System for automatically generating clinical test item statistical analysis plan and report

Through the system of automatically generating statistical analysis plans and reports of clinical trial projects, the problem of inconsistent manual writing and data management in clinical trials is solved, automated generation and in-depth analysis are realized, and work efficiency and data quality are improved.

CN119069057AInactive Publication Date: 2024-12-03BEIJING KANGTERIKE STATISTICAL TECH CO LTD

Patent Information

Application Number
CN202411185561.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-12-03
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The preparation and review of statistical analysis plans and reports in clinical trials relies on manual labor, which leads to a lot of time consumption and errors, and the existing data management is not unified, making data integrity, traceability and security difficult to guarantee, analysis tools and algorithms It is relatively basic and cannot deeply explore potential information in the data.

Method used

Design a system that automatically generates statistical analysis plans and reports of clinical trial projects, including data management module, template management module, automatic generation module, screening and verification module, analysis and feedback module, user interface module and system integration and interface module, through Automate the generation of statistical analysis plans and reports, providing in-depth data analysis and optimization suggestions to ensure data integrity and security.

Benefits of technology

Significantly reduce manual writing and proofreading time, improve work efficiency, reduce error rates, ensure the accuracy and consistency of statistical analysis plans and reports, deeply analyze clinical trial data, generate detailed statistical indicators and charts, and improve the scientificity of trial results and Reliability, guarantee the integrity and security of data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119069057A_ABST
    Figure CN119069057A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of clinical tests, and discloses a system for automatically generating a clinical test item statistical analysis plan and report, which comprises a data management module, a template management module, an automatic generation module, a screening and checking module, an analysis and feedback module, a user interface module and a system integration and interface module. The data management module is responsible for receiving, storing and managing various data of clinical tests, allowing a user to perform addition, deletion, modification and check operations on the data, and providing data backup and recovery functions; the template management module supports creation and management of templates for statistical analysis plans and reports. By automatically generating the statistical analysis plan and report, the time for manual compiling and proofreading is remarkably shortened, the working efficiency is improved, the system automatic generation process reduces human intervention, errors caused by human negligence or understanding deviation are reduced, and the accuracy and consistency of the statistical analysis plan and report are ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of clinical trials, and in particular to a system for automatically generating statistical analysis plans and reports for clinical trial projects. Background Art

[0002] Clinical trials are systematic trials that aim to detect or verify the clinical medicine, pharmacology and other pharmacodynamic effects, adverse reactions, or absorption, distribution, metabolism and excretion of an investigational drug in order to determine the efficacy and safety of the drug.

[0003] With the development of the pharmaceutical R&D field, clinical trials have become an indispensable part of the new drug R&D process. Clinical trials usually involve complex data collection, processing and analysis, which directly affects the scientificity and reliability of the test results. At present, in clinical trial projects, the preparation and review of statistical analysis plans and reports mainly rely on manual completion, and the manual operation mode has many shortcomings:

[0004] The preparation of statistical analysis plans and reports requires manual input and data proofreading, which is a time-consuming and laborious process. When faced with a large amount of data, the preparation and proofreading process takes a lot of time, which undoubtedly prolongs the test cycle and increases the test cost. In the process of manual preparation and proofreading, errors are prone to occur due to operator negligence or misunderstanding, resulting in data differences in statistical analysis plans and reports and deviations in analysis results.

[0005] In the existing technology, the functions of receiving, storing, editing, backing up and restoring clinical trial data are scattered in different systems or platforms, and there is a lack of unified data management, which makes it difficult to ensure the integrity, traceability and security of the data. In addition, the existing clinical trial data analysis tools and algorithms are relatively basic and cannot deeply explore the potential information in the data, and cannot generate detailed statistical indicators and charts, which affects the scientific nature of the trial results.

