population pharmacokinetic (PPK) model. This pharmacokinetic model may include pharmacokinetic parameters. Pharmacokinetic parameters are constants that reflect the dynamic changes of a drug in the body, and can quantitatively describe the kinetic characteristics and changes in action of a drug over time. In this embodiment, the pharmacokinetic parameters include, for example, clearance rate CL, apparent volume of distribution V, and first-order absorption rate constant K. a .

[0035] The pharmacokinetic model in this embodiment also includes random perturbation terms. These random perturbation terms contain information other than the model's principal variables. In one exemplary embodiment, the pharmacokinetic model may include the following random perturbation terms: a random perturbation term λ for clearance CL, a random perturbation term μ for apparent volume of distribution V, and a random perturbation term v for plasma drug concentration C.

[0036] like Figure 2 As shown, the method 200 for determining the random perturbation term of the pharmacokinetic model in this embodiment may include steps 210-230, which will be described below.

[0037] Step 210: Obtain the range of values ​​for the random perturbation term.

[0038] In this embodiment, the range of values ​​for the random disturbance term can be obtained based on the maximum and minimum values ​​that the random disturbance term can take.

[0039] In this embodiment, the range of values ​​for the random disturbance term can be determined based on existing research results, existing practical experience, or a combination of both.

[0040] In an optional embodiment, the predicted and actual values ​​of pharmacokinetic parameters or blood drug concentrations can be determined first based on known sample data, and then the range of values ​​for the random disturbance term can be determined based on the predicted and actual values ​​of pharmacokinetic parameters or blood drug concentrations.

[0041] It is easy to understand that the actual values ​​of pharmacokinetic parameters or blood drug concentrations can be obtained directly or indirectly from known sample data.

[0042] Predicted values ​​of pharmacokinetic parameters or blood drug concentrations can be obtained in the following ways: First, based on known sample data and a pre-defined standard formula, the predicted values ​​of pharmacokinetic parameters are determined, where the standard formula describes the relationship between pharmacokinetic parameters and covariates; second, based on known sample data and the predicted values ​​of pharmacokinetic parameters, the predicted values ​​of blood drug concentrations are determined.

[0043] For example, the standard formula for the relationship between pharmacokinetic parameters and the number of covariates is as follows:

[0044]

[0045]

[0046] Where x 11 x 12 x 21 x 22 Covariates are selected from variables such as population information, combined medication use, smoking status, and infection status. The number and categories of covariates selected are determined based on the actual situation; α1 and α2 represent population parameters; β11 β 12 β 21 β 22 Equal coefficients represent the degree of influence of covariates on parameters; η1 and η2 represent random perturbation terms at the pharmacokinetic parameter level.

[0047] By calculating the deviation between the predicted and actual values, the distribution or range of values ​​of the random disturbance term can be obtained.

[0048] In an optional embodiment, the number of random perturbation terms is multiple. Different random perturbation terms can have different value ranges. For each random perturbation term, a corresponding value range can be determined.

[0049] In an optional embodiment, a random perturbation term is added to the standard formula for predicting pharmacokinetic parameters to describe the inter-individual variation of pharmacokinetic parameters. Then, the parameter prediction results are nested into the final blood drug concentration prediction formula, and a second layer of random perturbation term is added to describe the difference between the predicted blood drug concentration value and the actual blood drug concentration value.

[0050] Step 220: Based on the range of values ​​for the random disturbance term, multiple initial values ​​for the random disturbance term are randomly selected.

[0051] In this embodiment, random sampling is performed within the range of values ​​for the random disturbance term, and the result obtained is the initial value of the random disturbance term.

[0052] In this embodiment, multiple initial values ​​can be obtained through multiple random samplings. The number of initial values ​​can be set according to the actual situation, and this embodiment does not limit this. For example, the number of initial values ​​may be 50, 100, or 200.

[0053] When there are multiple random perturbation terms, for each random perturbation term, multiple initial values ​​are randomly selected from its range. For example, for the random perturbation term λ of the clearance rate CL, 100 initial values ​​are randomly selected from its range. For the random perturbation term μ of the apparent volume of distribution V, 100 initial values ​​are randomly selected from its range. For the random perturbation term v of the blood drug concentration C, 100 initial values ​​are randomly selected from its range.

