Game equilibrium optimization-based wound pain management scheme generation method

Abstract

The invention discloses a wound pain management scheme generation method based on game equilibrium optimization. The method comprises the following steps: S1, collecting an individual data set of a patient; s2, constructing a multivariable interaction model for wound pain management based on the acquired patient individualized data set; s3, generating an initial grey wolf population by using the patient individualized data set and the multivariate interaction model, wherein each individual in the initial grey wolf population represents a pain management scheme; s4, defining a revenue function of each variable and calculating an equilibrium solution; s5, performing iterative optimization on the grey wolf population based on a grey wolf optimization algorithm, and adjusting positions and fitness values of individuals in the grey wolf population by introducing a game equilibrium analysis result as a dynamic constraint condition to gradually approach a global optimal pain management scheme; and S6, outputting an optimal pain management scheme finally output by the grey wolf optimization algorithm to clinical medical care personnel in a suggestion form. The reliability and the practicability of a treatment scheme are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of wound pain management technology, and in particular to a method for generating wound pain management schemes based on game equilibrium optimization. Background Technology

[0002] With the development of intelligent medical technology, pain management has gradually become an important research direction in clinical medicine. Wound pain management, as a key area, is directly related to the patient's recovery quality and life comfort. Effective pain management requires comprehensive consideration of multiple factors such as the patient's individual characteristics, medication use, and wound healing progress in order to provide scientific and efficient treatment plans.

[0003] Most existing pain management solutions rely primarily on the experience and judgment of healthcare professionals or fixed procedures. This approach often fails to adequately consider individual patient differences. Specifically, in terms of drug selection and dosage adjustment, existing technologies often use universal dosages as a standard, failing to incorporate personalized data such as the patient's age, gender, medical history, and tolerance. This can lead to increased side effects or insufficient efficacy for some patients during medication. Furthermore, traditional pain management solutions lack dynamic feedback mechanisms and cannot be flexibly adjusted according to the patient's changing condition in real time. For example, failure to optimize the treatment plan in a timely manner when wound healing speeds up or drug tolerance improves may result in overtreatment or treatment delays.

[0004] In recent years, with the increasing application of data-driven medical decision support systems in the field of pain management, some solutions based on statistical analysis and simple optimization models have been proposed. However, existing technologies have revealed the following shortcomings in practical applications: First, most models cannot simultaneously balance the multiple objectives of pain relief, drug side effects, and wound healing speed, and the optimization results often focus only on a single variable; second, existing methods lack the ability to model complex interactive variables, making it difficult to comprehensively describe the nonlinear relationships between individualized patient data; third, the optimization models of existing technologies are usually static and cannot dynamically respond to real-time changes in the patient's condition, thus limiting the adaptability and effectiveness of treatment plans.

[0005] In summary, existing technologies have significant shortcomings in terms of personalized, dynamic, and multi-objective optimization of pain management, which directly affects the treatment effect and rehabilitation efficiency of patients. There is an urgent need for a new method that uses integrated game equilibrium optimization and intelligent algorithms to solve the problem of multi-variable dynamic optimization and generate scientific wound pain management plans. Summary of the Invention

[0006] One objective of this invention is to propose a method for generating wound pain management plans based on game equilibrium optimization. The dynamic feedback mechanism of this invention effectively solves the problem that traditional methods cannot respond to changes in the patient's condition, and significantly improves the reliability and practicality of the treatment plan.

[0007] A method for generating a wound pain management scheme based on game equilibrium optimization according to an embodiment of the present invention includes the following steps:

[0008] S1. Collect individualized patient datasets;

[0009] S2. Based on the collected individualized patient datasets, construct a multivariate interaction model for wound pain management;

[0010] S3. Generate an initial gray wolf population using individualized patient datasets and a multivariate interaction model, where each individual in the initial gray wolf population represents a pain management strategy;

[0011] S4. Use game equilibrium theory to conduct equilibrium analysis on each pain management scheme in the gray wolf population. Treat the multiple variables in the pain management scheme as game participants, define the payoff function of each variable and calculate the equilibrium solution.

[0012] S5. Based on the gray wolf optimization algorithm, the gray wolf population is iteratively optimized. By introducing the game equilibrium analysis results as dynamic constraints, the position and fitness value of individuals in the gray wolf population are adjusted to gradually approach the globally optimal pain management scheme.

[0013] S6. The optimal pain management plan, which is finally output by the Grey Wolf optimization algorithm, is presented to clinical medical staff in the form of a suggestion.

