Clinical data management system and method based on cloud computing
Through a cloud-based clinical data management system, using intelligent algorithms and cloud computing technology, the shortcomings of traditional systems in data collection, storage and security are solved, and efficient, safe and compliant clinical data management is achieved.
Patent Information
- Application Number
- CN202411185250.7
- 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
The existing clinical data management system requires manual intervention when processing multi-source heterogeneous data, resulting in cumbersome integration and error-prone, low storage resource utilization, difficulty in real-time monitoring and response to security threats, and difficulty in achieving continuous compliance management.
A clinical data management system based on cloud computing was designed, including intelligent data acquisition and integration module, optimized data storage and management module, advanced data analysis and reporting module, dynamic user and permission management module, compliance data security and sharing module, The intelligent monitoring and compliance audit module uses technologies such as intelligent algorithms, cloud computing, differential privacy and homomorphic encryption to achieve automated data collection, storage, analysis, security and compliance management.
Improve data acquisition efficiency and consistency, optimize storage resource utilization, support massive data analysis, ensure data privacy and security, realize real-time compliance management, reduce the risk of manual intervention and errors, and improve the stability and security of the system. sex.
Smart Images

Figure CN119069055A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data management, and in particular to a clinical data management system and method based on cloud computing. Background Art
[0002] With the development of medical information technology, clinical data management plays an increasingly important role in modern medical care and scientific research. Clinical data comes from various sources and formats, including electronic medical records, laboratory results, imaging data, and data generated by various medical devices. Efficient collection, integration, and management of heterogeneous data have become major challenges facing medical information systems. The existing clinical data management system has the following deficiencies:
[0003] Traditional systems rely on manual intervention when processing heterogeneous data from multiple sources. The data integration process is cumbersome and error-prone, which affects the efficiency of data collection, reduces the consistency and accuracy of data between different sources, and increases uncertainty in clinical decision-making.
[0004] Traditional fixed storage methods are difficult to cope with dynamic changes in data volume, which easily leads to waste of storage resources. The lack of effective compression and hierarchical management methods leads to low utilization of storage space, slow data access speed, and the overall management efficiency needs to be improved urgently.
[0005] Existing systems usually rely on centralized computing resources to process massive amounts of clinical data, making it difficult to analyze large amounts of data within a reasonable time and difficult for clinicians and researchers to understand.
[0006] The sensitivity of medical data makes it face the risk of privacy leakage during storage, transmission and sharing. The system is unable to deal with potential security threats in a timely manner in terms of real-time monitoring and security incident response. In addition, compliance audits usually rely on manual operations, which is time-consuming and labor-intensive, and it is difficult to achieve continuous real-time compliance management, affecting the stability and security of the system.
[0007] Therefore, those skilled in the art provide a clinical data management system and method based on cloud computing to solve the problems raised in the above background technology. Summary of the invention
[0008] In view of the deficiencies in the prior art, the present invention provides a clinical data management system and method based on cloud computing to solve the problems raised in the above background technology.
[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions: a clinical data management system based on cloud computing, including an intelligent data acquisition and integration module, an optimized data storage and management module, an advanced data analysis and reporting module, a dynamic user and authority management module, a compliance data security and sharing module, and an intelligent monitoring and compliance audit module;
[0010] The intelligent data collection and integration module is responsible for automatically collecting data from multiple sources and using intelligent algorithms to standardize and integrate heterogeneous data;
[0011] The optimized data storage and management module utilizes the elastic storage capacity of cloud computing to securely store and manage large-scale clinical data through intelligent compression, hierarchical storage and efficient data retrieval algorithms;
[0012] The advanced data analysis and reporting module processes and analyzes massive clinical data through distributed computing and adaptive analysis algorithms to generate easy-to-understand reports and prediction models;
[0013] The dynamic user and authority management module uses dynamic authority allocation and multi-factor identity authentication mechanisms, and monitors user behavior in real time to prevent potential security threats;
[0014] The compliance data security and sharing module ensures the security of data during storage and transmission, protects data privacy using differential privacy and homomorphic encryption technologies, and supports cross-institutional data sharing and collaboration through intelligent compliance checks and conflict detection;
[0015] The intelligent monitoring and compliance audit module monitors the system's operating status in real time, uses anomaly detection and graph analysis algorithms to track security events, and at the same time, through an automated compliance audit mechanism, ensures that system operations and data management continue to comply with regulatory requirements.
[0016] Preferably, the intelligent data acquisition and integration module includes a data source interface unit, a data standardization unit, a data cleaning unit and a data integration unit; the data source interface unit is responsible for connecting and communicating with various data sources; the data standardization unit uses an intelligent algorithm to standardize data from different sources and unify the data format, encoding and structure; the data cleaning unit automatically detects and repairs outliers and missing values in the data; the data integration unit integrates data from different sources and uses a deep learning algorithm to identify and match related data;
[0017] The optimized data storage and management module includes an elastic storage unit, an intelligent compression unit, a hierarchical storage unit and an efficient retrieval unit; the elastic storage unit uses the elastic storage technology of cloud computing to dynamically allocate storage resources according to the amount of data; the intelligent compression unit reduces the storage space occupied by an intelligent compression algorithm; the hierarchical storage unit stores data in storage media at different levels according to the importance and access frequency of the data; the efficient retrieval unit uses an optimized retrieval algorithm to quickly find and access the stored data;
[0018] The advanced data analysis and reporting module includes a distributed computing unit, an adaptive analysis unit, a report generation unit and a prediction model unit; the distributed computing unit utilizes the distributed computing resources of cloud computing to process large-scale data; the adaptive analysis unit uses an adaptive learning algorithm to automatically adjust the analysis model according to different clinical needs; the report generation unit automatically generates easy-to-understand analysis reports combined with natural language generation technology; the prediction model unit generates a prediction model based on historical data and analysis results to assist in clinical decision-making and the formulation of preventive measures.
