Medical safety process supervision method and system based on data analysis

Through full-process monitoring and action decomposition technology, abnormal operations are identified, risks are evaluated and automated supervision firmware is designed, which solves the problems of low error operation recognition accuracy and slow warning speed in traditional medical safety process supervision methods, and improves the monitoring efficiency of medical safety process and Security.

CN118969229AInactive Publication Date: 2024-11-15CHENGDU YANCHONGWEI TECHNOLOGY CO LTD

Patent Information

Application Number
CN202411450903.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-17
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional medical safety process supervision methods based on data analysis have problems such as low error operation identification accuracy and slow warning speed, making it difficult to effectively monitor and prevent risks in the medical process.

Method used

Through electronic monitoring equipment, the full-process experimental operation process monitoring is carried out, the operation process actions are decomposed, behavior parameters are extracted, abnormal operation action characteristics are refined and posture risk assessment is performed, safety boundaries are reconstructed, potential risk points are identified, and automated supervision firmware is designed to Realize real-time monitoring and early warning.

Benefits of technology

It improves the accuracy of misoperation identification, shortens the early warning reaction time, significantly improves the safety and efficiency of laboratory operations, and reduces the probability of equipment damage and experiment failure.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118969229A_ABST
    Figure CN118969229A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of safety process supervision, in particular to a medical safety process supervision method and system based on data analysis. The method comprises the following steps of performing whole-process experiment operation process monitoring through electronic monitoring equipment, and performing operation process action decomposition to obtain operation process action decomposition data; performing action behavior parameter extraction and action posture risk evolution evaluation on the operation process action decomposition data to obtain operation posture risk evolution evaluation data; performing hidden danger triggering risk point preposition identification based on the operation posture risk evolution evaluation data to obtain hidden danger triggering risk point preposition data; and performing automatic supervision process firmware design according to the pre-data of the hidden danger triggering risk point to obtain potential risk logic supervision firmware. According to the invention, the safety process supervision technology is optimized, so that the safety process supervision technology is more perfect.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of safety process supervision, and in particular to a medical safety process supervision method and system based on data analysis. Background Art

[0002] With the rapid development of medical technology and the deepening of hospital informatization, medical safety issues are becoming increasingly complex and diverse, including risk management during surgery, safety monitoring of drug use, and privacy protection of patient information. Traditional medical safety supervision relies on manual records and experience management, and has many drawbacks such as information islands, data opacity, and lagging supervision, making it difficult to timely discover and prevent risks and hidden dangers in the medical process. In addition, medical data is complex and multi-source, including electronic medical records, imaging data, surgical records, laboratory test results, nursing records, etc. How to effectively integrate these data and conduct scientific analysis is an important challenge to improve the efficiency of medical safety supervision. The medical safety process supervision method based on data analysis is proposed in this context. By integrating multi-source data, it conducts all-round real-time monitoring and analysis of medical behaviors and processes. This method uses advanced data analysis technologies such as data mining, correlation analysis, and anomaly detection to carefully monitor key nodes in the medical operation process and identify potential safety hazards and abnormal behaviors. For example, by analyzing operating room equipment data, doctor's surgical records and patient physiological monitoring data, abnormal situations during surgery can be discovered in real time and timely warnings can be issued; by monitoring drug usage data, unreasonable drug combinations or dosage issues can be discovered, thereby reducing medication risks. In addition, by combining patient historical data and real-time medical behavior data, individual patient medical risks can be personalized and accurate medical decision support can be provided. However, the traditional medical safety process supervision method based on data analysis has the problems of low accuracy in identifying misoperations and slow warning speed. Summary of the invention

[0003] Based on this, it is necessary to provide a medical safety process supervision method and system based on data analysis to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a medical safety process supervision method based on data analysis is provided, the method comprising the following steps: Step S1: monitor the whole process of the experiment operation process through electronic monitoring equipment to obtain the whole process of the experiment operation process monitoring video; decompose the operation process action based on the whole process of the experiment operation process monitoring video to obtain the operation process action decomposition data; Step S2: extracting action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; performing abnormal operation action feature refinement processing on the operation process action behavior parameters to obtain abnormal operation action feature refinement data; performing action posture risk evolution assessment on the abnormal operation action feature refinement data to obtain operation posture risk evolution assessment data; Step S3: reconstructing the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain action posture safety boundary data; performing pre-identification of hidden danger triggering risk points based on the action posture safety boundary data to obtain hidden danger triggering risk point pre-data; Step S4: Design automated supervision process firmware based on the pre-data of hidden danger triggering risk points to obtain potential risk logic supervision firmware; embed the potential risk logic supervision firmware into the data cloud platform to perform security process supervision.