[0014] Optionally, S1 includes:

[0015] S11. Collect the patient's age A and gender G, where G represents the patient's gender, with a value of 1 indicating male and 0 indicating female;

[0016] S12. Collect the patient's pain score. ,in, The patient's pain score at time t is calculated, with a range of [0, 10]. A higher score indicates a more severe pain.

[0017] S13. Collect information on the patient's medication use. :

[0018] ;

[0019] in, This represents the dosage of the i-th drug used by the patient at time t, and n represents the total number of drugs used by the patient at the same time.

[0020] S14. Collect data on the patient's wound healing progress. :

[0021] ;

[0022] in, This indicates the area of ​​the wound that has healed. Indicates the initial area of ​​the wound. The value range is [0, 1], and the closer the value is to 1, the higher the degree of wound healing;

[0023] S15. Construct a patient-specific dataset D:

[0024] .

[0025] Optionally, S2 includes:

[0026] S21. Construct a wound healing rate model based on the patient's individualized dataset D, and obtain the wound healing rate value by differencing the wound healing progress at adjacent time steps. ;

[0027] S22. Construct a drug effect model based on individualized patient datasets, including the drug dosage used by patients at time t. According to drug action coefficient Perform a weighted summation to obtain the drug effect value at time t. ;

[0028] S23. Construct a patient tolerance model based on individualized patient datasets, relating patient age A, patient gender G, and medication usage. Combined with tolerance weighting factor Adjust the patient's tolerance level to obtain the patient's tolerance value. ;

[0029] S24. Construct a pain relief effect model based on individualized patient datasets, including patient pain scores. Progress of wound healing Combined, the pain relief effect value is obtained. :

[0030] ;

[0031] in, The pain relief effect value over time t is used to reflect the level of pain relief in response to the interaction between wound healing process and pain perception. The pain score at time t is used to reflect the patient's pain level at that moment;

[0032] S25. Adjust the wound healing speed value. Drug efficacy value Patient tolerance value and pain relief effect value This is integrated into a multivariate interaction model to simultaneously characterize the interactions among the aforementioned variables:

[0033] ;

[0034] in, For a multivariate interaction model at time t, Used to measure the cumulative effect of multiple drug doses at that moment.

[0035] Optionally, S3 includes:

[0036] S31. Define the multi-layered gray wolf population structure:

[0037] ;

[0038] in, This represents the gray wolf population at level l, where L is the population level and each individual in the population... This indicates a pain management plan, including medication dosage. Medication time Adjunctive treatment measures and personalized adjustment factors ;

[0039] S32. Differentiated initialization of each population layer is performed by combining individualized patient datasets and a multivariate interaction model:

[0040] ;

[0041] in, This represents the drug dosage set for the j-th individual in the l-th population. The dosage of the i-th drug corresponding to the j-th individual in the l-th layer is generated by uniformly distributing it within the constraint range;

[0042] S33. Adjust the weights of drug effects in each population layer using a dynamic mapping function. :

[0043] ;

[0044] in, The effect weight of the i-th drug in the l-th layer population. For the model based on the number of layers l and multivariate interaction The function is dynamically adjusted to adapt to different levels of optimization needs;

[0045] S34. Assign dynamic, personalized adjustment factors to each individual in the population. The adjustment factor is learned based on patient-specific data D and fitness history:

[0046] ;

[0047] in, Let be the dynamic adjustment factor for the j-th individual in the l-th layer of the population. This represents the pain relief effect value over time t. For pain scores at time t, the numerator represents the overall contribution of the individual regimen to the patient's experience, while the denominator reflects the effect of drug dosage in the regimen.

[0048] Optionally, S4 includes:

[0049] S41. Treat each pain management scheme in the gray wolf population as a player in a game, and define the player set:

[0050] ;

[0051] Where N is the number of individuals in the gray wolf population, and each player's strategy includes drug dosage, drug administration time, and auxiliary treatment measures;

[0052] S42. Define each player based on individualized patient datasets and multivariate interaction models. payoff function :

[0053] ;

[0054] in, For players The payoff function is used to comprehensively evaluate the merits of pain management programs. These are weighting coefficients used to balance the objective variables in the payoff function. The pain relief effect value over time t represents the effect of pain reduction. The dosage of the i-th drug used by the j-th player reflects the cost of the drug. For patient tolerance, The progress of wound healing demonstrates the contribution of the treatment plan to the recovery process;

[0055] S43. Establish a game equilibrium optimization model, and convert the player's payoff function... With strategy set Combined, determine the constraints for the equilibrium solution:

[0056] ;

[0057] in, For the strategy combination corresponding to the equilibrium solution, such that any player Without changing the strategies of other players, the payoff function The return value shall not be lower than that of any other strategy combination;

[0058] S44. Calculate the equilibrium solution of the gray wolf population using the Nash equilibrium algorithm:

[0059] .