[0019] Preferably, the dynamic user and permission management module includes a dynamic permission allocation unit, a multi-factor identity authentication unit and a user behavior monitoring unit; the dynamic permission allocation unit dynamically allocates access rights according to user roles and tasks; the multi-factor identity authentication unit performs user identity authentication in combination with multiple identity authentication methods; the user behavior monitoring unit monitors the user's operating behavior in the system in real time and uses a behavior analysis algorithm to identify abnormal activities;
[0020] The compliance data security and sharing module includes a data encryption unit, a compliance checking unit, a conflict detection unit and a security sharing unit; the data encryption unit uses differential privacy and homomorphic encryption technology to encrypt data; the compliance checking unit automatically checks the compliance of data processing and sharing through an intelligent compliance algorithm; the conflict detection unit detects and resolves data conflicts when multiple parties share and collaborate on data; the security sharing unit supports cross-institutional data security sharing;
[0021] The intelligent monitoring and compliance audit module includes a real-time monitoring unit, an anomaly detection unit, a graph analysis unit and an automated compliance audit unit; the real-time monitoring unit monitors the operating status of the system in real time, identifies and reports abnormal situations in a timely manner; the anomaly detection unit uses a deep learning anomaly detection algorithm to automatically identify abnormal events in system operations; the graph analysis unit performs path tracking and event correlation analysis in large-scale data to quickly locate and resolve security incidents; the automated compliance audit unit performs compliance audits regularly or in real time and automatically generates compliance reports.
[0022] Preferably, in the data cleaning unit, outliers and missing values in the data are automatically detected and repaired by an outlier detection algorithm, and the algorithm formula is:
[0023]
[0024] Where x is the data point, μ is the mean, and σ is the standard deviation;
[0025] In the data integration unit, data from different sources are integrated through the cosine similarity algorithm to identify and match related data. The algorithm formula is:
[0026]
[0027] Among them, A and B are vector representations of data items, and ∥A∥ and ∥B∥ are the moduli of the vectors.
[0028] Preferably, in the intelligent compression unit, Huffman coding is used to reduce the storage space occupied, and its algorithm formula is:
[0029]
[0030] Among them, L(C) is the compressed code length, p i is the probability of symbol i, C i is the encoding length of the symbol;
[0031] In the efficient retrieval unit, the local sensitive hashing algorithm is used to quickly search and access the stored data, and its algorithm formula is:
[0032] hash(v)=sign(Wv+b),
[0033] Among them, v is the data vector, W is the random matrix, and b is the bias vector.
[0034] Preferably, in the distributed computing unit, large-scale data is processed by computing resources through the MapReduce algorithm, and its algorithm formula is:
[0035] Map(k1,v1)→list(k2,v2),
[0036] Reduce(k2,list(v2))→output(v3),
[0037] Among them, k1 and k2 are keys, and v1, v2, and v3 are values;
[0038] In the adaptive analysis unit, the analysis model is automatically adjusted according to different clinical needs through an online gradient descent algorithm, and its algorithm formula is:
[0039]
[0040] Among them, θ t is the model parameter, η is the learning rate, and J is the loss function;
[0041] In the prediction model unit, a prediction model is generated by a time series prediction algorithm to assist in clinical decision-making and formulation of preventive measures:
[0042]
[0043] Among them, X t is the predicted time series data, α is a constant term, βi is the autoregressive coefficient, ∈ t For noise.
[0044] Preferably, in the multi-factor identity authentication unit, user identity authentication is performed by an identity authentication method based on Bayesian decision making, and its algorithm formula is:
[0045]
[0046] Among them, A is the event of identity authentication passing, and B is the observed authentication information;
[0047] In the user behavior monitoring unit, abnormal activities are identified through an abnormal detection algorithm based on behavior analysis, and its algorithm formula is:
[0048]
[0049] Where X is the observed behavior, μ is the mean, and σ is the standard deviation.
[0050] Preferably, in the data encryption unit, data is encrypted by a homomorphic encryption algorithm, and the algorithm formula is:
[0051] E(m1×m2)=E(m1)×E(m2),
[0052] Among them, m1 and m2 are plain text, and E is the encryption function.
[0053] Preferably, in the real-time monitoring unit, the operating status of the system is monitored in real time through a sliding window algorithm, and abnormal conditions are identified and reported in a timely manner. The algorithm formula is:
[0054]
[0055] Among them, X i is the data point in the window, n is the window size;
[0056] In the anomaly detection unit, abnormal events in system operation are automatically identified through a Gaussian mixture model, and its algorithm formula is:
[0057]
[0058] Where P(x) is the probability density of data point x, π k is the weight of the kth Gaussian component, μ k and Σ k are the mean and covariance matrices;
[0059] In the graph analysis unit, path tracking and event association analysis are performed through a depth-first search algorithm to quickly locate and resolve security incidents. The algorithm formula is:
[0060] DFS(G,v)=visit(v),
[0061] Among them, G is a graph, v is the starting node, and visit(v) is a visit operation.