[0006] Therefore, those skilled in the art provide a system for automatically generating statistical analysis plans and reports for clinical trial projects to solve the problems raised in the above-mentioned background technology. Summary of the invention

[0007] In view of the deficiencies of the prior art, the present invention provides a system for automatically generating statistical analysis plans and reports for clinical trial projects to solve the problems raised in the above-mentioned background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: a system for automatically generating statistical analysis plans and reports for clinical trial projects, including a data management module, a template management module, an automatic generation module, a screening and verification module, an analysis and feedback module, a user interface module, and a system integration and interface module;

[0009] The data management module is responsible for receiving, storing and managing various types of data of clinical trials, allowing users to add, delete, modify and query data, and providing data backup and recovery functions; the template management module supports the creation and management of templates for statistical analysis plans and reports; the automatic generation module automatically generates a detailed statistical analysis plan based on the input data and template to provide comprehensive data analysis and result summary; the screening and verification module provides tools and interfaces to help users manually check and verify the accuracy of data and correct errors; the analysis and feedback module conducts in-depth analysis of clinical trial data, generates various statistical indicators and charts, and provides optimization suggestions based on the analysis results; the user interface module provides an intuitive and friendly user interface to facilitate users to input data, manage templates, generate analysis plans and view reports; the system integration and interface module supports integration with other systems and provides standardized data input and output interfaces.

[0010] Preferably, the data management module includes a data receiving unit, a data storage unit, a data editing unit, a data backup unit, and a data recovery unit; the data receiving unit is responsible for receiving data from different data sources; the data storage unit stores the received data in a database; the data editing unit provides a user interface and tools to allow the user to add, delete, modify and query data; the data backup unit backs up data regularly; the data recovery unit can restore data from the backup when the data is lost or damaged;

[0011] The template management module includes a template creation unit, a template editing unit, and a template management unit; the template creation unit allows users to create new statistical analysis plans and report templates, and can define the format and content of the template according to requirements; the template editing unit supports editing and updating of existing templates and adjusting template content; the template management unit provides template management functions.

[0012] Preferably, the automatic generation module includes a plan generation unit and a report generation unit; the plan generation unit automatically generates a statistical analysis plan based on the input data and template; the report generation unit automatically generates an analysis report based on the statistical analysis results and template, providing a comprehensive data summary and result display;

[0013] The screening and verification module includes a manual verification unit and a machine screening unit; the manual verification unit provides a user interface and tools to help users manually verify the accuracy of data and correct errors; the machine screening unit uses automated algorithms and rules to screen and analyze data and automatically detect potential problems and anomalies.

[0014] Preferably, the analysis and feedback module includes a data analysis unit and a feedback generation unit; the data analysis unit performs in-depth analysis on the clinical trial data to generate various statistical indicators, charts and visualization results; the feedback generation unit provides optimization suggestions and improvement plans based on the data analysis results to help optimize the trial design and process;

[0015] The system integration and interface module includes an external system integration unit and an interface management unit; the external system integration unit supports integration with other systems; and the interface management unit provides standardized data input and output interfaces.

[0016] Preferably, the data receiving unit effectively identifies and deletes duplicate records through a hash function, and its algorithm formula is:

[0017] Hash(D i )=h(D i )whereD i ∈Data Set,

[0018] Among them, D i is the i-th data received, h(D i ) is the data D i The generated hash value, if h(D i )∈H, it indicates that the data is duplicate and should be deleted;

[0019] The data storage unit uses Huffman coding to compress data, and its algorithm formula is:

[0020]

[0021] Among them, C(w) is the length of the encoded data, P(w i ) is the symbol w i The probability of Length(w i ) is the symbol w i The encoding length of

[0022] When the data editing unit performs an editing operation on the data, a transaction management algorithm is used, and the algorithm formula is:

[0023] Atomicity:T i ={t1,t2,…,t n},

[0024] Consistency:C(s)={s1,s2,…,s m},

[0025] Isolation:

[0026] Durability:D(T i )(Changes persist),

[0027] Among them, T i Indicates transaction, s i Indicates the database status;

[0028] The data backup unit uses an incremental backup strategy to optimize the backup speed and storage space, and its algorithm formula is:

[0029] ΔB n =B n -B n-1 ,

[0030] Among them, B n is the data backed up for the nth time, ΔB n It is the difference data between the previous backup.

[0031] The data recovery unit uses reverse operation combined with log files to recover data, and its algorithm formula is:

[0032]

[0033] Among them, R n Indicates recovery to the state of the nth backup, ΔB i It is the changed data after the i-th backup.

[0034] Preferably, the template creation unit utilizes natural language generation technology in machine learning to automatically generate template content based on context and semantics, and its algorithm formula is:

[0035]

[0036] Among them, P(T|C) is the probability of generating template T under context C, t i is the i-th element in the template, n is the total number of elements in the template;

[0037] The template editing unit manages different versions of templates and can use a version control algorithm based on a directed acyclic graph, and the algorithm formula is:

[0038] V n =V n-1 +ΔT n ,

[0039] Among them, V n is the nth version of the template, ΔT n is the change of the nth version compared to the previous version;

[0040] When managing templates, the template management unit may use a collaborative filtering algorithm to recommend templates to users, and the algorithm formula is:

[0041]

[0042] Among them, R(u,t) is the preference prediction of user u for template t, μ is the global average, and b u and b t is the deviation between user u and template t, q u and p t are the feature vectors of users and templates.