[0060] Optionally, S5 includes:

[0061] S51. Define the population update rule for the gray wolf optimization algorithm, which updates the current position of each individual in the gray wolf population. Assuming the current pain management plan, initialize the positions of the best individual α wolf, the second best individual β wolf, and the third best individual δ wolf in the population:

[0062] ;

[0063] in, This represents the pain management scheme corresponding to the optimal solution in the t-th iteration of the population;

[0064] S52. Using the results of game equilibrium analysis As a dynamic constraint, the fitness function of individuals in the gray wolf population is adjusted. :

[0065] ;

[0066] in, This represents the fitness function based on a multivariate interaction model. Let represent the payoff value of the j-th individual in the game equilibrium analysis. These are the dynamic constraint weight coefficients;

[0067] S53. For each individual in the population Update the position and adjust the individual's location:

[0068]

[0069] ;

[0070] ;

[0071] in, This represents the new position after the (t+1)th iteration. C is the adjustment parameter, a is the linearly decreasing control variable during the iteration process, and r is a random number. Indicates the current position of the alpha wolf;

[0072] S54. Introduce a game equilibrium-based dynamic feedback mechanism to update the dynamic part of the individual fitness function:

[0073] ;

[0074] in, For dynamic feedback weighting coefficients, This indicates the cumulative deviation between wound healing speed and patient tolerance over a time frame, reflecting the directionality of dynamic adjustments to the treatment plan;

[0075] S55. Iterate and update until the stopping condition is met. In other words, the gray wolf optimization algorithm is considered to have converged when the rate of change of the fitness function of the α wolves in the population is lower than a threshold, and the globally optimal pain management solution is output. .

[0076] Optionally, S6 includes:

[0077] S61. The global optimal solution finally output by the Grey Wolf optimization algorithm. This translates into specific pain management plans, including the types of medications, dosage allocation, medication timing, and adjunctive treatment measures tailored to the individual needs of each patient.

[0078] S62. Adjust the description format of the generated pain management plan based on the patient's age, gender, pain score, and wound healing progress in the individualized patient dataset;

[0079] S63. Output the drug types and dosages as a daily medication list for the patient, listing the drug name, daily dosage, time of administration, and dosage per administration.

[0080] S64. Output the medication time as the specific time of day, and adjust the medication interval according to the patient's lifestyle and tolerance;

[0081] S65. Output adjunctive treatment measures as clear operational recommendations, providing specific treatment frequency, duration, and precautions based on the patient's wound healing progress;

[0082] S66. Integrate the output of drug types, dosage allocation, medication time and adjuvant treatment measures to form a pain management plan report. The pain management plan report shall include the scientific basis of the plan, indications and precautions, and indicate the basis and scope for dynamic adjustment.

[0083] The beneficial effects of this invention are:

[0084] (1) This invention models multiple variables in pain management as game participants and introduces game equilibrium theory to design a targeted payoff function. It comprehensively weighs multiple objectives such as pain relief effect, drug side effects and wound healing speed. Compared with traditional single-objective optimization techniques, the game equilibrium optimization mechanism of this invention can find a dynamic equilibrium point among multiple variables and combine individual patient data to achieve real-time adjustment of the plan. This solves the problem that the optimization results in the prior art are limited to a single variable and provides patients with a more scientific pain management plan.

[0085] (2) This invention proposes a multi-layer gray wolf population structure that combines global optimization with personalized adjustment. The drug action weights of gray wolf populations at different levels are adjusted by a dynamic mapping function, and a dynamic personalized adjustment factor is introduced to enable the optimization process to be precisely adjusted according to the real-time changes in the patient's condition. Compared with the traditional gray wolf optimization algorithm, the multi-layer structure not only significantly improves the global search capability of the population, but also enhances the flexibility and accuracy of the algorithm in generating personalized treatment plans by introducing dynamic constraints and weights of individualized patient data.