[0054] Step 230: Determine the target value of the random disturbance term based on the multiple initial values ​​of the random disturbance term.

[0055] In this embodiment, the target value of the random disturbance term is determined based on multiple initial values ​​rather than a single initial value, which reduces the dependence of the calculation result of the random disturbance term on the initial value.

[0056] In an optional embodiment, step 230 can be implemented based on an evolutionary algorithm. Further, step 230 can be implemented based on a differential evolutionary algorithm. By applying the differential evolutionary algorithm, high optimization efficiency, simple parameter settings, and good robustness can be achieved.

[0057] In an optional embodiment, step 230 may further include the following steps: First, establishing an initial population based on multiple initial values ​​of the random perturbation term, wherein each initial value of the random perturbation term corresponds to an initial individual in the initial value population; Second, performing at least one mutation operation, crossover operation, and selection operation on the initial population to obtain at least one generation of evolved population; Third, obtaining the target value of the random perturbation term based on the at least one generation of evolved population.

[0058] In an optional embodiment, the above mutation operation may include: first adding the first initial individual and the second initial individual in the initial population to obtain the sum, and then adding the sum to the third initial individual in the initial population according to a set weight to obtain the mutated individual.

[0059] Other methods can also be used to perform mutation operations. For example, for each initial individual in the initial population, at least one random individual corresponding to the initial individual is randomly selected according to the range of values ​​of the random perturbation term. Then, based on the initial individual and the corresponding random individual, the mutated individual corresponding to the initial individual is obtained.

[0060] In this embodiment, since the value of the random individual is restricted to the range of the random perturbation term, the value of the mutated individual and the target value of the random perturbation term are also restricted to the range of the random perturbation term. This can prevent the target value of the random perturbation term from exceeding the range of the random perturbation term and ensure the standardization of the target value of the random perturbation term.

[0061] In an optional embodiment, the above crossover operation may include: first, selecting an initial individual and a mutated individual, and generating a random number; second, comparing the random number with a preset crossover operator, and determining whether to replace the initial individual with the mutated individual based on the comparison result.

[0062] For example, if the random number corresponding to the initial individual is less than the preset crossover operator, the initial individual is replaced with the corresponding mutated individual. If the random number corresponding to the initial individual is greater than or equal to the preset crossover operator, the original initial individual is retained.

[0063] When there are multiple random perturbation terms, these terms collectively form a random perturbation term vector. Each random perturbation term corresponds to a component of the random perturbation term vector. The aforementioned crossover operation can occur between the initial random perturbation term vector (each component of which is an initial individual) and the mutated random perturbation term vector (at least one component of which is a mutated individual), and is performed on a component-by-component basis.

[0064] In an optional embodiment, there are multiple random perturbation terms. The values ​​of each random perturbation term collectively form a parameter value combination. The initial individual of the random perturbation term is used to form the initial parameter value combination, and the mutated individual of the random perturbation term is used to form the mutated parameter value combination. The above selection operation may include: first, obtaining the predicted blood drug concentration values ​​of the initial parameter value combination and the mutated parameter value combination on known monitoring data, respectively; second, comparing the blood drug concentration prediction accuracy of the initial parameter value combination and the mutated parameter value combination on known monitoring data, and determining whether to retain the initial parameter value combination or the mutated parameter value combination based on the comparison result.

[0065] For example, if the prediction accuracy of blood drug concentration based on known monitoring data is greater than that based on known monitoring data using the combination of initial parameter values, then the combination of initial parameter values ​​is retained; otherwise, the combination of variable parameter values ​​is retained.

[0066] The formula for calculating the accuracy of blood drug concentration prediction is as follows:

[0067] C pred =f(K a ,CL,V,dose,time,ε)

[0068] Among them, C pred k is the predicted value of blood drug concentration. a ε is the first-order absorption rate constant, V is the apparent distribution volume, CL is the clearance rate, ε represents the random perturbation term at the blood drug concentration prediction level, and the random perturbation term at the parameter level is included in the standard formula for the relationship between pharmacokinetic parameters and covariates.