[0043] Preferably, during the manual verification process of the manual verification unit, the system may use a human-computer interaction optimization algorithm to dynamically adjust the user interface and the verification sequence, and the algorithm formula is:

[0044]

[0045] Among them, Q(s,a) is the value function of taking action a in state s, α is the learning rate, r is the reward, and γ is the discount factor. is the optimal future value in the next state s′;

[0046] The machine screening unit can use an automated anomaly detection algorithm to identify anomalies in the data, and its algorithm formula is:

[0047]

[0048] where h(x) is the anomaly score of data point x, E(h(x)) is the expected path length for the data point to be isolated, and c(n) is a constant that depends on the number of samples n.

[0049] Preferably, the data analysis unit may use multivariate regression analysis to analyze the relationship between multiple variables, and its algorithm formula is:

[0050] Y=β0+β1X1+β2X2+…+β n X n +∈,

[0051] Among them, Y is the dependent variable, X1, X2, X n is the independent variable, β0 is the intercept term, β1, β2, β n is the regression coefficient of the independent variable, ∈ is the error term;

[0052] The feedback generation unit may use a reinforcement learning algorithm to optimize the generated feedback, and the algorithm formula is:

[0053]

[0054] Among them, π *(s) is the optimal strategy in state s, Q * (s,a) is the optimal value function for taking action a in state s.

[0055] Preferably, the external system integration unit may use a service orchestration algorithm during the external system integration process to ensure that the functions of multiple heterogeneous systems can be smoothly integrated and collaborated. The algorithm formula is:

[0056]

[0057] Among them, O is the best service orchestration solution, w ij is the weight between service i and service j, indicating their dependency or priority, x ij is a binary variable of whether service i and service j are included in the orchestration scheme;

[0058] In order to prevent the API interface from being abused, the interface management unit can use a token bucket algorithm to implement the rate limit of the interface. The algorithm formula is:

[0059] B(t)=min(C,B(t-1)+R·(tt last )),

[0060] Where B(t) is the number of tokens in the bucket at time t, C is the capacity of the bucket, R is the token generation rate, and t last Is the time of the last request.

[0061] The present invention provides a system for automatically generating statistical analysis plans and reports for clinical trial projects. It has the following beneficial effects:

[0062] 1. The present invention significantly reduces the time of manual writing and proofreading and improves work efficiency by automatically generating statistical analysis plans and reports. The system's automated generation process reduces human intervention and reduces errors caused by human negligence or misunderstanding, thereby ensuring the accuracy and consistency of statistical analysis plans and reports.

[0063] 2. The present invention provides comprehensive data receiving, storage, editing, backup and recovery functions through the data management module in the system to ensure the integrity, traceability and security of clinical trial data. Users can create and edit statistical analysis templates according to specific needs to meet the special needs of different types of clinical trials and improve the adaptability and flexibility of the system.

[0064] 3. The present invention integrates advanced data analysis algorithms and tools, and can then deeply analyze clinical trial data, generate detailed statistical indicators and charts, and provide scientific and reliable analysis results for clinical trials. It also makes optimization suggestions through the feedback module to improve the scientific nature of trial design and results. The system integration and interface module supports seamless integration with other external systems, provides standardized data input and output interfaces, ensures the scalability of the system and compatibility with existing systems, improves the accuracy and efficiency of data verification, and ensures the quality of clinical trial data. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 is a system diagram of the present invention;

[0066] Figure 2 It is a schematic diagram of a data management module of the present invention;

[0067] Figure 3 This is a schematic diagram of a template management module of the present invention;

[0068] Figure 4 It is a schematic diagram of the automatic generation module of the present invention;

[0069] Figure 5 It is a schematic diagram of the screening and verification module of the present invention;

[0070] Figure 6 It is a schematic diagram of the analysis and feedback module of the present invention;

[0071] Figure 7 It is a schematic diagram of the system integration and interface module of the present invention. DETAILED DESCRIPTION