[0062] A clinical data management method based on cloud computing includes the following steps:
[0063] S1. The system connects and communicates with various data sources through the data source interface unit to automatically acquire clinical data, and then standardizes the collected heterogeneous data to unify the data format, coding and structure. Secondly, it automatically detects and repairs abnormal values and missing values in the data. Finally, it integrates data from different sources to identify and match related data.
[0064] S2, dynamically allocate cloud storage resources according to data volume and demand, then compress the data to reduce storage space usage, and secondly, store the data in different storage levels to quickly find and access the stored data;
[0065] S3. Process large-scale data, automatically adjust the analysis model according to different clinical needs, and automatically generate easy-to-understand analysis reports for reference by clinicians and researchers to generate prediction models to assist clinical decision-making and the formulation of preventive measures;
[0066] S4. Dynamically allocate and adjust access rights based on user roles and tasks to authenticate users, identify abnormal activities, and prevent potential security threats;
[0067] S5. Encrypt data, automatically check compliance during data processing and sharing, and detect and resolve data conflicts when multiple parties share and collaborate on data;
[0068] S6. Monitor the operating status of the system in real time, identify and report abnormal situations in a timely manner, and automatically identify abnormal events in system operations, quickly locate and resolve security incidents, and automatically generate compliance reports.
[0069] The present invention provides a clinical data management system and method based on cloud computing. It has the following beneficial effects:
[0070] 1. The present invention uses an intelligent data collection and integration module, data standardization and deep learning algorithms to automatically collect and integrate heterogeneous data from multiple sources, improve the efficiency of data collection, ensure the consistency and accuracy of data between different sources, and reduce human intervention and errors.
[0071] 2. The present invention optimizes the data storage and management modules and utilizes the elastic storage technology and intelligent compression algorithm of cloud computing to achieve dynamic allocation of storage resources and efficient use of storage space, thereby reducing storage costs. It can also perform hierarchical management based on the importance and access frequency of data, thereby improving the overall efficiency of data management.
[0072] 3. The advanced data analysis and reporting module of the present invention can efficiently process massive data and automatically generate easy-to-understand reports and predictive models through distributed computing, adaptive analysis and natural language generation technology, so that clinicians and researchers can quickly obtain valuable analysis results and support more accurate clinical decision-making and scientific research.
[0073] 4. The compliance data security and sharing module of the present invention adopts differential privacy and homomorphic encryption technology to ensure the privacy and security of data during storage and transmission, and intelligent compliance checking and conflict detection, supports cross-institutional data security sharing, and ensures compliance in data sharing and collaboration.
[0074] 5. The intelligent monitoring and compliance audit module of the present invention can continuously monitor the system operation status, quickly identify and respond to security incidents, and generate compliance reports through real-time monitoring, anomaly detection and automated compliance audit mechanisms. The real-time and automated system management and audit mechanisms improve the stability and security of the system. BRIEF DESCRIPTION OF THE DRAWINGS
[0075] Figure 1 is a system diagram of the present invention;
[0076] Figure 2 It is a schematic diagram of the intelligent data collection and integration module of the present invention;
[0077] Figure 3 This is a schematic diagram of an optimized data storage and management module of the present invention;
[0078] Figure 4 It is a schematic diagram of the advanced data analysis and reporting module of the present invention;
[0079] Figure 5 A schematic diagram of a dynamic user and authority management module of the present invention;
[0080] Figure 6 A schematic diagram of a compliance data security and sharing module of the present invention;
[0081] Figure 7 It is a schematic diagram of the intelligent monitoring and compliance audit module of the present invention;
[0082] Figure 8 It is a flow chart of the present invention. DETAILED DESCRIPTION
[0083] 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.
[0084] The present invention is described in detail below in conjunction with the accompanying drawings:
[0085] Example:
[0086] Please see attached Figure 1 - Attachment Figure 8 , an embodiment of the present invention provides a clinical data management system based on cloud computing, including an intelligent data acquisition and integration module, an optimized data storage and management module, an advanced data analysis and reporting module, a dynamic user and authority management module, a compliance data security and sharing module, and an intelligent monitoring and compliance audit module;
[0087] The intelligent data collection and integration module is responsible for automatically collecting data from multiple sources and using intelligent algorithms to standardize and integrate heterogeneous data;
[0088] The optimized data storage and management module utilizes the elastic storage capacity of cloud computing to securely store and manage large-scale clinical data through intelligent compression, hierarchical storage, and efficient data retrieval algorithms;
[0089] The advanced data analysis and reporting module processes and analyzes massive clinical data through distributed computing and adaptive analysis algorithms to generate easy-to-understand reports and prediction models;
[0090] The dynamic user and permission management module uses dynamic permission allocation and multi-factor authentication mechanisms, and monitors user behavior in real time to prevent potential security threats;
[0091] The compliance data security and sharing module ensures the security of data during storage and transmission, protects data privacy using differential privacy and homomorphic encryption technologies, and supports cross-institutional data sharing and collaboration through intelligent compliance checks and conflict detection;
[0092] The intelligent monitoring and compliance audit module monitors the system's operating status in real time, uses anomaly detection and graph analysis algorithms to track security incidents, and at the same time, through an automated compliance audit mechanism, ensures that system operations and data management continue to comply with regulatory requirements.