[0005] The present invention monitors the whole process of experimental operation process through electronic monitoring equipment, which can systematically capture every detail of the whole process of the experiment. The core of this stage is to obtain a comprehensive experimental process video to ensure that all operation steps, behaviors and environmental factors are recorded. The video data provides a detailed basis for subsequent analysis, allowing a comprehensive review of the experimental process. Based on these monitoring videos, the operation process action decomposition is further performed. This process includes breaking down complex operation steps into specific actions so that each link can be analyzed more carefully. This decomposition not only helps to identify the specific details of each step of the operation, but also reveals the potential operation mode and behavior rules, laying a solid foundation for subsequent parameter extraction and behavior analysis. The ultimate goal of this stage is to form structured operation process data so that subsequent analysis can be more accurate and efficient. After completing the operation process action decomposition, the next step is to extract action behavior parameters from these data. This process involves extracting key behavior parameters from the operation process, such as action speed, frequency, angle, etc. These parameters help to quantify the characteristics of each action, making it possible to standardize and compare behaviors. The extracted parameters will then be used for the refinement of abnormal operation action features. By conducting an in-depth analysis of abnormal operations, we can identify features that are significantly different from normal behavior patterns, which indicate potential risks or operational errors. Next, we conduct an action posture risk evolution assessment on these refined data, which is to predict and evaluate the risk evolution of different operating postures in long-term use. This assessment can reveal safety hazards in operation and provide a scientific basis for improving operating specifications and optimizing the operating environment, thereby ultimately improving the safety and efficiency of the experiment. After obtaining the operation posture risk evolution assessment data, the next step is to reconstruct the action posture safety boundary based on the abnormal operation action feature refinement data. The core of this process is to determine the safety range and boundaries of each operation posture. These safety boundaries are defined based on the historical data of the action and the risk assessment results, ensuring that operations within these boundaries are relatively safe. Through precise boundary setting, the safety risks caused by improper posture during operation can be significantly reduced. Subsequently, the risk point of hidden danger triggering is pre-identified based on the constructed action posture safety boundary data. The purpose of this stage is to identify potential points that trigger risks during operation. Through the analysis of the safety boundary, those operating behaviors or environmental factors that may cause safety hazards can be discovered in advance. Pre-identification of hidden danger triggering risk points enables appropriate measures to be taken before problems occur, thereby greatly improving the safety and reliability of laboratory operations. This preventive risk management strategy not only helps protect the safety of laboratory personnel, but also reduces the probability of equipment damage and experimental failure. Based on the pre-identification data of hidden danger triggering risk points, the next step is to design an automated supervision process firmware. The focus of this stage is to develop a logical supervision firmware that can automatically monitor and manage laboratory operation risks.This firmware will design an automated supervision logic for potential risk points based on the pre-identified data, including functions such as real-time monitoring, alarm triggering and automatic adjustment. This process enables the laboratory to respond in a timely manner when abnormal operations occur, thereby reducing the need for manual intervention and improving operational safety and efficiency. Finally, the potential risk logic supervision firmware is embedded in the data cloud platform. This step involves integrating the designed firmware into the existing laboratory management system so that it can perform safety process supervision in real time. Through this integration, the laboratory can achieve comprehensive safety monitoring, automatically identify and respond to various potential risks, and ensure the standardization and safety of the experimental process. The embedded system can not only provide real-time feedback on risk information in the operation, but also provide targeted safety suggestions and measures, thereby further optimizing the laboratory's operating environment and safety management level. Therefore, the present invention is an optimization of a traditional medical safety process supervision method based on data analysis, which solves the problem that a traditional medical safety process supervision method based on data analysis has low accuracy in identifying misoperations and slow warning speed, improves the accuracy of misoperation identification, and improves the speed of warning.