[0072] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0073] The present invention is described in detail below in conjunction with the accompanying drawings:

[0074] Example:

[0075] Please refer to the attached Figure 1 - Attachment Figure 7 , an embodiment of the present invention provides a system for automatically generating statistical analysis plans and reports for clinical trial projects, including a data management module, a template management module, an automatic generation module, a screening and verification module, an analysis and feedback module, a user interface module, and a system integration and interface module;

[0076] The data management module is responsible for receiving, storing and managing various types of data of clinical trials, allowing users to add, delete, modify and query data, and providing data backup and recovery functions; the template management module supports the creation and management of templates for statistical analysis plans and reports; the automatic generation module automatically generates detailed statistical analysis plans based on the input data and templates to provide comprehensive data analysis and results summary; the screening and verification module provides tools and interfaces to help users manually check and verify the accuracy of data and correct errors; the analysis and feedback module conducts in-depth analysis of clinical trial data and generates various statistical indicators and charts to provide optimization suggestions based on the analysis results; the user interface module provides an intuitive and friendly user interface to facilitate users to input data, manage templates, generate analysis plans and view reports; the system integration and interface module supports integration with other systems and provides standardized data input and output interfaces.

[0077] The benefits of the data management module are that it provides complete data receiving, storage, editing, backup and recovery functions, allowing users to conveniently manage data in clinical trials, ensuring efficient data processing, and safeguarding the integrity, traceability and security of clinical trial data, reducing the risk of data loss or damage; the benefits of the template management module are that it reduces duplication of work when creating new analysis plans and reports, improves overall work efficiency, and ensures that templates always meet the latest requirements and standards; the benefits of the automatic generation module are that it reduces manual intervention, significantly improves generation efficiency, and reduces the risk of errors in manual writing and proofreading during the automation process, ensuring that the generated plans and reports are highly accurate and consistent, helping decision makers better understand Understand and utilize clinical trial data; The benefits of the screening and verification module help users identify and correct potential errors, ensure data accuracy, quickly screen out abnormal data, improve the efficiency of data verification, and ensure the quality of clinical trial data; The benefits of the analysis and feedback module provide in-depth data insights, support scientific decision-making, help improve trial design and processes, and enhance the overall effectiveness and scientificity of clinical trials; The benefits of the user interface module improve the ease of use of the system, simplify complex operations, reduce users' learning costs and operation time, and improve overall usage efficiency; The benefits of the system integration and interface module simplify the data exchange process between different systems, improve the efficiency and accuracy of data transmission, and support the addition of new systems or new functions.

[0078] The data management module includes a data receiving unit, a data storage unit, a data editing unit, a data backup unit, and a data recovery unit; the data receiving unit is responsible for receiving data from different data sources; the data storage unit stores the received data in a database; the data editing unit provides a user interface and tools to allow users to add, delete, modify, and query data; the data backup unit backs up data regularly; the data recovery unit can restore data from the backup when data is lost or damaged;

[0079] The template management module includes template creation unit, template editing unit and template management unit; the template creation unit allows users to create new statistical analysis plans and report templates, and can define the format and content of the template according to requirements; the template editing unit supports editing and updating of existing templates and adjusting template content; the template management unit provides template management functions.

[0080] The data receiving unit improves the comprehensiveness and accuracy of data and helps to quickly respond to data needs in clinical trials; the data storage unit helps to ensure the availability of data during clinical trials and improve data access speed and management efficiency; the data editing unit can adjust data according to needs to avoid problems caused by data lags or errors; the data backup unit ensures the security and integrity of data and reduces the risk of business interruption caused by data loss or damage; the data recovery unit reduces the impact of data loss on clinical trials and provides a high degree of flexibility;

[0081] The template creation unit ensures that the template can accurately match the requirements of different clinical trials, reduce duplication of work, and improve overall work efficiency; the template editing unit ensures that the template is always kept up to date and most compliant, improving the reuse value of the template; the template management unit ensures the orderliness and controllability of template resources, avoids errors caused by version confusion, and ensures that the template version used is always correct and valid.

[0082] The automatic generation module includes a plan generation unit and a report generation unit; the plan generation unit automatically generates a statistical analysis plan based on the input data and template; the report generation unit automatically generates an analysis report based on the statistical analysis results and template, providing a comprehensive data summary and result display;

[0083] The screening and verification module includes a manual verification unit and a machine screening unit; the manual verification unit provides a user interface and tools to help users manually verify the accuracy of the data and correct errors; the machine screening unit uses automated algorithms and rules to screen and analyze data and automatically detect potential problems and anomalies.