[0093] Benefits of the intelligent data collection and integration module: Reduce manual intervention, reduce the cost and time of data collection, improve efficiency, ensure data consistency and accuracy, and reduce human errors in data processing; Benefits of the cultural data storage and management module: Optimize storage costs, achieve efficient use of resources, reduce storage costs, ensure rapid retrieval and access to key data, and improve system response speed; Benefits of the advanced data analysis and reporting module: Improve the speed and efficiency of data analysis, ensure the accuracy and relevance of analysis results, and improve the effectiveness of prevention and treatment; Benefits of the dynamic user and permission management module: Avoid excessive or insufficient permissions, enhance system security, ensure that authenticated users can access the system, and prevent unauthorized access; Benefits of the compliance data security and sharing module: Ensure the security of data during storage and transmission, protect patient privacy, and reduce legal and compliance risks; Benefits of the intelligent monitoring and compliance audit module: Ensure the stability and reliability of the system, respond and handle problems quickly, prevent the expansion of potential security incidents, improve the efficiency of problem solving, and reduce the workload and errors of manual audits.
[0094] The intelligent data collection and integration module includes a data source interface unit, a data standardization unit, a data cleaning unit, and a data integration unit; the data source interface unit is responsible for connecting and communicating with various data sources; the data standardization unit uses intelligent algorithms to standardize data from different sources and unify data formats, encodings, and structures; the data cleaning unit automatically detects and repairs outliers and missing values in the data; the data integration unit integrates data from different sources and uses deep learning algorithms to identify and match related data;
[0095] The optimized data storage and management module includes an elastic storage unit, an intelligent compression unit, a hierarchical storage unit, and an efficient retrieval unit; the elastic storage unit uses the elastic storage technology of cloud computing to dynamically allocate storage resources according to the amount of data; the intelligent compression unit reduces the storage space occupied by an intelligent compression algorithm; the hierarchical storage unit stores data in storage media at different levels according to the importance and access frequency of the data; the efficient retrieval unit uses an optimized retrieval algorithm to quickly find and access stored data;
[0096] The advanced data analysis and reporting module includes a distributed computing unit, an adaptive analysis unit, a report generation unit and a predictive model unit; the distributed computing unit uses the distributed computing resources of cloud computing to process large-scale data; the adaptive analysis unit uses an adaptive learning algorithm to automatically adjust the analysis model according to different clinical needs; the report generation unit automatically generates easy-to-understand analysis reports combined with natural language generation technology; the predictive model unit generates a predictive model based on historical data and analysis results to assist in clinical decision-making and the formulation of preventive measures.
[0097] The benefits of the data source interface unit are to reduce the need for manual intervention and the risk of human error; the benefits of the data standardization unit are to improve data consistency and availability and support cross-platform data analysis and application; the benefits of the data cleaning unit are to reduce errors in the analysis process and improve the credibility of the analysis results after cleaning; the benefits of the data integration unit are to form a comprehensive data set, providing a richer foundation for subsequent complex analysis and decision support;
[0098] Benefits of elastic storage unit: avoid waste of resources, optimize storage costs, and flexibly respond to the growth or decrease of data volume, ensuring that the system can operate efficiently under various loads; Benefits of intelligent compression unit: reduce storage costs and ensure that data can be quickly decompressed and accessed when needed; Benefits of tiered storage unit: store data in different layers of media, ensure that key data can be quickly accessed, and optimize the allocation of storage resources; Benefits of efficient retrieval unit: improve the query and retrieval speed of large-scale data sets, enhance system responsiveness, and support real-time analysis and decision-making;
[0099] Benefits of the distributed computing unit: Efficiently process and analyze massive clinical data, shorten data processing time, handle complex computing tasks, and support large-scale data analysis; Benefits of the adaptive analysis unit: Automatically adjust the analysis model to ensure the accuracy and relevance of the analysis results, reduce the time for model tuning, and improve the efficiency and accuracy of data analysis; Benefits of the report generation unit: Reduce the time and workload of manual report writing, support clinical decision-making and research discoveries; Benefits of the predictive model unit: Help clinicians foresee possible medical events in advance, optimize treatment plans, and improve patient care quality and treatment outcomes.