[0069] In an optional embodiment, the initial parameter value combination can be substituted into the formula for calculating the blood drug concentration prediction accuracy to obtain the formula for calculating the blood drug concentration prediction accuracy after assignment. Then, the input data contained in the known monitoring data can be substituted into the formula for calculating the blood drug concentration prediction accuracy after assignment to obtain the blood drug concentration prediction value corresponding to the initial parameter value combination. Finally, the blood drug concentration prediction value corresponding to the initial parameter value combination is compared with the actual blood drug concentration value contained in the known monitoring data to obtain the blood drug concentration prediction accuracy of the initial parameter value combination on the known monitoring data. Similarly, the blood drug concentration prediction accuracy of the variable parameter value combination on the known monitoring data can be obtained. Input data includes, for example, the dosage (Dose) and the time length (time) between the blood drug concentration measurement time and the last administration time.

[0070] In an optional embodiment, before performing at least one mutation, crossover, and selection operation on the initial population, the method may further include: first, obtaining the blood drug concentration prediction accuracy of each initial individual in the initial population on known monitoring data; second, determining whether there are initial individuals in the initial population whose blood drug concentration prediction accuracy meets a preset accuracy requirement; and finally, if there are initial individuals in the initial population whose blood drug concentration prediction accuracy meets the preset accuracy requirement, determining the target value of the random perturbation term based on the initial individuals whose blood drug concentration prediction accuracy meets the preset accuracy requirement.

[0071] If the initial population contains individuals whose blood drug concentration prediction accuracy meets the preset accuracy requirements, the target value of the random perturbation term can be determined based on the initial individual with the highest blood drug concentration prediction accuracy among these initial individuals. This method helps improve the overall computation speed.

[0072] In an optional embodiment, mutation, crossover, and selection operations can be performed cyclically until a stopping condition is met. This stopping condition may be, for example, that the number of executions reaches a preset number, or that the prediction accuracy of individuals in the evolutionary population based on known monitoring data reaches a preset accuracy.

[0073] In an optional embodiment, after step 230, the method may further include: determining the predicted blood drug concentration based on the target value of the random perturbation term; and recommending the dosage based on the predicted blood drug concentration.

[0074] Optionally, the method for determining the random perturbation term of the pharmacokinetic model in this embodiment can be applied to the dosage recommendation of valproic acid.

[0075] The recommended dosage is determined by both deterministic and unknown random factors. After obtaining the initial dosage based on the deterministic factors, it is necessary to estimate the impact of unknown factors on the dosage. This embodiment uses a differential evolution algorithm to estimate two levels of unknown factors that may affect the final prediction of blood drug concentration. The first level is the pharmacokinetic parameter level, representing the random fluctuations of pharmacokinetic parameters. The second level, nested within the first level, represents the magnitude of the unexplained difference between the predicted and observed blood drug concentrations. Simultaneously estimating the random perturbation terms at both levels using differential evolution increases the probability of selecting suitable random perturbation terms, avoiding getting stuck in local optima during the search for the distribution of random perturbation terms for pharmacokinetic parameters and the distribution of random perturbation terms between the predicted and observed blood drug concentrations. This avoids problems such as low accuracy in blood drug concentration prediction due to improper determination of the random perturbation term distribution and the tediousness of constantly correcting the combination of random perturbation terms.

[0076] This embodiment dynamically tracks the current search status of random perturbation terms by leveraging the unique memory capability of the differential evolution method to adjust the search strategy, thereby increasing the global convergence capability and robustness of the random perturbation term search. The differential evolution method mutates and crossovers the random perturbation terms at two levels to obtain an intermediate population of random perturbation terms, thus increasing the selection strategy for random perturbation terms. The differential evolution method is based on a greedy selection method (also known as a greedy algorithm, which means that when solving a problem, it always makes the best choice at the moment, and through multiple greedy selections, it finally obtains the optimal solution for the entire problem) to select, retaining random perturbation terms with higher accuracy in predicting blood drug concentration from the intermediate population according to a pre-set objective function.