[0084] The plan generation unit significantly reduces the time and effort of manual writing, improves work efficiency, avoids differences or errors caused by manual writing, ensures the reliability of the statistical analysis process, and helps to make timely decisions; the report generation unit reduces human intervention, reduces the risk of errors in the report generation process, improves the quality of the report, and helps all parties to accurately grasp the test results;

[0085] The manual verification unit helps users manually verify the accuracy of data, making the verification process more efficient and convenient, and preventing erroneous data from affecting the analysis results; the machine screening unit reduces omissions and errors in the manual verification process and helps to correct them in a timely manner.

[0086] The analysis and feedback module includes a data analysis unit and a feedback generation unit. The data analysis unit conducts in-depth analysis of clinical trial data and generates various statistical indicators, charts, and visualization results. The feedback generation unit provides optimization suggestions and improvement plans based on the data analysis results to help optimize the trial design and process.

[0087] The system integration and interface module includes an external system integration unit and an interface management unit; the external system integration unit supports integration with other systems; the interface management unit provides standardized data input and output interfaces.

[0088] The data analysis unit helps decision makers better understand the information behind the data, provides analysis views from different angles, and meets the diverse data analysis needs in clinical trials; the feedback generation unit improves the scientificity and effectiveness of the trial design, and improves the overall quality and efficiency of clinical trials;

[0089] The external system integration unit can improve the efficiency of cross-system collaborative work, reduce repeated data input and processing, and reduce operational complexity; the interface management unit reduces format conversion problems in data transmission, improves data processing efficiency, simplifies the data integration process between the system and external systems, reduces technical complexity, and reduces errors in the integration process.

[0090] The data receiving unit uses a hash function to effectively identify and delete duplicate records. The algorithm formula is:

[0091] Hash(D i )=h(D i )where D i ∈Data Set,

[0092] Among them, D i is the i-th data received, h(D i ) is the data D i The generated hash value, if h(D i )∈H, it indicates that the data is duplicate and should be deleted;

[0093] The data storage unit uses Huffman coding for data compression, and its algorithm formula is:

[0094]

[0095] Among them, C(w) is the length of the encoded data, P(w i ) is the symbol w i The probability of Length(w i ) is the symbol w i The encoding length of

[0096] When the data editing unit performs editing operations on data, it uses a transaction management algorithm, and its algorithm formula is:

[0097] Atomicity:T i ={t1,t2,…,t n},

[0098] Consistency:C(s)={s1,s2,…,s m},

[0099] Isolation:

[0100] Durability:D(T i )(Changes persist),

[0101] Among them, T i Indicates transaction, s i Indicates the database status;

[0102] The data backup unit uses an incremental backup strategy to optimize backup speed and storage space. The algorithm formula is:

[0103] ΔB n =B n -B n-1 ,

[0104] Among them, B n is the data backed up for the nth time, ΔB n It is the difference data between the last backup;

[0105] The data recovery unit uses reverse operation combined with log files to recover data. The algorithm formula is:

[0106]

[0107] Among them, R n Indicates recovery to the state of the nth backup, ΔB i It is the changed data after the i-th backup.

[0108] The hash function deduplication algorithm quickly identifies and deletes duplicate records by calculating the hash value of the data, thereby ensuring the uniqueness and integrity of the data. The high efficiency of the hash function enables fast data processing, effectively reducing the waste of storage space and redundancy in data processing;

[0109] The Huffman coding algorithm uses the probability of symbols to encode data, effectively reducing the storage length of the overall data. This algorithm significantly saves space when storing data and improves storage efficiency.

[0110] The transaction management algorithm ensures the reliability and consistency of database operations. During the data editing process, transaction management guarantees the atomicity, consistency, isolation and persistence of operations, that is, it ensures that all data operations are completed or completely rolled back, thereby ensuring the integrity and security of the data.

[0111] Compared with full backup, incremental backup significantly reduces the time and storage resources of each backup, improves the efficiency and availability of system backup, and reduces the impact of backup on system performance.