[0100] The dynamic user and permission management module includes a dynamic permission allocation unit, a multi-factor authentication unit and a user behavior monitoring unit; the dynamic permission allocation unit dynamically allocates access rights according to user roles and tasks; the multi-factor authentication unit combines multiple authentication methods to authenticate users; the user behavior monitoring unit monitors the user's operating behavior in the system in real time and uses behavior analysis algorithms to identify abnormal activities;
[0101] The compliance data security and sharing module includes a data encryption unit, a compliance check unit, a conflict detection unit, and a security sharing unit; the data encryption unit uses differential privacy and homomorphic encryption technology to encrypt data; the compliance check unit automatically checks the compliance of data processing and sharing through intelligent compliance algorithms; the conflict detection unit detects and resolves data conflicts when multiple parties share and collaborate on data; and the security sharing unit supports cross-institutional data security sharing;
[0102] The intelligent monitoring and compliance audit module includes a real-time monitoring unit, an anomaly detection unit, a graph analysis unit and an automated compliance audit unit; the real-time monitoring unit monitors the system's operating status in real time, identifies and reports abnormal situations in a timely manner; the anomaly detection unit uses a deep learning anomaly detection algorithm to automatically identify abnormal events in system operations; the graph analysis unit performs path tracking and event correlation analysis in large-scale data to quickly locate and resolve security incidents; the automated compliance audit unit performs compliance audits regularly or in real time and automatically generates compliance reports.
[0103] The benefits of the dynamic permission allocation unit are to reduce the risk of permission abuse, improve the flexibility of the system, ensure that permission allocation is highly matched with actual needs, reduce unnecessary permission exposure, and reduce potential security risks; the benefits of the multi-factor authentication unit are to effectively prevent security vulnerabilities that may be caused by a single authentication method, enhance the overall security of the system, and protect the system from unauthorized access; the benefits of the user behavior monitoring unit are to respond to abnormal activities in a timely manner to prevent the spread of potential security threats;
[0104] Benefits of the data encryption unit: Enhance the flexibility of data use, maintain security, and always keep the data encrypted during data processing; Benefits of the compliance inspection unit: Reduce the time and errors of manual inspections, improve compliance efficiency, and reduce compliance risks; Benefits of the conflict detection unit: Effectively handle data conflicts, prevent collaboration failures or misunderstandings caused by inconsistent data, and automate conflict detection and resolution to reduce human intervention, thereby improving the efficiency and reliability of cross-institutional collaboration; Benefits of the security sharing unit: Promote collaborative research and data-driven medical decision-making, and prevent data leaks;
[0105] The benefits of the real-time monitoring unit are to prevent failures and reduce system downtime; the benefits of the anomaly detection unit are to automatically identify abnormal events in system operations to prevent security incidents from worsening; the benefits of the graph analysis unit are to quickly identify multi-level and complex security event relationships and improve the efficiency of event handling; the benefits of the automated compliance review unit are to reduce manual participation, ensure that system operations continue to comply with regulatory requirements, improve the speed and efficiency of compliance reviews, and reduce the workload of manual reviews.
[0106] In the data cleaning unit, the outlier detection algorithm is used to automatically detect and repair outliers and missing values in the data. The algorithm formula is:
[0107]
[0108] Where x is the data point, μ is the mean, and σ is the standard deviation;
[0109] In the data integration unit, data from different sources are integrated through the cosine similarity algorithm to identify and match related data. The algorithm formula is:
[0110]
[0111] Among them, A and B are vector representations of data items, and ∥A∥ and ∥B∥ are the moduli of the vectors.
[0112] If the absolute value of the Z-Score exceeds a preset threshold, the data point is considered an outlier. This can identify abnormal data points that are far away from the majority of data points. Once an outlier is identified, the data can be cleaned by replacing, deleting or correcting it. The repair process improves the overall quality of the data set and reduces the deviation caused by outliers.
[0113] The cosine similarity algorithm is used to measure the similarity between vectors. Its value is between -1 and 1. In practical applications, the focus is usually on the range between 0 and 1. A value of 1 indicates that the vector directions are exactly the same, that is, the data items are highly similar. A value of 0 indicates complete dissimilarity. In the process of data integration, cosine similarity can be used to measure the similarity between data items from different data sources. It is used for high-dimensional data to ensure that data from different sources can be effectively identified and matched to form a more complete and consistent data set.
[0114] In the intelligent compression unit, Huffman coding is used to reduce the storage space occupied. The algorithm formula is:
[0115]
[0116] Among them, L(C) is the compressed code length, p i is the probability of symbol i, C i is the encoding length of the symbol;
[0117] In the efficient retrieval unit, the local sensitive hashing algorithm is used to quickly find and access stored data. The algorithm formula is:
[0118] hash°v)=sign(Wv+b),
[0119] Among them, v is the data vector, W is the random matrix, and b is the bias vector.
[0120] Huffman coding is a lossless compression algorithm based on the frequency of symbol occurrence. By assigning shorter codes to high-frequency symbols and longer codes to low-frequency symbols, Huffman coding can effectively reduce the overall storage space occupied by data. This formula calculates the total length of the symbol after encoding, and the storage efficiency of the compressed data is improved;
[0121] The Locality Sensitive Hash (LSH) algorithm is an algorithm used for fast approximate nearest neighbor search in high-dimensional data. By mapping high-dimensional data to a low-dimensional space, the LSH algorithm enables similar data items to have the same or similar hash values in the low-dimensional space, and can quickly find similar data items, significantly reducing the dimensions and amount of data that need to be compared.