[0077] Figure 3 A flowchart illustrating a specific example of a method for determining the stochastic perturbation term of a pharmacokinetic model according to this disclosure is shown. Figure 3 As shown, this example includes the following process: First, the initial iteration range of the random perturbation terms at the pharmacokinetic parameter level and the blood drug concentration prediction level is set according to the distribution of the random perturbation terms. Second, the objective function of differential evolution is defined as a function of the random perturbation terms at both the pharmacokinetic parameter level and the blood drug concentration prediction level. A random perturbation term is added to the standard formula for pharmacokinetic parameter prediction to describe the inter-individual variation of the pharmacokinetic parameters. Then, the parameter prediction results are nested into the final blood drug concentration prediction formula, adding a second layer of random perturbation terms to describe the difference between the predicted and actual values. Finally, the absolute value of the difference between the predicted and actual blood drug concentrations is used as the objective function of differential evolution, which simultaneously includes the random perturbation terms at both the pharmacokinetic parameter level and the blood drug concentration prediction level. Third, the population size is set, which is the candidate solution population for the random perturbation terms. Fourth, the number of iterations is set, and each iteration performs mutation crossover selection. By adjusting the number of iterations, the random perturbation terms are iterated until the objective function converges to the global optimum.

[0078] The specific process of differential evolution involves adding a weighted vector between two random perturbation vectors from the candidate solution population to a third random perturbation vector to generate a new vector, achieving mutation. Then, the mutated vector is crossed with the unmutated vector. The resulting new random perturbation vector is compared with the objective function obtained from the unmutated vector, and the better individual is selected and retained for the next generation. During the evolutionary process, the best vector in each generation is evaluated to record the minimization process. This can yield a very high convergence rate, guiding the search process towards the globally optimal solution in the given vector space.

[0079] This example incorporates a feedback loop. Based on the differential evolution, the model parameters are continuously adjusted to determine whether the accuracy of the blood drug concentration prediction has reached the preset accuracy. The model parameters are then adjusted based on the prediction accuracy until the accuracy of the output prediction value reaches the preset accuracy.

[0080] After obtaining the optimal solution of the random perturbation term, the value of the optimal solution is fixed, and blood drug concentration is predicted based on it to obtain accurate blood drug concentration prediction results.

[0081] Due to the incompleteness of our understanding of the patients under investigation, many unknown influencing factors cannot be incorporated into the model and must be represented by a random disturbance term. Furthermore, the impact of numerous minor influencing factors on the explained variable is minimal. Considering model simplicity and the potentially high cost of obtaining data for many variables, these minor variables are often omitted, and their effects are incorporated into the random disturbance term. This embodiment employs a differential evolution algorithm to dynamically search for the optimal random disturbance term to improve the final prediction accuracy.

[0082] In this embodiment, differential evolution is used to determine multiple random perturbation terms. It has fewer controllable parameters, strong global search capability, and the crossover and mutation selection has inherent parallelism, which can be used for collaborative search. It is efficient and accurate in the process of multi-objective optimization of random perturbation terms.

[0083] Figure 4 An exemplary block diagram of an apparatus for determining random perturbation terms of a pharmacokinetic model according to embodiments of the present disclosure is shown. Figure 4 As shown, the device 400 for determining the random perturbation term of the pharmacokinetic model includes: a value range acquisition module 410, used to acquire the value range of the random perturbation term; an initial value determination module 420, used to randomly extract multiple initial values ​​of the random perturbation term based on the value range of the random perturbation term; and a target value determination module 430, used to determine the target value of the random perturbation term based on the multiple initial values ​​of the random perturbation term.

[0084] It should be understood that Figure 4The various modules of the device 400 shown can be connected to the reference. Figure 2 The steps in method 200 described correspond to each other. Therefore, the operations, features, and advantages described above for method 200 also apply to apparatus 400 and its included modules. For the sake of brevity, some operations, features, and advantages will not be repeated here.

[0085] In an optional embodiment, the above-mentioned random perturbation term includes random perturbation terms of pharmacokinetic parameters and random perturbation terms of blood drug concentration.

[0086] In an optional embodiment, the target value determination module 430 is further configured to: determine the target value of the random perturbation term based on the multiple initial values ​​of the random perturbation term using a differential evolution algorithm.