[0112] The template creation unit uses the natural language generation technology in machine learning to automatically generate template content based on context and semantics. Its algorithm formula is:

[0113]

[0114] Among them, P(T|C) is the probability of generating template T under context C, t i is the i-th element in the template, n is the total number of elements in the template;

[0115] The template editing unit manages different versions of templates and can use a version control algorithm based on a directed acyclic graph. The algorithm formula is:

[0116] V n =V n-1 +ΔT n ,

[0117] Among them, V n is the nth version of the template, ΔT n is the change of the nth version compared to the previous version;

[0118] When managing templates, the template management unit can use collaborative filtering algorithms to recommend templates to users. The algorithm formula is:

[0119]

[0120] Among them, R(u,t) is the preference prediction of user u for template t, μ is the global average, and b u and b t is the deviation between user u and template t, q u and p t are the feature vectors of users and templates.

[0121] The natural language generation algorithm uses natural language generation technology to automatically generate template content based on a given context C. By gradually generating each element of the template and taking into account the previously generated content and context, it ensures that the generated template content is semantically coherent and logically clear. It can automatically generate templates that meet specific requirements, reduce the workload of manual writing, and improve the accuracy and consistency of template content.

[0122] The directed acyclic graph version control algorithm is used to manage different versions of templates, ensuring that changes in each version can be tracked and recorded when modifying or updating the template, preventing version confusion and ensuring that users always use the correct version during template editing, reducing the risk of errors;

[0123] The collaborative filtering recommendation algorithm is used for template recommendation in the template management unit. By analyzing the user's historical behavior and the characteristics of the template, the collaborative filtering algorithm can predict the user's preference for a specific template. This algorithm combines the global average, user preferences and template characteristics to achieve personalized recommendations and improve the efficiency and satisfaction of template use.

[0124] During the manual verification process, the system can use the human-computer interaction optimization algorithm to dynamically adjust the user interface and verification sequence. The algorithm formula is:

[0125]

[0126] Among them, Q(s,a) is the value function of taking action a in state s, α is the learning rate, r is the reward, and γ is the discount factor. is the optimal future value in the next state s′;

[0127] The machine screening unit can use an automated anomaly detection algorithm to identify anomalies in the data. The algorithm formula is:

[0128]

[0129] where h(x) is the anomaly score of data point x, E(h(x)) is the expected path length for the data point to be isolated, and c(n) is a constant that depends on the number of samples n.

[0130] The human-computer interaction optimization algorithm is used to optimize the decision-making and behavior in the human-computer interaction process. Through the reinforcement learning algorithm, the system can dynamically adjust the user interface and verification order so that users can complete tasks more efficiently during the verification process. After each user performs an operation in a specific state, the system will update the value function Q(s,a) of the operation according to the operation result, and gradually learn the optimal interaction strategy. Ultimately, the optimization can improve the efficiency of manual verification and reduce the user's operation burden;

[0131] The automated anomaly detection algorithm is used to automatically detect anomalies in the data, indicating that the data point is an anomaly. The algorithm can quickly identify anomalies in the data, help to promptly discover and deal with data anomalies that may affect the analysis results, and ensure data quality and accuracy of analysis.

[0132] The data analysis unit can use multivariate regression analysis to analyze the relationship between multiple variables. The algorithm formula is:

[0133] Y=β0+β1X1+β2X2+…+β n X n +∈,

[0134] Among them, Y is the dependent variable, X1, X2, X n is the independent variable, β0 is the intercept term, β1, β2, β n is the regression coefficient of the independent variable, ∈ is the error term;

[0135] The feedback generation unit can use a reinforcement learning algorithm to optimize the generated feedback. The algorithm formula is:

[0136]

[0137] Among them, π * (s) is the optimal strategy in state s, Q * (s,a) is the optimal value function for taking action a in state s.

[0138] Multivariate regression analysis algorithm analysis is widely used to predict, explain and establish relationships between variables. In clinical trial data analysis, the method helps to understand how multiple factors work together to affect the trial results;

[0139] Reinforcement learning algorithms can continuously adjust and optimize feedback strategies to maximize expected benefits or effects, which means that the system can gradually improve the generated feedback based on user reactions and environmental changes, making it more effective and accurate, ultimately improving user experience and system performance.

[0140] During the external system integration process, the external system integration unit can use the service orchestration algorithm to ensure that the functions of multiple heterogeneous systems can be smoothly integrated and collaborated. The algorithm formula is:

[0141]

[0142] Among them, O is the best service orchestration solution, w ij is the weight between service i and service j, indicating their dependency or priority, x ij is a binary variable of whether service i and service j are included in the orchestration scheme;

[0143] To prevent API interfaces from being abused, the interface management unit can use a token bucket algorithm to implement interface rate limiting. The algorithm formula is:

[0144] B(t)=min(C,B(t-1)+R·(tt last )),

[0145] Where B(t) is the number of tokens in the bucket at time t, C is the capacity of the bucket, R is the token generation rate, and tlast Is the time of the last request.

