Biomedical Experiment Optimization System Based on Graphical Model

Abstract

The present invention relates to the field of test optimization technology, specifically to a biomedical test optimization system based on a graphical model, and the biomedical test optimization system based on a graphical model includes a data collection module, a graphical weight adjustment module, a path exploration module, and an optimization adjustment module. In the present invention, by synchronously integrating drug concentration and biomarker reaction data through real-time data, not only can the entire test be covered, but also real-time updating and accurate mapping of data can be ensured, greatly enhancing the reliability and effectiveness of the test, and dynamic data can update the weights of nodes and edges in the graphical model in real time, so that the model can accurately reflect the actual interaction effects between drugs, and the updated data is used for path analysis and pharmacological action relationship marking, further enhancing the pertinence and scientificity of the test, being able to quickly identify key variables and action pathways, optimizing drug combination schemes, significantly improving R&D efficiency and cost control, and greatly shortening the drug R&D cycle.

Description

Technical Field

[0001] The present invention relates to the technical field of test optimization, and in particular to a biomedical test optimization system based on a graphic model. Background Art

[0002] The field of experimental optimization technology involves the use of scientific methods, algorithms, and systematic tools to improve and enhance the design and execution of research experiments. The core goal is to optimize the parameters and conditions of the experiment to reduce resource consumption, improve data quality and efficiency, and make the experimental results more reliable and effective. It involves the application of a variety of technologies, including statistical analysis, machine learning, computational models, and automated processes to ensure the accuracy and repeatability of the experimental process when controlling variables and processing large amounts of data. It has a wide range of applications in drug development, biotechnology, agricultural science, and other fields.

[0003] Among them, the biomedical test optimization system is a specific topic used in biomedical research and development to improve the efficiency and output quality of research experiments. By integrating advanced data analysis tools and automation technology, the system can help researchers design more accurate experimental plans and optimize experimental conditions, thereby promoting drugs from the laboratory to the market faster. In addition, the system can also be used to monitor the progress of the experiment and adjust the experimental parameters in real time to cope with unforeseen variables and challenges, ultimately achieving the goal of improving the efficiency of the entire R&D cycle and reducing costs.

[0004] The main shortcoming of the existing technology in actual operation is that the processing and feedback mechanism of real-time data during the test is not flexible enough. Traditional test optimization technology often relies on preset static models and parameters, which is difficult to deal with sudden variables and complex biological reactions that occur during the test, resulting in insufficient flexibility and adaptability of the test, affecting the accuracy and repeatability of the test results. In addition, data processing in the existing technology is usually delayed, and data updates and analysis cannot be carried out synchronously, resulting in a slow response to the adjustment of test conditions, and unable to immediately feedback key information during the test. There is a lack of dynamic adjustment capabilities, which often leads to waste of resources, increased test costs, and prolonged the time it takes for drugs to go from the laboratory to the market. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a biomedical experiment optimization system based on a graphic model.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: A biomedical experiment optimization system based on a graphical model comprises:

[0007] The data collection module configures the data interface, collects drug concentration and biomarker data, integrates the collected information and performs real-time data synchronization, covers the entire test process, and outputs a biological dynamic data set;

[0008] The graph weight adjustment module adjusts the nodes and edges in the graph by weight update according to the biodynamic data set, recalculates the connection strength between nodes each time the data is updated, reflects the interaction effect between drugs in real time according to the calculation results, maps the real-time interaction between drugs, and generates an adjusted interaction weight graph;

[0009] The pathway exploration module uses the adjusted interaction weight graph to perform pathway analysis, identify key action pathways between drugs, analyze the pharmacological action relationships of pathway nodes based on the identification results, mark key nodes and links, and output key pathway exploration results;

[0010] The optimization and adjustment module evaluates the effect of the combination scheme by simulating the impact of the drug combination on the pathological process based on the critical path exploration results, makes targeted adjustments based on the evaluation results, optimizes the drug dosage and combination scheme, and forms a dynamic optimization plan for medical trials.