[0086] (3) In the process of optimizing the gray wolf, the present invention combines the real-time condition data of the patient and designs a dynamic feedback mechanism. By dynamically adjusting the fitness function, the optimization direction of the scheme is corrected in real time based on the differences in the wound healing speed and tolerance of the patient. This makes the generated pain management scheme always highly adaptable to the changes in the patient's condition. The dynamic feedback mechanism effectively solves the problem that traditional methods cannot respond to changes in the patient's condition and significantly improves the reliability and practicality of the treatment scheme. Attached Figure Description

[0087] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0088] Figure 1 This is a flowchart of a method for generating a wound pain management scheme based on game equilibrium optimization proposed in this invention;

[0089] Figure 2 This diagram illustrates the pain management optimization process that combines the gray wolf optimization algorithm with game equilibrium in the wound pain management scheme generation method proposed in this invention. Detailed Implementation

[0090] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0091] refer to Figures 1-2A method for generating a wound pain management scheme based on game equilibrium optimization includes the following steps:

[0092] S1. Collect individualized patient datasets;

[0093] S2. Based on the collected individualized patient datasets, construct a multivariate interaction model for wound pain management;

[0094] S3. Generate an initial gray wolf population using individualized patient datasets and a multivariate interaction model, where each individual in the initial gray wolf population represents a pain management strategy;

[0095] S4. Use game equilibrium theory to conduct equilibrium analysis on each pain management scheme in the gray wolf population. Treat the multiple variables in the pain management scheme as game participants, define the payoff function of each variable and calculate the equilibrium solution.

[0096] S5. Based on the gray wolf optimization algorithm, the gray wolf population is iteratively optimized. By introducing the game equilibrium analysis results as dynamic constraints, the position and fitness value of individuals in the gray wolf population are adjusted to gradually approach the globally optimal pain management scheme.

[0097] S6. The optimal pain management plan, which is finally output by the Grey Wolf optimization algorithm, is presented to clinical medical staff in the form of a suggestion.

[0098] In this embodiment, S1 includes:

[0099] S11. Collect the patient's age A and gender G, where G represents the patient's gender, with a value of 1 indicating male and 0 indicating female;

[0100] S12. Collect the patient's pain score. ,in, The patient's pain score at time t is calculated, with a range of [0, 10]. A higher score indicates a more severe pain.

[0101] S13. Collect information on the patient's medication use. :

[0102] ;

[0103] in, This represents the dosage of the i-th drug used by the patient at time t, and n represents the total number of drugs used by the patient at the same time.

[0104] S14. Collect data on the patient's wound healing progress. :

[0105] ;

[0106] in, This indicates the area of ​​the wound that has healed. Indicates the initial area of ​​the wound. The value range is [0, 1], and the closer the value is to 1, the higher the degree of wound healing;

[0107] S15. Construct a patient-specific dataset D:

[0108] .

[0109] In this embodiment, S2 includes:

[0110] S21. Construct a wound healing rate model based on the patient's individualized dataset D, and obtain the wound healing rate value by differencing the wound healing progress at adjacent time steps. ;

[0111] S22. Construct a drug effect model based on individualized patient datasets, including the drug dosage used by patients at time t. According to drug action coefficient Perform a weighted summation to obtain the drug effect value at time t. ;

[0112] S23. Construct a patient tolerance model based on individualized patient datasets, relating patient age A, patient gender G, and medication usage. Combined with tolerance weighting factor Adjust the patient's tolerance level to obtain the patient's tolerance value. ;

[0113] S24. Construct a pain relief effect model based on individualized patient datasets, including patient pain scores. Progress of wound healing Combined, the pain relief effect value is obtained. :

[0114] ;

[0115] in, The pain relief effect value over time t is used to reflect the level of pain relief in response to the interaction between wound healing process and pain perception. The pain score at time t is used to reflect the patient's pain level at that moment;

[0116] S25. Adjust the wound healing speed value. Drug efficacy value Patient tolerance value and pain relief effect value This is integrated into a multivariate interaction model to simultaneously characterize the interactions among the aforementioned variables:

[0117] ;

[0118] in, For a multivariate interaction model at time t, Used to measure the cumulative effect of multiple drug doses at that moment.

[0119] In this embodiment, S3 includes:

[0120] S31. Define the multi-layered gray wolf population structure:

[0121] ;

[0122] in, This represents the gray wolf population at level l, where L is the population level and each individual in the population... This indicates a pain management plan, including medication dosage. Medication time Adjunctive treatment measures and personalized adjustment factors ;

[0123] S32. Differentiated initialization of each population layer is performed by combining individualized patient datasets and a multivariate interaction model:

[0124] ;

[0125] in, This represents the drug dosage set for the j-th individual in the l-th population. The dosage of the i-th drug corresponding to the j-th individual in the l-th layer is generated by uniformly distributing it within the constraint range;

[0126] S33. Adjust the weights of drug effects in each population layer using a dynamic mapping function. :

[0127] ;

[0128] in, The effect weight of the i-th drug in the l-th layer population. For the model based on the number of layers l and multivariate interaction The function is dynamically adjusted to adapt to different levels of optimization needs;

[0129] S34. Assign dynamic, personalized adjustment factors to each individual in the population. The adjustment factor is learned based on patient-specific data D and fitness history:

[0130] ;

[0131] in, Let be the dynamic adjustment factor for the j-th individual in the l-th layer of the population. This represents the pain relief effect value over time t. For pain scores at time t, the numerator represents the overall contribution of the individual regimen to the patient's experience, while the denominator reflects the effect of drug dosage in the regimen.