[0146] The service orchestration algorithm is used to optimize service orchestration and ensure that different heterogeneous systems can collaborate efficiently during the external system integration process. Service orchestration involves the dependencies and execution order between multiple services. By modeling and weighing the relationships, the service orchestration algorithm can find the optimal service execution solution.

[0147] The token bucket algorithm is used to implement rate limiting of the interface and prevent the API interface from being abused. The token bucket algorithm is a flow control mechanism. The number of tokens B(t) in the bucket represents the number of requests that the system can handle. The capacity C of the bucket limits the number of tokens to prevent unlimited accumulation of tokens. By controlling the generation and consumption rate of tokens, the token bucket algorithm effectively limits the frequency of API calls, prevents the system from crashing or performance degradation due to overloaded requests, and ensures the stability and fairness of the service.

[0148] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A system for automatically generating statistical analysis plans and reports for clinical trial projects, characterized in that: It includes data management module, template management module, automatic generation module, screening and verification module, analysis and feedback module, user interface module, system integration and interface module; The data management module is responsible for receiving, storing and managing various types of clinical trial data, allowing users to add, delete, modify and query data, and providing data backup and recovery functions; the template management module supports the creation and management of templates for statistical analysis plans and reports; The automatic generation module automatically generates a detailed statistical analysis plan based on the input data and template to provide a comprehensive data analysis and result summary; The screening and verification module provides tools and interfaces to help users manually check and verify the accuracy of data and correct errors; the analysis and feedback module conducts in-depth analysis of clinical trial data and generates various statistical indicators and charts to provide optimization suggestions based on the analysis results; The user interface module provides an intuitive and user-friendly user interface, which facilitates users to input data, manage templates, generate analysis plans and view reports; The system integration and interface module supports integration with other systems and provides standardized data input and output interfaces.

2. A system for automatically generating clinical trial project statistical analysis plans and reports according to claim 1, characterized in that: The data management module includes a data receiving unit, a data storage unit, a data editing unit, a data backup unit, and a data recovery unit; the data receiving unit is responsible for receiving data from different data sources; The data storage unit stores the received data in a database; The data editing unit provides a user interface and tools to allow users to add, delete, modify and query data; the data backup unit backs up data regularly; The data recovery unit can restore data from the backup when data is lost or damaged; The template management module includes a template creation unit, a template editing unit, and a template management unit; The template creation unit allows users to create new statistical analysis plans and report templates, and can define the format and content of the templates according to requirements; The template editing unit supports editing and updating of existing templates and adjusting template content; The template management unit provides a template management function.

3. A system for automatically generating clinical trial project statistical analysis plans and reports according to claim 1, characterized in that: The automatic generation module includes a plan generation unit and a report generation unit; The plan generation unit automatically generates a statistical analysis plan based on the input data and template; the report generation unit automatically generates an analysis report based on the statistical analysis results and template, providing a comprehensive data summary and result display; The screening and verification module includes a manual verification unit and a machine screening unit; the manual verification unit provides a user interface and tools to help users manually verify the accuracy of data and correct errors; The machine screening unit uses automated algorithms and rules to screen and analyze data to automatically detect potential problems and anomalies.

4. A system for automatically generating clinical trial project statistical analysis plans and reports according to claim 1, characterized in that: The analysis and feedback module includes a data analysis unit and a feedback generation unit; the data analysis unit performs in-depth analysis on clinical trial data to generate various statistical indicators, charts and visualization results; the feedback generation unit provides optimization suggestions and improvement plans based on the data analysis results to help optimize the trial design and process; The system integration and interface module includes an external system integration unit and an interface management unit; the external system integration unit supports integration with other systems; and the interface management unit provides standardized data input and output interfaces.