[0006] Preferably, step S1 comprises the following steps: Step S11: monitor the whole process of experimental operation process through electronic monitoring equipment to obtain a monitoring video of the whole process of experimental operation process; Step S12: performing contrast optimization processing on the whole-process experimental operation process monitoring video to obtain an operation process monitoring contrast-optimized video; Step S13: switching and integrating the operation process monitoring contrast-optimized video from multiple angles to obtain a multi-angle integrated video of the operation process monitoring; Step S14: Decompose the operation process actions based on the multi-angle integrated video of the operation process monitoring to obtain the operation process action decomposition data.

[0007] The present invention uses electronic monitoring equipment to monitor the operation of the entire process of the experiment, which is the first step to ensure that the experimental process is accurately recorded. The key to this stage is to fully capture every operational detail in the experiment, including operating steps, equipment status and environmental conditions. The acquired full-process experimental operation process monitoring video provides a complete view of the entire experimental process, ensuring that all operating actions can be traced back and analyzed. The advantage of this comprehensive record is that it can truly reproduce the actual operation process of the experiment, providing basic data support for subsequent data processing and analysis. The result of this step is a detailed experimental video, which lays a solid foundation for subsequent image processing and motion analysis. The contrast optimization processing of the full-process experimental operation process monitoring video is to improve the visualization effect of the video and make the details in the experiment clearer and more obvious. The optimization processing can enhance the visibility of important operation details in the video, especially in low light or low contrast. By adjusting the contrast, the operation steps, action details and equipment status in the video can be made more prominent, reducing the information loss caused by visual blur or low contrast. The result of this step is an operation process monitoring contrast-optimized video, and such optimization makes subsequent motion analysis and data extraction more accurate, ensuring the reliability of the analysis. The purpose of integrating the contrast-optimized video of the operation process monitoring with multiple angles is to obtain a comprehensive view of the operation process through the integration of different perspectives. By switching from multiple angles, the experimental operation can be observed from different perspectives, which is particularly important for capturing the subtle differences and angle changes in the operation. The video after integrating these perspectives enables the observer to understand the entire process of the operation from multiple angles, so as to analyze the operation behavior more comprehensively. The benefit of this step is that it can provide multi-dimensional information, which helps to more accurately identify the details and potential problems in the operation, and provide more comprehensive data support for the subsequent action decomposition. Decomposing the operation process based on the multi-angle integrated video of the operation process monitoring is to break down the complex experimental operation process into more detailed action steps. Through this process, each specific action in the experiment can be identified and marked, and these actions can be classified according to order and function. The action decomposition data can help to understand the specific details of each operation step and its role in the entire experimental process in detail. This refined data makes the subsequent behavior analysis more targeted, can accurately identify abnormalities or potential risks in the operation, and provides detailed basic data for further parameter extraction and risk assessment.

[0008] Preferably, step S2 comprises the following steps: Step S21: Obtaining standardized operation process data; Step S22: extracting action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; Step S23: performing abnormal operation action identification on the operation process action behavior parameters according to the standardized operation process data to obtain abnormal operation action data; Step S24: performing abnormal operation action feature refinement processing on the operation process action behavior parameters according to the abnormal operation action data to obtain abnormal operation action feature refinement data; Step S25: Perform action posture risk evolution assessment on the abnormal operation action feature refinement data to obtain operation posture risk evolution assessment data.