[0122] In the distributed computing unit, the MapReduce algorithm is used to calculate resources and process large-scale data. The algorithm formula is:
[0123] Map(k1,v1)→list(k2,v2),
[0124] Reduce(k2,list(v2))→output(v3),
[0125] Among them, k1 and k2 are keys, and v1, v2, and v3 are values;
[0126] In the adaptive analysis unit, the analysis model is automatically adjusted according to different clinical needs through the online gradient descent algorithm. The algorithm formula is:
[0127] θ t+1 =θ t -η* θ J(θ t ;x t ,y t ),
[0128] Among them, θ t is the model parameter, η is the learning rate, and J is the loss function;
[0129] In the prediction model unit, a prediction model is generated through a time series prediction algorithm to assist in clinical decision-making and the formulation of preventive measures:
[0130]
[0131] Among them, X t is the predicted time series data, α is a constant term, β i is the autoregressive coefficient, ∈ t For noise.
[0132] MapReduce is a distributed computing framework used to process large data sets. The Map function splits the input data into key-value pairs and performs preliminary processing to generate a list of intermediate key-value pairs. The Reduce function aggregates or further processes the value list with the same intermediate key to generate the final output.
[0133] The online gradient descent algorithm is an incremental learning algorithm that can update model parameters in real time when data streams continue to arrive, so that the model can quickly adapt to new data changes and ultimately find the optimal solution for model parameters, keeping the model up to date and accurate.
[0134] Time series prediction algorithms are used in autoregressive models. The value at the predicted moment is based on the data at the current time point. By modeling the historical data of the time series, trends and seasonal changes in the data can be captured, and future changes in the condition can be predicted. This helps clinicians identify potential risks in advance, develop preventive measures or adjust treatment plans, thereby improving the quality of patient care and treatment outcomes and reducing the incidence of adverse events.
[0135] In the multi-factor authentication unit, user authentication is performed through an authentication method based on Bayesian decision making, and its algorithm formula is:
[0136]
[0137] Among them, A is the event of identity authentication passing, and B is the observed authentication information;
[0138] In the user behavior monitoring unit, abnormal activities are identified through an anomaly detection algorithm based on behavior analysis. The algorithm formula is:
[0139]
[0140] Where X is the observed behavior, μ is the mean, and σ is the standard deviation.
[0141] The Bayesian decision-making based identity authentication algorithm uses Bayes’ theorem to combine prior information and verification information to calculate the posterior probability that the user’s identity is true. By evaluating the probability, the system can determine the authenticity of the user’s identity and reduce the risk of misjudgment;
[0142] The anomaly detection algorithm calculates the Z-Score of user behavior characteristics. The system can identify operations that are significantly different from normal behavior patterns, such as abnormal login time, access to unauthorized resources, etc. The anomalies indicate potential security threats. If abnormal behavior is detected, it can immediately trigger an early warning or take corresponding measures to quickly respond to possible security incidents and reduce damage.
[0143] In the data encryption unit, data is encrypted using a homomorphic encryption algorithm, and its algorithm formula is:
[0144] E(m1×m2)=E(m1)×E(m2),
[0145] Among them, m1 and m2 are plain text, and E is the encryption function.
[0146] Homomorphic encryption allows mathematical operations to be performed directly on encrypted data without decrypting the data. The formula specifically indicates that a multiplication operation is performed on the encrypted ciphertext, and the result is equivalent to performing a multiplication operation on the plaintext before encrypting it. This means that the system can calculate the data without exposing the plaintext, which is useful for protecting data privacy. Through homomorphic encryption, the system can process data in an encrypted state and perform operations without decrypting the data, reducing the risk of data leakage during the calculation process and ensuring the privacy of the data.
[0147] In the real-time monitoring unit, the sliding window algorithm is used to monitor the operating status of the system in real time, and abnormal situations are identified and reported in a timely manner. The algorithm formula is:
[0148]
[0149] Among them, X i is the data point in the window, n is the window size;
[0150] In the anomaly detection unit, the Gaussian mixture model is used to automatically identify abnormal events in system operations. The algorithm formula is:
[0151]
[0152] Where P(x) is the probability density of data point x, π k is the weight of the kth Gaussian component, μ k and Σ k are the mean and covariance matrices;
[0153] In the graph analysis unit, the depth-first search algorithm is used to perform path tracking and event correlation analysis to quickly locate and resolve security incidents. The algorithm formula is:
[0154] DFS(G,v)=visit(v),
[0155] Among them, G is a graph, v is the starting node, and visit(v) is a visit operation.
[0156] The sliding window algorithm is used to calculate the real-time average value of data within the sliding window. The formula helps to monitor the operation status of the system in real time, effectively smooth out short-term fluctuations, help identify long-term trends or abnormal behaviors of the system, and promptly discover and report potential abnormal situations.
[0157] The Gaussian mixture model is used to model complex data distribution. It accurately describes the multimodal distribution of data through the linear combination of multiple Gaussian distributions, which means that the data point is an outlier.
[0158] Depth-first search is an algorithm for traversing or searching a graph. By traversing the nodes and edges in the graph, DFS can effectively discover the dependencies and causal chains between events, helping security analysts quickly locate the root cause of the problem and deeply analyze the propagation path of the event.
[0159] A clinical data management method based on cloud computing includes the following steps:
[0160] S1. The system connects and communicates with various data sources through the data source interface unit to automatically acquire clinical data, and then standardizes the collected heterogeneous data to unify the data format, coding and structure. Secondly, it automatically detects and repairs abnormal values and missing values in the data. Finally, it integrates data from different sources to identify and match related data.