[0011] As a further solution of the present invention, the steps of acquiring the biological dynamic data set are:

[0012] Configure data interfaces and sensors to automatically collect drug concentration and biomarker data and obtain raw data sets;

[0013] According to the original data set, key parameters are extracted, data filtering is performed and the data is preliminarily calibrated, using the formula, ;

[0014] Calculate data processing efficiency index , and obtain comprehensive analysis data, among which, represents the drug concentration value after calibration, represents the biomarker response intensity value, Represents the time window size;

[0015] The comprehensive analysis data are used to perform pattern recognition and statistical analysis, identify and verify the response pattern of biomarkers, capture the entire test response, and generate a biological dynamic data set.

[0016] As a further solution of the present invention, the steps for calculating the connection strength and direction between nodes are:

[0017] Based on the biological dynamic data set, extract drug interaction data that affects node connections, perform preliminary screening on nodes and edges, and generate preliminary screening results;

[0018] Based on the preliminary screening results, the weights are updated using the formula: ;

[0019] Calculate weights for new nodes and edges , generate updated weight results, where, is the old weight, is the learning rate, is the frequency of new interactions between nodes, and Is a node and The degree of

[0020] Using the updated weight result, the formula is adopted: ;

[0021] Recalculate the connection strength between nodes , output the recalculation result of the connection strength between nodes, where, Is a node and The new weights between is the maximum value among all new weights.

[0022] As a further solution of the present invention, the step of obtaining the adjusted interaction weight map is:

[0023] Integrate the recalculation results of the connection strength between the nodes, convert the data to map the interaction effects between the drugs, and generate visual data preparation results by assigning visual weights to each node to reflect the strength of the effect in the network;

[0024] Applying the visualization data preparation results, dynamically updating the drug interaction network diagram, revealing real-time interaction data by adjusting the size of nodes and the thickness of edges, and generating an updated network diagram structure;

[0025] According to the updated network graph structure, visualization processing is performed to reflect the interaction changes caused by data update in real time, and an adjusted interaction weight graph is generated.

[0026] As a further embodiment of the present invention, the step of identifying the key action pathway between drugs is:

[0027] calling the adjusted interaction weight graph, analyzing the mutual relationship between drugs based on the pharmacological action intensity, and outputting basic analysis data;

[0028] Based on the adjusted interaction weight graph and the basic analysis data, the formula is adopted: ;

[0029] Calculate new path weights , generate the optimal path analysis results, where represents the original path weight, Represents a connected node and the relative importance of

[0030] Analyze the optimal path analysis results, identify key action pathways, including key nodes and connections in the path, and generate key action pathway confirmation results.

[0031] As a further solution of the present invention, the steps for obtaining the critical path exploration results are:

[0032] Perform pharmacological action analysis on each node in the key action pathway confirmation results, evaluate the interaction between drugs through pharmacokinetic and drug target data, and output the pharmacological action analysis results;

[0033] Based on the pharmacological action analysis results, key nodes are determined, drug action relationships are combined with key nodes and links, and the positions of drugs in the graph are marked to obtain drug marking information;

[0034] The drug labeling information is integrated, the drug interaction network is graphically displayed, the pharmacological effect of each key node is recorded, and the key path exploration results are output.

[0035] As a further solution of the present invention, the steps for evaluating the effect of the combination solution are:

[0036] According to the critical path exploration results, drugs associated with the target pathological process are selected to generate a drug combination model;

[0037] Based on the drug combination model, pathological simulation is performed using the formula, ;

[0038] Calculating the pathological impact of drug combinations , generate simulated drug effect results, where Represents the concentration value of the drug. represents the duration of treatment, represents the half-life of the drug;

[0039] Analyze the simulated drug effect results, perform statistical analysis, compare the drug effects with the expected goals, confirm the actual efficacy of the scheme, and obtain the combination scheme effect evaluation results.

[0040] As a further solution of the present invention, the steps for obtaining the dynamic optimization plan of the medical experiment are:

[0041] Based on the combination effect evaluation results, analyze the drug combination performance, determine the drug dosage and combination parameters that need to be adjusted through pharmacodynamic and clinical data comparison, and generate a drug parameter adjustment list;

[0042] According to the drug parameter adjustment list, the drug dosage is adjusted, the combination scheme is reconfigured, and the drug combination configuration is optimized by simulating the efficacy prediction under different dosages and combinations to generate an optimized drug combination scheme;

[0043] Based on the optimized drug combination scheme, a new clinical trial design is carried out, including randomization, blinding setting and control group design, to generate a dynamic optimization plan for medical trials.