[0132] In this embodiment, S4 includes:

[0133] S41. Treat each pain management scheme in the gray wolf population as a player in a game, and define the player set:

[0134] ;

[0135] Where N is the number of individuals in the gray wolf population, and each player's strategy includes drug dosage, drug administration time, and auxiliary treatment measures;

[0136] S42. Define each player based on individualized patient datasets and multivariate interaction models. payoff function :

[0137] ;

[0138] in, For players The payoff function is used to comprehensively evaluate the merits of pain management programs. These are weighting coefficients used to balance the objective variables in the payoff function. The pain relief effect value over time t represents the effect of pain reduction. The dosage of the i-th drug used by the j-th player reflects the cost of the drug. For patient tolerance, The progress of wound healing demonstrates the contribution of the treatment plan to the recovery process;

[0139] S43. Establish a game equilibrium optimization model, and convert the player's payoff function... With strategy set Combined, determine the constraints for the equilibrium solution:

[0140] ;

[0141] in, For the strategy combination corresponding to the equilibrium solution, such that any player Without changing the strategies of other players, the payoff function The return value shall not be lower than that of any other strategy combination;

[0142] S44. Calculate the equilibrium solution of the gray wolf population using the Nash equilibrium algorithm:

[0143] .

[0144] In this embodiment, S5 includes:

[0145] S51. Define the population update rule for the gray wolf optimization algorithm, which updates the current position of each individual in the gray wolf population. Assuming the current pain management plan, initialize the positions of the best individual α wolf, the second best individual β wolf, and the third best individual δ wolf in the population:

[0146] ;

[0147] in, This represents the pain management scheme corresponding to the optimal solution in the t-th iteration of the population;

[0148] S52. Using the results of game equilibrium analysis As a dynamic constraint, the fitness function of individuals in the gray wolf population is adjusted. :

[0149] ;

[0150] in, This represents the fitness function based on a multivariate interaction model. Let represent the payoff value of the j-th individual in the game equilibrium analysis. These are the dynamic constraint weight coefficients;

[0151] S53. For each individual in the population Update the position and adjust the individual's location:

[0152]

[0153] ;

[0154] ;

[0155] in, This represents the new position after the (t+1)th iteration. C is the adjustment parameter, a is the linearly decreasing control variable during the iteration process, and r is a random number. Indicates the current position of the alpha wolf;

[0156] S54. Introduce a game equilibrium-based dynamic feedback mechanism to update the dynamic part of the individual fitness function:

[0157] ;

[0158] in, For dynamic feedback weighting coefficients, This indicates the cumulative deviation between wound healing speed and patient tolerance over a time frame, reflecting the directionality of dynamic adjustments to the treatment plan;

[0159] S55. Iterate and update until the stopping condition is met. In other words, the gray wolf optimization algorithm is considered to have converged when the rate of change of the fitness function of the α wolves in the population is lower than a threshold, and the globally optimal pain management solution is output. .

[0160] In this embodiment, S6 includes:

[0161] S61. The global optimal solution finally output by the Grey Wolf optimization algorithm. This translates into specific pain management plans, including the types of medications, dosage allocation, medication timing, and adjunctive treatment measures tailored to the individual needs of each patient.

[0162] S62. Adjust the description format of the generated pain management plan based on the patient's age, gender, pain score, and wound healing progress in the individualized patient dataset;

[0163] S63. Output the drug types and dosages as a daily medication list for the patient, listing the drug name, daily dosage, time of administration, and dosage per administration.

[0164] S64. Output the medication time as the specific time of day, and adjust the medication interval according to the patient's lifestyle and tolerance;

[0165] S65. Output adjunctive treatment measures as clear operational recommendations, providing specific treatment frequency, duration, and precautions based on the patient's wound healing progress;

[0166] S66. Integrate the output of drug types, dosage allocation, medication time and adjuvant treatment measures to form a pain management plan report. The pain management plan report shall include the scientific basis of the plan, indications and precautions, and indicate the basis and scope for dynamic adjustment.