[0087] In an optional embodiment, the target value determination module 430 is further configured to: establish an initial population based on a plurality of initial values ​​of the random perturbation term, wherein each initial value of the random perturbation term corresponds to an initial individual in the initial value population; perform at least one mutation operation, crossover operation, and selection operation on the initial population to obtain at least one generation of evolutionary population; and obtain the target value of the random perturbation term based on the at least one generation of evolutionary population.

[0088] In an optional embodiment, the mutation operation includes: adding the first initial individual and the second initial individual in the initial population to obtain a sum; and adding the sum to the third initial individual in the initial population according to a set weight to obtain the mutated individual.

[0089] In an optional embodiment, the above crossover operation includes: selecting an initial individual and a mutated individual, and generating a random number; comparing the random number with a preset crossover operator, and determining whether to replace the initial individual with the mutated individual based on the comparison result.

[0090] In an optional embodiment, the number of the above-mentioned random perturbation items is multiple, and the value of each of the above-mentioned random perturbation items together forms a parameter value combination. The initial individual of the above-mentioned random perturbation items is used to form an initial parameter value combination, and the mutated individual of the above-mentioned random perturbation items is used to form a mutated parameter value combination. The selection operation includes: obtaining the blood drug concentration prediction values ​​of the above-mentioned initial parameter value combination and the above-mentioned mutated parameter value combination on known monitoring data respectively; comparing the blood drug concentration prediction accuracy of the above-mentioned initial parameter value combination and the above-mentioned mutated parameter value combination on known monitoring data, and determining whether to retain the above-mentioned initial parameter value combination or the above-mentioned mutated parameter value combination based on the comparison result.

[0091] In an optional embodiment, the target value determination module 430 is further configured to: repeatedly execute the above-mentioned mutation operation, crossover operation and selection operation until the number of executions reaches a preset number, or the prediction accuracy of individuals in the above-mentioned evolutionary population based on known monitoring data reaches a preset accuracy.

[0092] In an optional implementation, the value range acquisition module 410 is further configured to: determine the predicted and actual values ​​of pharmacokinetic parameters or blood drug concentrations based on known sample data; and determine the value range of the random disturbance term based on the predicted and actual values ​​of the pharmacokinetic parameters or blood drug concentrations.

[0093] In an optional embodiment, the value range acquisition module 410 is further configured to: determine the predicted value of the pharmacokinetic parameter based on the known sample data and the preset standard formula, wherein the standard formula is used to describe the relationship between the pharmacokinetic parameter and the covariate; and determine the predicted value of the blood drug concentration based on the known sample data and the predicted value of the pharmacokinetic parameter.

[0094] In an optional embodiment, the device 400 further includes a dosage recommendation module (not shown) for: determining a blood drug concentration prediction result based on the target value of the random perturbation term; and recommending a dosage based on the blood drug concentration prediction result.

[0095] Figure 5 A schematic block diagram of an example electronic device 500 that can be used to implement embodiments of the present disclosure is shown. See also Figure 5 The present invention describes a structural block diagram of an electronic device 500 that can serve as a server or client of the present disclosure, which is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein. Figure 5As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. The RAM 503 may also store various programs and data required for the operation of the device 500. The computing unit 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504. Multiple components in the device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, optical disk, etc.; and a communication unit 509, such as a network card, modem, wireless transceiver, etc. The communication unit 509 allows the device 500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0096] The computing unit 501 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as the method for determining the random perturbation term of a pharmacokinetic model. For example, in some embodiments, the method for determining the random perturbation term of a pharmacokinetic model can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 508. In some embodiments, part or all of the computer program can be loaded and / or installed on device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the method for determining the random perturbation term of a pharmacokinetic model described above can be performed. Alternatively, in other embodiments, the computing unit 501 may be configured by any other suitable means (e.g., by means of firmware) to perform a method for determining random perturbation terms of the pharmacokinetic model.

[0097] The various illustrative logics, logic blocks, modules, circuits, and algorithmic processes described in conjunction with the aspects disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. The interchangeability of hardware and software has been generally described in terms of functionality and illustrated in the various illustrative components, modules, circuits, and processes described above. Whether this functionality is implemented in hardware or software depends on the specific application and the design constraints on the overall system.