[0044] Compared with the prior art, the advantages and positive effects of the present invention are:

[0045] In the present invention, the drug concentration and biomarker reaction data are integrated synchronously through real-time data, which not only covers the entire test process, but also ensures real-time updating and accurate mapping of data, greatly enhancing the reliability and effectiveness of the test. Dynamic data updates the weights of nodes and edges in the graphical model in real time, so that the model can accurately reflect the actual interaction effects between drugs. The updated data is used for path analysis and pharmacological action relationship marking, further enhancing the pertinence and scientific nature of the test, and can quickly identify key variables and action pathways, optimize drug combination schemes, significantly improve R&D efficiency and cost control, and greatly shorten the drug development cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a system flow chart of the present invention;

[0047] Figure 2 A flowchart for obtaining a biological dynamic data set of the present invention;

[0048] Figure 3 The flowchart of the calculation of the connection strength and direction between nodes of the present invention;

[0049] Figure 4 A flowchart of obtaining the adjusted interaction weight graph of the present invention;

[0050] Figure 5 A flowchart for identifying key action pathways between drugs of the present invention;

[0051] Figure 6 A flowchart for obtaining the key path exploration results of the present invention;

[0052] Figure 7 A flow chart for evaluating the effect of the combination scheme of the present invention;

[0053] Figure 8 The present invention is a flowchart for obtaining the dynamic optimization scheme of medical trials. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0055] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0056] See also Figure 1 , the biomedical trial optimization system based on graphical models includes:

[0057] The data collection module configures the data interface, collects drug concentration and biomarker data, integrates the collected information and performs real-time data synchronization, covers the entire test process, and outputs a biological dynamic data set;

[0058] The graph weight adjustment module adjusts the nodes and edges in the graph by updating the weights according to the biological dynamic data set. Each time the data is updated, the connection strength between nodes is recalculated. The interactive effects between drugs are reflected in real time according to the calculation results, the real-time interactions between drugs are mapped, and the adjusted interactive weight graph is generated.

[0059] The pathway exploration module uses the adjusted interaction weight graph to perform pathway analysis, identify key action pathways between drugs, analyze the pharmacological action relationships of pathway nodes based on the identification results, mark key nodes and links, and output key pathway exploration results;

[0060] Based on the results of critical path exploration, the optimization and adjustment module simulates the impact of drug combinations on pathological processes, evaluates the effectiveness of the combination scheme, makes targeted adjustments based on the evaluation results, optimizes drug dosages and combination schemes, and forms a dynamic optimization plan for medical trials.

[0061] The biological dynamic data set includes drug concentration data, biomarkers and time point data. The adjusted interaction weight graph includes node update data, edge weight update data and real-time interaction graph. The key path exploration results include action pathway diagram, node pharmacological relationship analysis results and key connection point marking results. The dynamic optimization plan for medical trials includes dosage adjustment optimization plan, combination optimization proposal and effect evaluation report.

[0062] See also Figure 2 , the steps to obtain the biological dynamic data set are:

[0063] Configure data interfaces and sensors to automatically collect drug concentration and biomarker data and obtain raw data sets;

[0064] Configure data interfaces and sensor devices. The main tasks include selecting appropriate hardware interfaces and sensor types to realize the automatic collection of drug concentration and biomarker data. First, select sensors with specific sensitivity and range according to the chemical properties of drugs and biomarkers, such as using a mass spectrometer to detect drug concentration and using an enzyme reader to capture biomarker reactions; second, install and calibrate sensors to ensure the accuracy and reliability of data, including environmental adaptability testing and signal calibration; finally, configure the data interface system, including the installation and debugging of software and hardware, to ensure that the equipment can monitor and record changes in drugs and biomarkers in real time. The data interface needs to be able to process signals from different sensors and convert the signals into digital information for subsequent analysis. Through this process, the original data set is obtained, laying the foundation for subsequent data processing and analysis.