[0009] The acquisition of standardized operation process data by the present invention is a basic step to ensure that the experimental operation meets the predetermined standards and specifications. Standardized operation process data generally refers to standardized steps and procedures in the operation process. These data can help formulate standard operation processes by comparing actual operations with standardized operation standards. Through standardization, it can be ensured that each experimental step is carried out in accordance with best practices, thereby minimizing the variability and uncertainty in the operation. This process provides a clear operation benchmark, so that subsequent data analysis can specifically identify which operations are in line with the standards, which have deviations or need to be improved. This is of great significance for improving the accuracy and consistency of the experiment, and lays a standardized data foundation for anomaly detection and risk assessment. The operation process action decomposition data is subjected to action behavior parameter extraction, the purpose of which is to extract key behavioral features from detailed operation actions. These parameters include the frequency, duration, speed, angle, etc. of the action, which can quantify the characteristics of each operation step. The advantage of this step is that the complex behavior in the operation process is converted into analyzable parameter data, so that the subsequent behavior analysis is more systematic and objective. The extracted parameter data provides a basis for identifying anomalies in the operation and optimizing the operation process, making the monitoring and improvement of the experimental operation more accurate and well-founded. The identification of abnormal operation actions based on the action behavior parameters of the operation process according to the standardized operation process data is to identify those actions that deviate from the normal operation range by comparing the actual operation parameters with the standard standards. This process involves comparing the extracted action behavior parameters with the standard operation data to determine which operation actions do not meet the specifications or have abnormalities. This identification can help discover potential problems in the operation process, such as errors in operation steps, the influence of environmental factors, or equipment failures. By identifying abnormal operation actions, early warnings and adjustments can be made, thereby effectively reducing operation risks and improving the safety and reliability of experiments. After identifying abnormal operation actions, the next step is to perform feature refinement on these abnormal actions. This process includes in-depth analysis of the specific characteristics of abnormal operations, such as the change pattern of actions, abnormal frequency, or performance under specific conditions. This refinement can help understand the specific causes and effects of abnormal operations and provide detailed abnormal data analysis results. Through feature refinement, the potential root causes of abnormal operations can be revealed and targeted suggestions can be provided for improving operation processes. The benefit of this step is that it can accurately locate problems, help to make effective corrections and optimizations, and further improve the standardization and safety of operations. The risk evolution assessment of action postures based on the refined data of abnormal operation action features is to evaluate the risk changes of different operation postures over time. This process involves long-term analysis of abnormal operation data to predict the risks these abnormal postures will bring in the future. For example, some operation postures may seem harmless in the short term, but long-term use may lead to equipment wear, operator fatigue, or safety hazards.Through this assessment, long-term potential risk trends can be identified and preventive improvement suggestions can be provided. The benefit of this step is that it can achieve forward-looking risk management, reduce future risks and damages by timely adjusting operating postures and processes, and ensure the long-term safety and efficiency of the experiment.

[0010] Preferably, step S24 comprises the following steps: Step S241: calculating the abnormal operation action dynamic deviation of the operation process action behavior parameters according to the abnormal operation action data to obtain the abnormal operation action dynamic deviation data; Step S242: performing a multi-dimensional parameter heterogeneous comparison on the operation process action behavior parameters based on the abnormal operation action dynamic deviation data to obtain multi-dimensional parameter heterogeneous data; Step S243: performing microscopic motion scale analysis according to the abnormal operation motion dynamic deviation data and the multi-dimensional parameter heterogeneous data to obtain multi-dimensional microscopic motion scale data; Step S244: performing a subtle motion feature deep analysis on the multi-dimensional micro-motion scale data to obtain multi-dimensional micro-motion scale deep analysis data; Step S245: performing abnormal operation action feature refinement processing according to the multi-dimensional micro-action scale depth analysis data to obtain abnormal operation action feature refinement data.