[0167] Example 1:

[0168] Example: In City A, the burn unit of a tertiary hospital in City A admitted a 45-year-old male patient, Mr. Zhang, who suffered extensive second-degree burns due to an accident, with the burn area accounting for 25% of his body surface area. After admission, Mr. Zhang exhibited severe wound pain, scoring 9 out of 10 (out of 10 being the most severe pain). Traditional pain relief methods had limited effect on him, and drug side effects such as nausea and drowsiness significantly affected his recovery process. The medical team hoped to alleviate the patient's suffering through a more scientific pain management plan, while avoiding the overuse and side effects of drugs and improving wound healing efficiency.

[0169] The hospital chose to use the method of this invention to generate a pain management plan in order to achieve dynamic, personalized, and scientific treatment. The following is the actual pain management process of Mr. Zhang during the first week after his admission:

[0170] After the patient was admitted to the hospital, the system collected Zhang’s individual data, including age (45 years old), gender (male), wound pain score (initial value of 9 points), medication use (initially oxycodone 15 mg / day, once every 8 hours), and wound healing progress (initial healing rate of 0%, total wound area of ​​450 square centimeters, and unhealed area of ​​450 square centimeters).

[0171] Through data analysis, a multivariate interaction model of Zhang was constructed, in which drug dosage, tolerability, pain relief effect and wound healing speed were included in the model. The preliminary assessment of the model showed that the drug use had a high influence coefficient on the pain relief effect, but significant side effects. At the same time, the tolerability model showed that the patient had a slight intolerance trend to the current dose.

[0172] The system initialized a multi-layered gray wolf population, with 20 individuals in each layer. Each individual represents a different pain management plan, including drug type (oxycodone, ibuprofen), dosage (10-30 mg / day), dosing time (6-12 hour interval), and adjunctive treatment measures (such as physical cooling and psychological counseling). Through game equilibrium analysis, a multi-objective optimization function was defined, with pain relief effect weighted at 0.5, side effect control weighted at 0.3, and wound healing speed weighted at 0.2.

[0173] After the optimization process took 2 minutes, the system output Zhang's first version of the optimal pain management plan:

[0174] Drug type: Oxycodone; Dosage: 12 mg / day; Dosage interval: 8 hours;

[0175] Supportive treatment measures: physical cooling twice a day, 30 minutes each time; psychological counseling for 20 minutes daily;

[0176] Objective: To reduce pain while keeping side effects within tolerable limits and improving healing efficiency.

[0177] On the third day, the system collected Zhang's real-time data. The wound healing progress reached 10% (the unhealed part was 405 square centimeters), and the pain score dropped to 6 points. However, the patient experienced mild drowsiness. The system combined with the dynamic feedback mechanism to automatically adjust and optimize the model, further reducing the drug dosage to 10 mg / day and suggesting extending the physical cooling time to 40 minutes.

[0178] The data analysis results for Zhang one week later showed:

[0179] The pain score dropped to 3 (compared to a score above 5 using traditional methods).

[0180] Wound healing progress reached 25% (compared to 15% with conventional methods during the same period);

[0181] Side effects are significantly reduced, with no obvious drowsiness or nausea (the incidence of side effects is 60% with traditional methods).

[0182] The training sample was used as evidence. Data source: historical hospital data, including 100 burn patients (aged 18-65 years) as the training sample. 50 patients were treated with the traditional method and 50 patients were treated with the method of this invention. Results: Under the method of this invention, the average pain score of patients decreased by 30%, the wound healing rate increased by 20%, and the incidence of side effects decreased by 50%.

[0183] As can be seen from this embodiment, the method of the present invention is significantly superior to traditional methods in terms of dynamic personalized adjustment, optimization efficiency and effect. Its scientificity and feasibility have been clinically verified, providing an important reference for technological innovation in the field of pain management.

[0184] This invention models multiple variables in pain management as game participants and introduces game equilibrium theory to design a targeted payoff function. It comprehensively weighs multiple objectives, including pain relief, drug side effects, and wound healing speed. Compared with traditional single-objective optimization techniques, the game equilibrium optimization mechanism of this invention can find a dynamic equilibrium point among multiple variables and combine individualized patient data to achieve real-time adjustment of the plan. This solves the problem that the optimization results in the prior art are limited to a single variable, and provides patients with a more scientific pain management plan.