[0065] According to the original data set, key parameters are extracted, data filtering is performed and the data is preliminarily calibrated. The formula is used. ;

[0066] Calculate data processing efficiency index , and obtain comprehensive analysis data, among which, represents the drug concentration value after calibration, represents the biomarker response intensity value, Represents the time window size;

[0067] The drug concentration measured at the data collection point, =0.05mg / L; the original measured biomarker response intensity, is 0.2, then ; Represents the time window. If the time window is set to 4 seconds, then .

[0068] Substitute the numerical calculation:

[0069] ;

[0070] The results show that under specific drug concentrations, biomarker response intensity and time windows, the combined effect of various factors is In fact, it reflects the comprehensive ability of data generation and processing under experimental conditions. This efficiency indicator is a key metric used to evaluate the performance of the entire data processing pipeline under the current experimental settings. Researchers can understand the efficiency of data generation and adjust the experimental conditions or data processing pipeline when necessary to optimize the experimental design and data quality.

[0071] Using comprehensive analytical data, pattern recognition and statistical analysis are performed to identify and verify the response patterns of biomarkers, capture the entire test response, and generate a biological dynamic data set;

[0072] The comprehensive analysis data are used for pattern recognition and statistical analysis. First, the comprehensive analysis data are preprocessed, including data cleaning and normalization, to improve data quality. Next, statistical methods are used to perform descriptive analysis on the data, such as calculating statistical indicators such as mean and variance. The indicators help to initially understand the basic characteristics of the data set. Then, machine learning algorithms such as cluster analysis and principal component analysis (PCA) are applied to perform pattern recognition on the data to identify the main variables and trends in the data. These patterns will be used to classify the responses of biomarkers. Finally, by analyzing the results, a biodynamic data set is generated to ensure that the data can accurately reflect the full response of the trial, thereby supporting further scientific research and clinical decision-making.

[0073] See also Figure 3 , the calculation steps of the connection strength and direction between nodes are:

[0074] Based on the biological dynamics data set, drug interaction data that affects node connections are extracted, and preliminary screening is performed on nodes and edges to generate preliminary screening results;

[0075] Drug interaction data that affects node connections are extracted from the biodynamic data set, and preliminary screening is performed on key nodes and edges to generate preliminary screening results. During the process, special attention is paid to those nodes and edges that show significant changes in the most recent data update cycle. Using real-time updated drug effect data, the nodes with the greatest impact are selected for key processing. By quantitatively analyzing the interaction strength and change trend between nodes, nodes and edges with significant drug interactions are screened out. The data is obtained by counting the frequency of drug interactions in the past 24 hours. The statistical data is sorted according to the frequency of drug reactions between nodes, and the top 20% of node pairs are selected for in-depth analysis. The generated preliminary screening results will directly affect the subsequent weight adjustment and connection strength calculation.

[0076] Based on the preliminary screening results, the weights are updated using the formula: ;

[0077] Calculate weights for new nodes and edges , generate updated weight results, where, is the old weight, is the learning rate, is the frequency of new interactions between nodes, and Is a node and The degree of

[0078] Get data , , , , .

[0079] Compute the product of node degrees:

[0080] ;

[0081] Calculate the frequency to degree ratio:

[0082] ;

[0083] Apply learning rate:

[0084] ;

[0085] The final result for ,The results show that under the current dataset and settings, the new weights are slightly lower than the old weights, ,reflecting that the interaction strength between nodes during this period has ,weakened compared to before.

[0086] Using the updated weight results, the formula is adopted, ;

[0087] Recalculate the connection strength between nodes , output the recalculation result of the connection strength between nodes, where, Is a node and The new weights between is the maximum value among all new weights;

[0088] According to the above calculation results (obtained from the previous calculation), and the maximum weight value =0.600, the calculation process is , the final normalized connection strength is 0.937. The results show that considering the maximum weight, the node and The connection strength between them is very close to the maximum potential strength, indicating that the interaction between them is strong and occupies an important position in the network. This result is very important for understanding and optimizing the key connection points in the network structure.