[0011] The present invention performs dynamic deviation calculation on the action behavior parameters of the operation process based on abnormal operation action data, and the main purpose is to identify and quantify the dynamic anomalies that occur during the operation process. This calculation focuses on the deviation between the actual performance of each operation step and the standard specification, especially those deviations that occur over time or environmental changes. By analyzing these dynamic deviations, it is possible to understand how abnormal operation actions occur at different time points or under different operating conditions. This deviation calculation can reveal the instability or inconsistency in the operation process, and provide data support for subsequent abnormal feature identification and risk assessment. Its beneficial effect is that the deviation information captured by dynamic analysis provides an in-depth perspective for a comprehensive understanding and improvement of the operation process, thereby effectively reducing operation errors and improving the stability and reliability of the experimental process. Multidimensional parameter heterogeneous comparison is performed on the dynamic deviation data of abnormal operation actions. The main purpose of this step is to analyze the characteristics of abnormal operations through different parameter dimensions and compare these characteristics with standard operation data. Multidimensional parameter heterogeneous comparison allows multiple operation parameters (such as speed, angle, time, etc.) to be considered simultaneously, and abnormal behaviors existing in different dimensions are identified. Such a comparison can reveal the relationship between the parameters and how they jointly affect the operation results, thereby helping to gain a deep understanding of the root cause of the abnormality. Through this multi-dimensional analysis, the abnormal conditions in the operation process can be comprehensively evaluated, and more complex potential problems can be identified, which provides a basis for accurate abnormal analysis and processing, thereby improving the overall optimization and improvement effect of the operation process. Micro-motion scale analysis is performed based on the dynamic deviation data of abnormal operation actions and multi-dimensional parameter heterogeneous data, aiming to deeply explore the detailed characteristics of abnormal operations. Micro-motion scale analysis focuses on small-scale actions in the operation process and analyzes how these small-scale actions lead to abnormalities in the overall operation. By refining the analysis to the micro level, specific details in abnormal operations can be identified, such as irregular operation gestures or slight posture deviations. The beneficial effect of this step is that it provides a deep understanding of operational abnormalities, so that details that are ignored by rough analysis can be discovered, so as to locate and process abnormalities more accurately. This meticulous analysis helps to improve the operation process, improve the accuracy and efficiency of operations, and reduce the hidden dangers in operations. The main purpose of deep analysis of subtle action features of multi-dimensional micro-motion scale data is to extract deeper features from micro-scale action data. This step identifies the specific role and impact of each subtle action in the overall operation through in-depth analysis. In-depth analysis of subtle action features helps reveal specific action features that show inconsistencies or anomalies during the operation process, and provides detailed explanations for understanding the causes of these features. This in-depth analysis can help understand the details of the abnormal operation and better identify the specific actions or operation links that cause the abnormality, thereby providing detailed basis for subsequent improvements and optimizations, and ensuring the reliability and safety of the operation process.Refine the abnormal operation action features based on the multi-dimensional micro-action scale deep analysis data. The core of this step is to further refine and organize the deep-level action features obtained from the micro-scale analysis in order to more accurately describe and classify the abnormal operation features. Through refinement, the specific features of the abnormal operation can be decomposed into more detailed categories and sub-features, which helps to identify potential problem areas in the operation process and make targeted improvements. The refinement of abnormal operation action features can not only enhance the understanding of abnormal operations, but also provide support for the formulation of more effective control measures and improvement strategies. The beneficial effect of this step is that it provides precise details for the optimization of the operation process, helping to improve the standardization and safety of the overall operation.

[0012] Preferably, step S25 comprises the following steps: Step S251: performing posture stress distribution analysis on the abnormal operation action feature refinement data to obtain posture stress distribution data; Step S252: performing microseismic force decoupling processing on the abnormal operation action feature refinement data according to the posture stress distribution data to obtain posture action microseismic force decoupling data; Step S253: performing displacement residual analysis on the posture action microseismic force decoupling data to obtain microseismic force displacement residual data; Step S254: Perform risk evolution assessment of the action posture based on the microseismic force decoupling data of the posture action and the microseismic force displacement residual data to obtain the risk evolution assessment data of the operation posture.