[0161] S2, dynamically allocate cloud storage resources according to data volume and demand, then compress the data to reduce storage space usage, and secondly, store the data in different storage levels to quickly find and access the stored data;
[0162] S3. Process large-scale data, automatically adjust the analysis model according to different clinical needs, and automatically generate easy-to-understand analysis reports for reference by clinicians and researchers to generate prediction models to assist clinical decision-making and the formulation of preventive measures;
[0163] S4. Dynamically allocate and adjust access rights based on user roles and tasks to authenticate users, identify abnormal activities, and prevent potential security threats;
[0164] S5. Encrypt data, automatically check compliance during data processing and sharing, and detect and resolve data conflicts when multiple parties share and collaborate on data;
[0165] S6. Monitor the operating status of the system in real time, identify and report abnormal situations in a timely manner, and automatically identify abnormal events in system operations, quickly locate and resolve security incidents, and automatically generate compliance reports.
[0166] 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 clinical data management system based on cloud computing, characterized in that: It includes intelligent data collection and integration module, optimized data storage and management module, advanced data analysis and reporting module, dynamic user and permission management module, compliance data security and sharing module, intelligent monitoring and compliance audit module; The intelligent data collection and integration module is responsible for automatically collecting data from multiple sources and using intelligent algorithms to standardize and integrate heterogeneous data; The optimized data storage and management module utilizes the elastic storage capacity of cloud computing to securely store and manage large-scale clinical data through intelligent compression, hierarchical storage and efficient data retrieval algorithms; The advanced data analysis and reporting module processes and analyzes massive clinical data through distributed computing and adaptive analysis algorithms to generate easy-to-understand reports and prediction models; The dynamic user and authority management module uses dynamic authority allocation and multi-factor identity authentication mechanisms, and monitors user behavior in real time to prevent potential security threats; The compliance data security and sharing module ensures the security of data during storage and transmission, protects data privacy using differential privacy and homomorphic encryption technologies, and supports cross-institutional data sharing and collaboration through intelligent compliance checks and conflict detection; The intelligent monitoring and compliance audit module monitors the system's operating status in real time, uses anomaly detection and graph analysis algorithms to track security events, and at the same time, through an automated compliance audit mechanism, ensures that system operations and data management continue to comply with regulatory requirements.
2. A clinical data management system based on cloud computing according to claim 1, characterized in that: The intelligent data acquisition and integration module includes a data source interface unit, a data standardization unit, a data cleaning unit and a data integration unit; the data source interface unit is responsible for connecting and communicating with various data sources; the data standardization unit uses intelligent algorithms to standardize data from different sources and unify data formats, encodings and structures; the data cleaning unit automatically detects and repairs outliers and missing values in the data; the data integration unit integrates data from different sources and uses deep learning algorithms to identify and match related data; The optimized data storage and management module includes an elastic storage unit, an intelligent compression unit, a hierarchical storage unit and an efficient retrieval unit; The elastic storage unit utilizes the elastic storage technology of cloud computing to dynamically allocate storage resources according to the amount of data; The intelligent compression unit reduces the storage space occupied by the intelligent compression algorithm; The hierarchical storage unit stores data in storage media at different levels according to the importance and access frequency of the data; The efficient retrieval unit uses an optimized retrieval algorithm to quickly find and access stored data; The advanced data analysis and reporting module includes a distributed computing unit, an adaptive analysis unit, a report generation unit and a prediction model unit; the distributed computing unit utilizes the distributed computing resources of cloud computing to process large-scale data; the adaptive analysis unit uses an adaptive learning algorithm to automatically adjust the analysis model according to different clinical needs; the report generation unit automatically generates easy-to-understand analysis reports combined with natural language generation technology; the prediction model unit generates a prediction model based on historical data and analysis results to assist in clinical decision-making and the formulation of preventive measures.
3. A clinical data management system based on cloud computing according to claim 1, characterized in that: The dynamic user and permission management module includes a dynamic permission allocation unit, a multi-factor identity authentication unit and a user behavior monitoring unit; the dynamic permission allocation unit dynamically allocates access rights according to user roles and tasks; The multi-factor authentication unit combines multiple authentication methods to perform user authentication; the user behavior monitoring unit monitors the user's operating behavior in the system in real time and uses a behavior analysis algorithm to identify abnormal activities; The compliance data security and sharing module includes a data encryption unit, a compliance checking unit, a conflict detection unit and a security sharing unit; the data encryption unit uses differential privacy and homomorphic encryption technology to encrypt data; the compliance checking unit automatically checks the compliance of data processing and sharing through an intelligent compliance algorithm; the conflict detection unit detects and resolves data conflicts when multiple parties share and collaborate on data; The secure sharing unit supports secure sharing of data across institutions; The intelligent monitoring and compliance audit module includes a real-time monitoring unit, an anomaly detection unit, a graph analysis unit, and an automated compliance audit unit; The real-time monitoring unit monitors the operating status of the system in real time, and promptly identifies and reports abnormal situations; the anomaly detection unit uses a deep learning anomaly detection algorithm to automatically identify abnormal events in system operations; the graph analysis unit performs path tracking and event correlation analysis in large-scale data to quickly locate and resolve security incidents; the automated compliance audit unit performs compliance audits regularly or in real time and automatically generates compliance reports.