[0089] See also Figure 4 , the steps to obtain the adjusted interaction weight map are:

[0090] Integrate the recalculation results of the connection strength between nodes, transform the data to map the interaction effects between drugs, and generate visual data preparation results by assigning visual weights to each node to reflect the strength of the effect in the network;

[0091] Integrate the previous calculation results of the strength of the connection between nodes. These results are obtained based on the node interaction strength that has changed significantly during the latest data update cycle. During the execution process, special attention is paid to those nodes that show significant changes due to drug interactions. The data is converted into a visual format and a visual weight is assigned to each node based on its strength of action in the network. The process includes calculating the visual weight of each node. The weight is quantified by evaluating the relative strength of the node's connection strength to other nodes in the network. The purpose of generating visual data preparation results is to provide the necessary input parameters for the next step of graphics update. This result will directly affect the visual performance of the network diagram.

[0092] Apply the visual data preparation results to dynamically update the drug interaction network diagram, reveal the real-time interaction data by adjusting the size of nodes and the thickness of edges, and generate an updated network diagram structure;

[0093] Apply the visualized data results to update the drug interaction network diagram, dynamically adjust the position of nodes and the thickness of edges in the network diagram to reflect the latest interaction data. The size of the node is adjusted according to the visual weight of its connection strength, and the thickness of the edge is expressed according to the real-time interaction strength between different drugs. The process includes using graphics algorithms to automatically adjust the visual properties of nodes and edges to ensure that the network diagram can reflect the changes in interactions between drugs in real time after each data update. The purpose of generating an updated network diagram structure is to provide an intuitive view of drug interaction effects, so that observers can understand the dynamic relationship between drugs.

[0094] According to the updated network graph structure, visualization processing is performed to reflect the interaction changes caused by data updates in real time and generate an adjusted interaction weight graph;

[0095] Using visualization tools such as Graphviz or Gephi, the interaction weight graph is rendered according to the updated network graph structure. The process includes importing the updated network data into the visualization software and applying a predefined visual style to render the graph, ensuring that the visual representation of each node and edge is consistent with its data-driven weight. The process of generating the final adjusted interaction weight graph includes not only visual rendering but also performance optimization to ensure fast rendering even with large amounts of data. The final result is a dynamic network graph that can reflect the interaction effects between drugs in real time, providing researchers with a powerful tool to observe and analyze the interactions between drugs.

[0096] See also Figure 5 , the steps for identifying key action pathways between drugs are:

[0097] Call the adjusted interaction weight graph to analyze the relationship between drugs based on the intensity of pharmacological effects and output basic analysis data;

[0098] The interactive weight graph with adjusted weights is called. The weights are adjusted based on the pharmacological effect strength and historical interaction data between drugs. For example, the effect strength from drug A to drug B may be adjusted up or down based on the response of the patient population and the results of clinical trials. These data are updated and integrated in real time by multiple medical databases through API interfaces, ensuring the timeliness and accuracy of the data. Using these precise data can provide a solid foundation for subsequent path analysis. At the same time, each interaction between drugs is regarded as an independent data point, ensuring the scientificity and practicality of the entire network graph in practical applications.

[0099] Based on the adjusted interaction weight graph and basic analysis data, the formula is used. ;

[0100] Calculate new path weights , generate the optimal path analysis results, where represents the original path weight, Represents a connected node and the relative importance of

[0101] Represents the original pathway weight, specifically expressed as drug and medications The original interaction strength between is empirical data extracted from the medical database and is assumed to be 0.05; represents the relative importance index between drug i and drug j, which is usually determined based on the clinical impact and frequency of use of the drug, and is assumed to be 2 (with values ​​ranging from 1 to 5, indicating increasing importance). First, calculate get , take the absolute value and square root to get , then calculate Finally, substituting these two results into the formula, we get The results show that according to the original interaction strength between drugs and their relative importance, the newly calculated pathway weight is 0.2342, which can be used to determine the key action pathways between drugs according to the weight.

[0102] Analyze the optimal path analysis results, identify the key action pathways, including the key nodes and connections in the pathways, and generate key action pathway confirmation results;

[0103] After obtaining the results of the optimal pathway analysis, the key action pathways between drugs are identified. This includes a detailed analysis of the weight of each drug node and determining which connections play a key role in the drug action network. Through network analysis, large-scale data sets are processed and high-weight nodes and connections are quickly identified. For example, if a drug plays a core role in the treatment of a disease, its related connections will appear as important connections on the critical path. During the analysis, the frequency and efficacy of drug use, as well as possible synergistic or antagonistic effects between different drugs, are also considered. These are factors that must be considered comprehensively when determining the critical path. In this way, the action network between drugs can be accurately depicted, providing a scientific basis for clinical decision-making.