5. A system for automatically generating statistical analysis plans and reports for clinical trial projects according to claim 2, characterized in that: The data receiving unit effectively identifies and deletes duplicate records through a hash function, and its algorithm formula is: Hash(D i )=h(D i ) where D i ∈Data Set, Among them, D i is the i-th data received, h(D i ) is the data D i The generated hash value, if h(D i )∈H, it indicates that the data is duplicate and should be deleted; The data storage unit uses Huffman coding to compress data, and its algorithm formula is: Among them, C(w) is the length of the encoded data, P(w i ) is the symbol w i The probability of Length(w i ) is the symbol w i The encoding length of When the data editing unit performs an editing operation on the data, a transaction management algorithm is used, and the algorithm formula is: Atomicity:T i ={t1,t2,...,t n }, Consistency:C(S)={s1,s2,...,s m }, Isolation: Durability:D(T i )(Changes persist), Among them, T i Indicates transaction, s i Indicates the database status; The data backup unit uses an incremental backup strategy to optimize the backup speed and storage space, and its algorithm formula is: ΔB n =B n -B n-1 , Among them, B n is the data backed up for the nth time, ΔB n It is the difference data between the last backup; The data recovery unit uses reverse operation combined with log files to recover data, and its algorithm formula is: Among them, R n Indicates recovery to the state of the nth backup, ΔB i It is the changed data after the i-th backup.

6. A system for automatically generating clinical trial project statistical analysis plans and reports according to claim 2, characterized in that: The template creation unit utilizes the natural language generation technology in machine learning to automatically generate template content based on context and semantics. The algorithm formula is: Among them, P(T|C) is the probability of generating template T under context C, t i is the i-th element in the template, n is the total number of elements in the template; The template editing unit manages different versions of templates and can use a version control algorithm based on a directed acyclic graph, and the algorithm formula is: V n =V n-1 +ΔT n , Among them, V n is the nth version of the template, ΔT n is the change of the nth version compared to the previous version; When managing templates, the template management unit may use a collaborative filtering algorithm to recommend templates to users, and the algorithm formula is: Among them, R(u,t) is the preference prediction of user u for template t, μ is the global average, and b u and b t is the deviation between user u and template t, q u and p t are the feature vectors of users and templates.

7. A system for automatically generating clinical trial project statistical analysis plans and reports according to claim 3, characterized in that: During the manual verification process, the system can use a human-computer interaction optimization algorithm to dynamically adjust the user interface and the verification sequence. The algorithm formula is: Among them, Q(s,a) is the value function of taking action a in state s, α is the learning rate, r is the reward, and γ is the discount factor. is the optimal future value in the next state s′; The machine screening unit can use an automated anomaly detection algorithm to identify anomalies in the data, and its algorithm formula is: where h(x) is the anomaly score of data point x, E(h(x)) is the expected path length for the data point to be isolated, and c(n) is a constant that depends on the number of samples n.

8. A system for automatically generating statistical analysis plans and reports for clinical trial projects according to claim 4, characterized in that: The data analysis unit may use multivariate regression analysis to analyze the relationship between multiple variables, and its algorithm formula is: Y=β0+β1X1+β2X2+…+β n X n +∈, Among them, Y is the dependent variable, X1, X2, X n is the independent variable, β0 is the intercept term, β1, β2, β n is the regression coefficient of the independent variable, ∈ is the error term; The feedback generation unit may use a reinforcement learning algorithm to optimize the generated feedback, and the algorithm formula is: Among them, π * (s) is the optimal strategy in state s, Q * (s,a) is the optimal value function for taking action a in state s.

9. A system for automatically generating clinical trial project statistical analysis plans and reports according to claim 4, characterized in that: The external system integration unit may use a service orchestration algorithm during the external system integration process to ensure that the functions of multiple heterogeneous systems can be smoothly integrated and collaborated. The algorithm formula is: Among them, O is the best service orchestration solution, w ij is the weight between service i and service j, indicating their dependency or priority, x ij is a binary variable of whether service i and service j are included in the orchestration scheme; In order to prevent the API interface from being abused, the interface management unit can use a token bucket algorithm to implement the rate limit of the interface. The algorithm formula is: B(t)=min(C,B(t-1)+R·(t-t last) ), Where B(t) is the number of tokens in the bucket at time t, C is the capacity of the bucket, R is the token generation rate, and t last Is the time of the last request.

Citation Information

Patent Citations

  • Disease queue management system

    CN110364267A

  • System for automatically generating clinical test item statistical analysis plan and report

    CN116013448A

  • Mental health management and data platform system

    CN116487004A

  • Clinical test item statistical analysis planning and reporting system

    CN117542464A

Cited By

  • Automatic reporting system based on data driving

    CN119886099A