[0185] This invention proposes a multi-layered gray wolf population structure that combines global optimization with personalized adjustment. By using a dynamic mapping function to adjust the drug action weights of gray wolf populations at different levels, and by introducing a dynamic personalized adjustment factor, the optimization process can be precisely adjusted according to the real-time changes in the patient's condition. Compared with the traditional gray wolf optimization algorithm, the multi-layered structure not only significantly improves the global search capability of the population, but also enhances the flexibility and accuracy of the algorithm in generating personalized treatment plans by introducing dynamic constraints and weights based on individualized patient data.

[0186] This invention incorporates real-time patient data during the gray wolf optimization process and designs a dynamic feedback mechanism. By dynamically adjusting the fitness function, the optimization direction of the plan is corrected in real time based on the differences in the patient's wound healing speed and tolerance. This ensures that the generated pain management plan always maintains a high degree of adaptability to changes in the patient's condition. The dynamic feedback mechanism effectively solves the problem that traditional methods cannot respond to changes in the patient's condition, and significantly improves the reliability and practicality of the treatment plan.

[0187] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for generating a wound pain management scheme based on game equilibrium optimization, characterized in that, Includes the following steps: S1. Collect individualized patient datasets; S2. Based on the collected individualized patient datasets, construct a multivariate interaction model for wound pain management; S3. Generate an initial gray wolf population using individualized patient datasets and a multivariate interaction model, where each individual in the initial gray wolf population represents a pain management strategy; S4. Use game equilibrium theory to conduct equilibrium analysis on each pain management scheme in the gray wolf population. Treat the multiple variables in the pain management scheme as game participants, define the payoff function of each variable and calculate the equilibrium solution. S5. Based on the gray wolf optimization algorithm, the gray wolf population is iteratively optimized. By introducing the game equilibrium analysis results as dynamic constraints, the position and fitness value of individuals in the gray wolf population are adjusted to gradually approach the globally optimal pain management scheme. S6. The optimal pain management plan, which is finally output by the Grey Wolf optimization algorithm, is presented to clinical medical staff in the form of a suggestion.

2. The method for generating a wound pain management scheme based on game equilibrium optimization according to claim 1, characterized in that, S1 includes: S11. Collect the patient's age A and gender G, where G represents the patient's gender, with a value of 1 indicating male and 0 indicating female; S12. Collect the patient's pain score. ,in, The patient's pain score at time t is calculated, with a value range of [0, 10]. The higher the score, the more severe the pain. S13. Collect information on the patient's medication use. : ； in, This represents the dosage of the i-th drug used by the patient at time t, and n represents the total number of drugs used by the patient at the same time. S14. Collect data on the patient's wound healing progress. : ； in, This indicates the area of ​​the wound that has healed. Indicates the initial area of ​​the wound. The value range is [0,1], and the closer the value is to 1, the higher the degree of wound healing; S15. Construct a patient-specific dataset D: 。 3. The method for generating a wound pain management scheme based on game equilibrium optimization according to claim 2, characterized in that, S2 includes: S21. Construct a wound healing rate model based on the patient's individualized dataset D, and obtain the wound healing rate value by differencing the wound healing progress at adjacent time steps. ; S22. Construct a drug effect model based on individualized patient datasets, including the drug dosage used by patients at time t. According to drug action coefficient Perform a weighted summation to obtain the drug effect value at time t. ; S23. Construct a patient tolerance model based on individualized patient datasets, relating patient age A, patient gender G, and medication usage. Combined with tolerance weighting factor Adjust the patient's tolerance level to obtain the patient's tolerance value. ; S24. Construct a pain relief effect model based on individualized patient datasets, including patient pain scores. Progress of wound healing Combined, the pain relief effect value is obtained. : ； in, The pain relief effect value over time t is used to reflect the level of pain relief in response to the interaction between wound healing process and pain perception. The pain score at time t is used to reflect the patient's pain level at that moment; S25. Adjust the wound healing speed value. Drug efficacy value Patient tolerance value and pain relief effect value This is integrated into a multivariate interaction model to simultaneously characterize the interactions among the aforementioned variables: ； in, For a multivariate interaction model at time t, Used to measure the cumulative effect of multiple drug doses at that moment.