[0104] See also Figure 6 , the steps to obtain the critical path exploration results are:

[0105] Conduct pharmacological analysis on each node in the key action pathway confirmation results, evaluate the interactions between drugs through pharmacokinetic and drug target data, and output the pharmacological analysis results;

[0106] When conducting pharmacological action analysis, we first collect drug data based on existing clinical research and drug interaction databases. These data include drug metabolic pathways, pharmacodynamic properties, and known drug interactions. This information is extracted from a large amount of medical literature using data mining technology. Statistical methods are then used to analyze the data to determine which drug combinations will produce synergistic or antagonistic effects in treatment. This approach allows us to accurately evaluate the pharmacological effects of each drug and their mutual influence in specific disease states, which provides a scientific basis for marking key nodes and connections.

[0107] Based on the results of pharmacological action analysis, key nodes are determined, the drug action relationships are combined with key nodes and links, the positions of drugs in the graph are marked, and drug marking information is obtained;

[0108] When confirming the key nodes, the importance score of the drug node can be calculated To further express the importance of the node, the formula is used Calculate, where Indicates the number of events in which a node directly affects other drugs in the drug network, represents the total number of drug interaction events. Assumptions and ,but This score reflects the activity of a drug in the entire treatment network, thus intuitively reflecting its importance.

[0109] Integrate drug labeling information, graphically display drug interaction networks, record the pharmacological effects of each key node, and output key path exploration results;

[0110] After completing the analysis of drug action relationships and node marking, all data are integrated and a report on the results of the key path exploration is output. The report lists in detail the pharmacological effects of each key node and link, as well as their role in the treatment strategy. The report includes a graphical network diagram that shows in detail the intensity of action and path direction between drugs. In addition, a detailed description of each drug is provided, such as mechanism of action, indications, side effects, and other information, which can be directly used to guide clinical treatment decisions and the optimization of drug use strategies.

[0111] See also Figure 7 , the evaluation steps of the combination scheme effect are:

[0112] According to the results of critical path exploration, drugs associated with the target pathological process are selected to generate a drug combination model;

[0113] When selecting drugs based on the results of critical path exploration, we first query the pathology database to find drug components that are highly correlated with specific pathological states, and select potentially effective drug combinations based on scientific literature on drug interactions. Then, we analyze the kinetic parameters of each drug at a specific dose, such as absorption rate, distribution volume, metabolic rate, etc., as well as its expected effect on the pathological process. Through relevant parameters, we establish a drug combination plan, predict the efficacy and safety of different combinations, and provide a theoretical basis for subsequent clinical trials and treatments, ultimately forming a drug combination model that combines actual medical needs and scientific research.

[0114] Based on the drug combination model, pathological simulation is performed using the formula, ;

[0115] Calculating the pathological impact of drug combinations , generate simulated drug effect results, where Represents the concentration value of the drug. represents the duration of treatment, represents the half-life of the drug;

[0116] represents the drug concentration, assuming it is 75 mg / L, represents the treatment time, which is 30 days. Represents the half-life of the drug, which is 24 hours. The calculation steps are as follows: First calculate mg·day / L, then divided by the half-life ,get mg·day / L / hour, and finally apply the formula The results showed that the influence of this drug combination on the pathological process was 9.68. This value reflects the pathological influence of the drug combination at a given concentration and treatment duration, and is an important parameter for evaluating the effectiveness of the drug combination.

[0117] Analyze the results of simulated drug effects, conduct statistical analysis, compare drug effects with expected targets, confirm the actual efficacy of the regimen, and obtain the results of the combined regimen effect evaluation;

[0118] Based on the results of simulated drug effects, in-depth analysis is conducted to compare the actual efficacy of the drug combination with the expected target. Specific operations include using analysis of variance (ANOVA) to test the differences in the effects of different drug combinations on pathological states, and performing regression analysis to evaluate the effect of treatment time on drug efficacy. At the same time, multiple comparison correction methods such as Bonferroni correction are applied to ensure the rigor of statistical results. In addition, sensitivity analysis is used to check the robustness of the model and confirm the reliability of the simulation results. This helps the research team to accurately evaluate the effect of each drug combination regimen, so that the final results are closer to clinical application needs and obtain the evaluation results of the combination regimen effect.