4. A cloud computing-based clinical data management system according to claim 2, characterized in that: In the data cleaning unit, outliers and missing values in the data are automatically detected and repaired through an outlier detection algorithm, and the algorithm formula is: Where x is the data point, μ is the mean, and σ is the standard deviation; In the data integration unit, data from different sources are integrated through the cosine similarity algorithm to identify and match related data. The algorithm formula is: Among them, A and B are vector representations of data items, and ∥A∥ and ∥B∥ are the moduli of the vectors.
5. A clinical data management system based on cloud computing according to claim 2, characterized in that: In the intelligent compression unit, Huffman coding is used to reduce the storage space occupied, and its algorithm formula is: Among them, L(C) is the compressed code length, p i is the probability of symbol i, C i is the encoding length of the symbol; In the efficient retrieval unit, the local sensitive hashing algorithm is used to quickly search and access the stored data, and its algorithm formula is: hash(v)=sign(Wv+b), Among them, v is the data vector, W is the random matrix, and b is the bias vector.
6. A cloud computing-based clinical data management system according to claim 2, characterized in that: In the distributed computing unit, large-scale data is processed by computing resources through the MapReduce algorithm, and its algorithm formula is: Map(k1,v1)→list(k2,v2), Reduce(k2,list(v2))→output(v3), Among them, k1 and k2 are keys, and v1, v2, and v3 are values; In the adaptive analysis unit, the analysis model is automatically adjusted according to different clinical needs through an online gradient descent algorithm, and its algorithm formula is: Among them, θ t is the model parameter, η is the learning rate, and J is the loss function; In the prediction model unit, a prediction model is generated by a time series prediction algorithm to assist in clinical decision-making and formulation of preventive measures: Among them, X t is the predicted time series data, α is a constant term, β i is the autoregressive coefficient, ∈ t For noise.
7. A cloud computing-based clinical data management system according to claim 3, characterized in that: In the multi-factor authentication unit, user identity authentication is performed through an authentication method based on Bayesian decision making, and its algorithm formula is: Among them, A is the event of identity authentication passing, and B is the observed authentication information; In the user behavior monitoring unit, abnormal activities are identified through an abnormal detection algorithm based on behavior analysis, and its algorithm formula is: Where X is the observed behavior, μ is the mean, and σ is the standard deviation.
8. The cloud computing-based clinical data management system according to claim 3, characterized in that: In the data encryption unit, data is encrypted by a homomorphic encryption algorithm, and the algorithm formula is: E(m1×m2)=E(m1)×E(m2), Among them, m1 and m2 are plain text, and E is the encryption function.
9. A cloud computing-based clinical data management system according to claim 3, characterized in that: In the real-time monitoring unit, the operating status of the system is monitored in real time through a sliding window algorithm, and abnormal situations are identified and reported in a timely manner. The algorithm formula is: Among them, X i is the data point in the window, n is the window size; In the anomaly detection unit, abnormal events in system operation are automatically identified through a Gaussian mixture model, and its algorithm formula is: Where P(x) is the probability density of data point x, π k is the weight of the kth Gaussian component, μ k and Σ k are the mean and covariance matrices; In the graph analysis unit, path tracking and event association analysis are performed through a depth-first search algorithm to quickly locate and resolve security incidents. The algorithm formula is: DFS°G,v)=visit°v), Among them, G is a graph, v is the starting node, and visit°v) is a visit operation.
10. A clinical data management method based on cloud computing, according to a clinical data management system based on cloud computing according to any one of claims 1 to 9, characterized in that: The steps include: S1. The system connects and communicates with various data sources through the data source interface unit to automatically acquire clinical data, and then standardizes the collected heterogeneous data to unify the data format, coding and structure. Secondly, it automatically detects and repairs abnormal values and missing values in the data. Finally, it integrates data from different sources to identify and match related data. S2, dynamically allocate cloud storage resources according to data volume and demand, then compress the data to reduce storage space usage, and secondly, store the data in different storage levels to quickly find and access the stored data; S3. Process large-scale data, automatically adjust the analysis model according to different clinical needs, and automatically generate easy-to-understand analysis reports for reference by clinicians and researchers to generate prediction models to assist clinical decision-making and the formulation of preventive measures; S4. Dynamically allocate and adjust access rights based on user roles and tasks to authenticate users, identify abnormal activities, and prevent potential security threats; S5. Encrypt data, automatically check compliance during data processing and sharing, and detect and resolve data conflicts when multiple parties share and collaborate on data; S6. Monitor the operating status of the system in real time, identify and report abnormal situations in a timely manner, automatically identify abnormal events in system operations, quickly locate and resolve security incidents, and automatically generate compliance reports.
Citation Information
Patent Citations
Identity authentication method and system
CN107483433A
Data management system for intelligent operation and maintenance service
CN117194399A
Data interaction system for archive management
CN117827743A
Marine emergency mobilization command platform system and processing method
CN118246708A
An intelligent data quality management system for improved AI / ML model performance across various industries.
DE202024104069U1
Cited By
Clinical test data storage and management system based on cloud computing
CN119479970A
Wounded information management and sharing method and system based on cloud computing
CN120032832A