[0098] Hardware and data processing apparatuses for implementing the various illustrative logics, logic blocks, modules, and circuits described in conjunction with the aspects disclosed herein may be implemented or performed by general-purpose single-chip or multi-chip processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general-purpose processor may be a microprocessor or any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors combined with a DSP core, or any other such configuration. In some aspects, specific processes and methods may be performed by circuits specific to a given function.

[0099] In one or more aspects, the described functionality can be implemented in hardware, digital electronic circuits, computer software, firmware (including the structures disclosed in this specification and their equivalents) or any combination thereof. The aspects of the subject matter described in this specification can also be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a computer storage medium for execution by a data processing apparatus or for controlling the operation of a data processing apparatus.

[0100] If implemented in software, the functionality can be stored or transferred as one or more instructions or code onto a computer-readable medium. The processes of the methods or algorithms disclosed herein can be implemented in a processor-executable software module that may reside on a computer-readable medium. Computer-readable media include computer storage media and communication media, including any medium capable of transferring a computer program from one place to another. Storage media can be any available medium accessible to a computer. By way of example and not limitation, this computer-readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disc storage, disk storage or other magnetic storage devices, or any other medium that can be used to store the required program code in the form of instructions or data structures and is accessible to a computer. Furthermore, any connection can be properly referred to as a computer-readable medium. The disks and discs used herein include high-density optical discs (CDs), laser discs, optical discs, digital versatile discs (DVDs), floppy disks, and Blu-ray discs, wherein disks typically magnetically copy data, while discs optically copy data using lasers. Combinations of the above should also be included within the scope of computer-readable media. Additionally, the operation of a method or algorithm may be one or any combination or set of code and instructions on a machine-readable and computer-readable medium, which may be incorporated into a computer program product.

[0101] The various embodiments in this disclosure are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device embodiments, equipment embodiments, computer-readable storage medium embodiments, and computer program product embodiments are basically similar to the method embodiments, so the descriptions are relatively simple, and relevant parts can be referred to the descriptions of the method embodiments.

Claims

1. A method for determining the stochastic perturbation term of a pharmacokinetic model, comprising: Obtain the range of values ​​for the random perturbation term; Based on the range of values ​​for the random perturbation term, multiple initial values ​​for the random perturbation term are randomly selected. Based on multiple initial values ​​of the random disturbance term, determine the target value of the random disturbance term; Based on the target value of the random perturbation term, the predicted blood drug concentration is determined, and the recommended dosage is made based on the predicted blood drug concentration. In this study, the differential evolution algorithm was used to estimate the unknown factors at two levels that affect the final prediction of blood drug concentration. The first level represents the random fluctuation of pharmacokinetic parameters, and the second level represents the magnitude of the unexplained difference between the predicted and actual blood drug concentration. The random perturbation terms at both levels were estimated using the differential evolution algorithm. The process of obtaining the range of values ​​for the random perturbation term includes: Based on known sample data, predictive and actual values ​​of pharmacokinetic parameters or blood drug concentrations are determined. Specifically, the predicted values ​​of the pharmacokinetic parameters are determined based on the known sample data and a preset standard formula, whereby the standard formula describes the relationship between pharmacokinetic parameters and covariates. The predicted values ​​of the blood drug concentrations are also determined based on the known sample data and the predicted values ​​of the pharmacokinetic parameters. The range of values ​​for the random perturbation term is determined based on the pharmacokinetic parameters or the predicted and actual values ​​of blood drug concentration. The range of values ​​is the deviation between the predicted value and the actual value. A random perturbation term is added to the standard formula for predicting pharmacokinetic parameters to describe the inter-individual variation of pharmacokinetic parameters. The parameter prediction results are then nested into the final blood drug concentration prediction formula, and a second layer of random perturbation term is added to describe the deviation between the predicted blood drug concentration value and the actual blood drug concentration value.

2. The method according to claim 1, wherein, The random perturbation term includes random perturbation terms of pharmacokinetic parameters and random perturbation terms of blood drug concentration.

3. The method according to claim 1, wherein, Determining the target value of the random disturbance term based on multiple initial values ​​of the random disturbance term includes: Based on multiple initial values ​​of the random perturbation term, the target value of the random perturbation term is determined using the differential evolution algorithm.