[0119] See also Figure 8 , the steps to obtain the dynamic optimization plan for medical trials are:

[0120] Based on the results of the combination effect evaluation, analyze the performance of the drug combination, determine the drug dosage and combination parameters that need to be adjusted through the comparison of pharmacodynamics and clinical data, and generate a list of drug parameter adjustments;

[0121] After evaluating the effects of drug combinations, the clinical performance of each drug combination was analyzed in detail, with special attention paid to those combinations that failed to achieve the expected efficacy. First, drug response information was extracted from a large amount of clinical trial data, and statistical analysis was performed to analyze the variability and statistical significance of the effects of each drug combination (achieved by calculating the standard deviation and variance). Then, based on the data, the need for dose adjustment was identified, and it was analyzed which drugs needed to have their doses increased to improve efficacy and which drugs might cause side effects due to excessive doses, providing a scientific basis for subsequent optimization of drug doses and treatment plans, and generating a list of drug parameter adjustments.

[0122] Adjust the drug parameter list, adjust the drug dosage, reconfigure the combination plan, optimize the drug combination configuration by simulating the efficacy prediction under different dosages and combinations, and generate the optimized drug combination plan;

[0123] When adjusting drug dosage, use the formula, Make drug dosage adjustment values: In this formula, and Represent the maximum and minimum therapeutic windows of drug concentration, respectively, obtained through pharmacokinetic studies, and set mg / L, mg / L; Represents the threshold value of therapeutic efficacy, which is determined by clinical trial data and is set to ; is an adjustment factor that is dynamically adjusted according to the severity of the pathological state and depends on the patient's response. ; The calculation process is as follows:

[0124] ;

[0125] The results showed that in order to achieve the preset efficacy threshold, the drug dose should be adjusted to 18 mg / L per day.

[0126] Based on the optimized drug combination scheme, conduct new clinical trial design, including randomization, blind setting and control group design, and generate dynamic optimization scheme for medical trials;

[0127] Based on the optimized drug dosage and combination strategy, a new clinical trial design was refined and formulated. First, the basic structure of the clinical trial was defined, including the selection of an appropriate patient population, the design of the randomization process, and the implementation of the blind test to ensure the fairness and scientific nature of the trial. Next, a control group was set up to compare the effect of the optimized drug combination with conventional treatment. In addition, the data collection method and time points were planned in detail to ensure that the efficacy and safety of the drug combination could be accurately evaluated. Through these detailed plans, it was ensured that the trial could dynamically adjust the drug dosage and combination strategy on the basis of scientific rigor, and ultimately a dynamic optimization medical trial program was formed that was both targeted and able to be adjusted according to clinical responses.

[0128] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A biomedical experiment optimization system based on a graphical model, characterized in that: The system comprises: The data collection module configures the data interface, collects drug concentration and biomarker data, integrates the collected information and performs real-time data synchronization, covers the entire test process, and outputs a biological dynamic data set; The graph weight adjustment module adjusts the nodes and edges in the graph by weight update according to the biodynamic data set, recalculates the connection strength between nodes each time the data is updated, reflects the interaction effect between drugs in real time according to the calculation results, maps the real-time interaction between drugs, and generates an adjusted interaction weight graph; The pathway exploration module uses the adjusted interaction weight graph to perform pathway analysis, identify key action pathways between drugs, analyze the pharmacological action relationships of pathway nodes based on the identification results, mark key nodes and links, and output key pathway exploration results; The optimization and adjustment module evaluates the effect of the combination scheme by simulating the effect of the drug combination on the pathological process according to the critical path exploration results, makes targeted adjustments according to the evaluation results, optimizes the drug dosage and combination scheme, and forms a dynamic optimization scheme for medical trials; The steps for calculating the connection strength and direction between the nodes are: Based on the biological dynamic data set, extract drug interaction data that affects node connections, perform preliminary screening on nodes and edges, and generate preliminary screening results; Based on the preliminary screening results, the weights are updated using the formula: ; Calculate weights for new nodes and edges , generate updated weight results, where, is the old weight, is the learning rate, is the frequency of new interactions between nodes, and Is a node and The degree of Using the updated weight result, the formula is adopted: ; Recalculate the connection strength between nodes , output the recalculation result of the connection strength between nodes, where, Is a node and The new weights between is the maximum value among all new weights; The steps for obtaining the adjusted interaction weight graph are: Integrate the recalculation results of the connection strength between the nodes, convert the data to map the interaction effects between the drugs, and generate visual data preparation results by assigning visual weights to each node to reflect the strength of the effect in the network; Applying the visualization data preparation results, dynamically updating the drug interaction network diagram, revealing real-time interaction data by adjusting the size of nodes and the thickness of edges, and generating an updated network diagram structure; According to the updated network graph structure, visualization processing is performed to reflect the interaction changes caused by data update in real time, and an adjusted interaction weight graph is generated.