4. The method for generating a wound pain management scheme based on game equilibrium optimization according to claim 1, characterized in that, S3 includes: S31. Define the multi-layered gray wolf population structure: ； in, This represents the gray wolf population at level l, where L is the population level and each individual in the population... This indicates a pain management plan, including medication dosage. Medication time Adjunctive treatment measures and personalized adjustment factors ; S32. Differentiated initialization of each population layer is performed by combining individualized patient datasets and a multivariate interaction model: ； in, This represents the drug dosage set for the j-th individual in the l-th population. The dosage of the i-th drug corresponding to the j-th individual in the l-th layer is generated by uniformly distributing it within the constraint range; S33. Adjust the weights of drug effects in each population layer using a dynamic mapping function. : ； in, The effect weight of the i-th drug in the l-th layer population. For the model based on the number of layers l and multivariate interaction The function is dynamically adjusted to adapt to different levels of optimization needs; S34. Assign dynamic, personalized adjustment factors to each individual in the population. The adjustment factor is learned based on patient-specific data D and fitness history: ； in, Let be the dynamic adjustment factor for the j-th individual in the l-th layer of the population. This represents the pain relief effect value over time t. For pain scores at time t, the numerator represents the overall contribution of the individual regimen to the patient's experience, while the denominator reflects the effect of drug dosage in the regimen.

5. The method for generating a wound pain management scheme based on game equilibrium optimization according to claim 1, characterized in that, S4 includes: S41. Treat each pain management scheme in the gray wolf population as a player in a game, and define the player set: ； Where N is the number of individuals in the gray wolf population, and each player's strategy includes drug dosage, drug administration time, and auxiliary treatment measures; S42. Define each player based on individualized patient datasets and multivariate interaction models. The profit function : ； in, For players The payoff function is used to comprehensively evaluate the merits of pain management programs. These are weighting coefficients used to balance the objective variables in the payoff function. The pain relief effect value over time t represents the effect of pain reduction. The dosage of the i-th drug used by the j-th player reflects the cost of the drug. For patient tolerance, The progress of wound healing demonstrates the contribution of the treatment plan to the recovery process; S43. Establish a game equilibrium optimization model, and convert the player's payoff function... With strategy set Combined, determine the constraints for the equilibrium solution: ； in, For the strategy combination corresponding to the equilibrium solution, such that any player Without changing the strategies of other players, the payoff function The return value shall not be lower than that of any other strategy combination; S44. Calculate the equilibrium solution of the gray wolf population using the Nash equilibrium algorithm: 。 6. The method for generating a wound pain management scheme based on game equilibrium optimization according to claim 1, characterized in that, S5 includes: S51. Define the population update rule for the gray wolf optimization algorithm, which updates the current position of each individual in the gray wolf population. Assuming the current pain management plan, initialize the positions of the best individual α wolf, the second best individual β wolf, and the third best individual δ wolf in the population: ； in, This represents the pain management scheme corresponding to the optimal solution in the t-th iteration of the population; S52. Using the results of game equilibrium analysis As a dynamic constraint, the fitness function of individuals in the gray wolf population is adjusted. : ； in, This represents the fitness function based on a multivariate interaction model. Let represent the payoff value of the j-th individual in the game equilibrium analysis. These are the dynamic constraint weight coefficients; S53. For each individual in the population Update the position and adjust the individual's location: ； ； in, This represents the new position after the (t+1)th iteration. C is the adjustment parameter, a is the linearly decreasing control variable during the iteration process, and r is a random number. Indicates the current position of the alpha wolf; S54. Introduce a game equilibrium-based dynamic feedback mechanism to update the dynamic part of the individual fitness function: ； in, For dynamic feedback weighting coefficients, This indicates the cumulative deviation between wound healing speed and patient tolerance over a time frame, reflecting the directionality of dynamic adjustments to the treatment plan; S55. Iterate and update until the stopping condition is met. In other words, the gray wolf optimization algorithm is considered to have converged when the rate of change of the fitness function of the α wolves in the population is lower than a threshold, and the globally optimal pain management solution is output. .

7. The method for generating a wound pain management scheme based on game equilibrium optimization according to claim 1, characterized in that, S6 includes: S61. The global optimal solution finally output by the Grey Wolf optimization algorithm. This translates into specific pain management plans, including the types of medications, dosage allocation, medication timing, and adjunctive treatment measures tailored to the individual needs of each patient. S62. Adjust the description format of the generated pain management plan based on the patient's age, gender, pain score, and wound healing progress in the individualized patient dataset; S63. Output the drug types and dosages as a daily medication list for the patient, listing the drug name, daily dosage, time of administration, and dosage per administration. S64. Output the medication time as the specific time of day, and adjust the medication interval according to the patient's lifestyle and tolerance; S65. Output adjunctive treatment measures as clear operational recommendations, providing specific treatment frequency, duration, and precautions based on the patient's wound healing progress; S66. Integrate the output of drug types, dosage allocation, medication time and adjuvant treatment measures to form a pain management plan report. The pain management plan report shall include the scientific basis of the plan, indications and precautions, and indicate the basis and scope for dynamic adjustment.

Images