4. The method according to claim 2, wherein, The step of determining the target value of the random perturbation term using a differential evolution algorithm based on multiple initial values ​​of the random perturbation term includes: An initial value population is established based on multiple initial values ​​of the random perturbation term, wherein each initial value of the random perturbation term corresponds to an initial individual in the initial value population; Perform at least one mutation, crossover, and selection operation on the initial population to obtain at least one generation of evolved population; The target value of the random perturbation term is obtained based on the at least one generation of the evolutionary population.

5. The method according to claim 4, wherein, The mutation operation includes: The first and second initial individuals in the initial population are added together to obtain the sum; The summation result is added to the third initial individual in the initial population according to the set weight to obtain the mutated individual.

6. The method according to claim 4, wherein, The crossover operation includes: Select an initial individual and a mutated individual, and generate a random number; The random number is compared with a preset crossover operator, and the result of the comparison determines whether to replace the initial individual with the mutated individual.

7. The method according to claim 4, wherein, The number of random perturbation terms is multiple, and the values ​​of each random perturbation term together form a parameter value combination. The initial individual corresponding to the random perturbation term is used to form the initial parameter value combination, and the mutated individual corresponding to the random perturbation term is used to form the mutated parameter value combination. The selection operation includes: The predicted blood drug concentration values ​​of the initial parameter value combination and the variable parameter value combination on known monitoring data are obtained respectively; The accuracy of blood drug concentration prediction based on known monitoring data is compared between the initial parameter value combination and the variable parameter value combination, and the result of the comparison determines whether to retain the initial parameter value combination or the variable parameter value combination.

8. The method according to claim 4, wherein, The method further includes: The mutation, crossover, and selection operations are performed repeatedly until the preset number of executions is reached, or the prediction accuracy of individuals in the evolutionary population based on known monitoring data reaches a preset accuracy.

9. A device for determining the stochastic perturbation term of a pharmacokinetic model, comprising: The value range acquisition module is used to obtain the value range of the random disturbance term; The initial value determination module is used to randomly extract multiple initial values ​​of the random disturbance term based on the value range of the random disturbance term; The target value determination module is used to determine the target value of the random disturbance term based on multiple initial values ​​of the random disturbance term; Based on the target value of the random perturbation term, the predicted blood drug concentration is determined, and the recommended dosage is made based on the predicted blood drug concentration. In this study, the differential evolution algorithm was used to estimate the unknown factors at two levels that affect the final prediction of blood drug concentration. The first level represents the random fluctuation of pharmacokinetic parameters, and the second level represents the magnitude of the unexplained difference between the predicted and actual blood drug concentration. The random perturbation terms at both levels were estimated using the differential evolution algorithm. The process of obtaining the range of values ​​for the random perturbation term includes: Based on known sample data, predictive and actual values ​​of pharmacokinetic parameters or blood drug concentrations are determined. Specifically, the predicted values ​​of the pharmacokinetic parameters are determined based on the known sample data and a preset standard formula, whereby the standard formula describes the relationship between pharmacokinetic parameters and covariates. The predicted values ​​of the blood drug concentrations are also determined based on the known sample data and the predicted values ​​of the pharmacokinetic parameters. The range of values ​​for the random perturbation term is determined based on the pharmacokinetic parameters or the predicted and actual values ​​of blood drug concentration. The range of values ​​is the deviation between the predicted value and the actual value. A random perturbation term is added to the standard formula for predicting pharmacokinetic parameters to describe the inter-individual variation of pharmacokinetic parameters. The parameter prediction results are then nested into the final blood drug concentration prediction formula, and a second layer of random perturbation term is added to describe the deviation between the predicted blood drug concentration value and the actual blood drug concentration value.

10. A computer program product comprising program code instructions that, when executed by a computer, cause the computer to perform the method of at least one of claims 1 to 8.

11. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to at least one of claims 1 to 8.

12. An electronic device, comprising: processor, A memory that communicates electronically with the processor; And instructions, which are stored in the memory and can be executed by the processor to cause the electronic device to perform the method according to at least one of claims 1 to 8.

Images