2. The biomedical experiment optimization system based on a graphical model according to claim 1, characterized in that: The steps of acquiring the biological dynamic data set are: Configure data interfaces and sensors to automatically collect drug concentration and biomarker data and obtain raw data sets; According to the original data set, key parameters are extracted, data filtering is performed and the data is preliminarily calibrated using the formula: ; Calculate data processing efficiency index , and obtain comprehensive analysis data, among which, represents the drug concentration value after calibration, represents the biomarker response intensity value, Represents the time window size; The comprehensive analysis data are used to perform pattern recognition and statistical analysis, identify and verify the response pattern of biomarkers, capture the entire test response, and generate a biological dynamic data set.

3. The biomedical experiment optimization system based on a graphical model according to claim 1, characterized in that: The steps for identifying the key action pathways between the drugs are: calling the adjusted interaction weight graph, analyzing the mutual relationship between drugs based on the pharmacological action intensity, and outputting basic analysis data; Based on the adjusted interaction weight graph and the basic analysis data, the formula is adopted: ; Calculate new path weights , generate the optimal path analysis results, where represents the original path weight, Represents a connected node and the relative importance of Analyze the optimal path analysis results, identify key action pathways, including key nodes and connections in the path, and generate key action pathway confirmation results.

4. The biomedical experiment optimization system based on a graphical model according to claim 3, characterized in that: The steps for obtaining the critical path exploration results are as follows: Perform pharmacological action analysis on each node in the key action pathway confirmation results, evaluate the interaction between drugs through pharmacokinetic and drug target data, and output the pharmacological action analysis results; Based on the pharmacological action analysis results, key nodes are determined, drug action relationships are combined with key nodes and links, and the positions of drugs in the graph are marked to obtain drug marking information; The drug labeling information is integrated, the drug interaction network is graphically displayed, the pharmacological effect of each key node is recorded, and the key path exploration results are output.

5. The biomedical experiment optimization system based on a graphical model according to claim 4, characterized in that: The steps for evaluating the effect of the combination scheme are: According to the critical path exploration results, drugs associated with the target pathological process are selected to generate a drug combination model; Based on the drug combination model, pathological simulation was performed using the formula: ; Calculating the pathological impact of drug combinations , generate simulated drug effect results, where Represents the concentration value of the drug. represents the duration of treatment, represents the half-life of the drug; Analyze the simulated drug effect results, perform statistical analysis, compare the drug effects with the expected goals, confirm the actual efficacy of the scheme, and obtain the combination scheme effect evaluation results.

6. The biomedical experiment optimization system based on a graphical model according to claim 5, characterized in that: The steps for obtaining the dynamic optimization plan of the medical trial are: Based on the combination effect evaluation results, analyze the drug combination performance, determine the drug dosage and combination parameters that need to be adjusted through pharmacodynamic and clinical data comparison, and generate a drug parameter adjustment list; According to the drug parameter adjustment list, the drug dosage is adjusted, the combination scheme is reconfigured, and the drug combination configuration is optimized by simulating the efficacy prediction under different dosages and combinations to generate an optimized drug combination scheme; Based on the optimized drug combination scheme, a new clinical trial design is carried out, including randomization, blinding setting and control group design, to generate a dynamic optimization plan for medical trials.

Images