[0013] The present invention performs posture stress distribution analysis on the abnormal operation action feature refinement data, the main purpose of which is to reveal the stress distribution of the posture on the operation object or human body during the operation process. This process helps identify the areas that cause excessive stress or stress concentration by calculating and visualizing the stress distribution under different postures during the operation process. The posture stress distribution data can provide information about the impact of posture on the overall operation, including how to cause overuse or fatigue of certain parts. Through this analysis, the stress distribution pattern in abnormal operation actions can be better understood, and then those postures that pose risks to the operation process or the health of operators can be identified. This analysis helps to optimize the operation posture, reduce injuries or equipment failures caused by improper posture, and thus improve the safety and efficiency of the overall operation. Microseismic force decoupling processing is performed on the abnormal operation action feature refinement data according to the posture stress distribution data. Microseismic force decoupling processing refers to further analyzing the small vibrations and force changes during the operation process on the basis of the posture stress distribution, and separating them from the total stress distribution. This processing can identify the local forces generated by those small vibrations, which are generated during the operation process and affect the operation quality or safety. Through decoupling processing, complex stress distribution data can be decomposed into more specific microseismic force information, so as to more accurately analyze and process the impact of microseismic on operation. This processing helps to identify and optimize the impact of small vibrations on posture during operation, reduce potential risks caused by microseismic, and further improve the accuracy and stability of operation. Perform displacement residual analysis on the microseismic force decoupling data of posture action. Displacement residual analysis focuses on the difference between the actual displacement and the expected displacement caused by microseismic or other factors during operation. By analyzing these displacement residuals, deviations or errors in the operation can be identified, so as to understand the gap between the actual effect of the operation posture and the ideal state. Displacement residual data can reveal displacement errors caused by microseismic force, posture changes and other factors, which can lead to reduced operation accuracy or equipment damage. Through detailed analysis of displacement residuals, the operation posture or process can be adjusted in a targeted manner, problems caused by displacement errors can be reduced, the accuracy and stability of operation can be improved, and the reliability of the overall operation can be improved. The risk evolution of action posture is evaluated based on the microseismic force decoupling data of posture action and the displacement residual data of microseismic force. The purpose of this process is to evaluate the risk evolution trend of different operating postures over time. By taking into account the microseismic force and displacement residual data, the potential risks of various postures in long-term use can be predicted, such as equipment wear, personnel health problems or operating errors. Risk evolution assessment helps identify which operating postures will lead to greater risks in the future and provide a basis for taking preventive measures in advance. This assessment helps to achieve forward-looking risk management, by adjusting operating postures and optimizing operating processes to reduce risks in long-term use, thereby improving the safety and effectiveness of operations and ensuring that the operating process remains stable and reliable in the long term.

[0014] Preferably, step S3 comprises the following steps: Step S31: reconstructing the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain the action posture safety boundary data; Step S32: performing operation space hidden danger association analysis according to the action posture safety boundary data to obtain operation space hidden danger association data; Step S33: pre-identify hidden danger triggering risk points on the hidden danger associated data of the operation space according to the action posture safety boundary data to obtain hidden danger triggering risk point pre-data.

[0015] The present invention reconstructs the safety boundary of the action posture based on the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data. The main purpose of this step is to redefine the safety boundary of the operation posture based on the existing risk assessment data. This means that on the basis of the original safety standards and operating specifications, the safety range of the operation posture is adjusted and optimized according to the specific abnormal operation characteristics and risk evolution trends. By reconstructing the safety boundary, it is possible to more accurately define which operation postures are within the safety range and which are beyond the safety limit, thereby effectively preventing potential operation risks. This reconstruction can dynamically adapt to changes in the operating environment, ensure that the safety boundary always meets the latest risk assessment results, thereby improving the safety and effectiveness of the operation and reducing safety accidents caused by improper posture. Based on the reconstructed action posture safety boundary data, an operation space hidden danger association analysis is performed. This step aims to identify potential hidden dangers and risk points by analyzing the relationship between each hidden danger in the operation space and the redefined safety boundary. The operation space hidden danger association analysis comprehensively considers various factors in the operation space (such as equipment, environment and personnel operation) to reveal how these factors interact with the safety boundary to produce hidden dangers. The analysis results can help determine the distribution of hidden dangers in the operation space and their relationship with the operation posture, providing a basis for further risk management and safety improvement. This kind of correlation analysis helps to systematically identify and deal with hidden dangers in the operation space, ensure that potential risks can be effectively avoided and reduced during operation, and improve the safety of the overall operation environment. Pre-identify the risk points that trigger hidden dangers based on the safety boundary data of the action posture for the operation space hidden danger association data. The core of this step is to use the safety boundary and hidden danger association data to predict and identify specific risk points that will cause safety problems in the future. Through pre-identification, potential hidden dangers can be discovered and resolved in advance before problems occur during the actual operation process, avoiding these hidden dangers from causing actual risks or accidents during operation. This pre-identification not only helps to take preventive measures in a timely manner, but also optimizes the operation process and safety management strategy to reduce the potential accident rate. Its beneficial effect is to improve the predictability and prevention capabilities of operations, ensure that the safety of the operating environment is always effectively controlled, thereby protecting the safety of personnel and equipment and maintaining the stability and reliability of the operation process.

[0016] Preferably, step S31 includes the following steps: Step S311: performing operation posture microstress path analysis on abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain posture microstress path data; Step S312: performing gravity center offset calculation on posture microstress path data to obtain microstress gravity center offset data; Step S313: performing time-series posture fatigue evaluation according to the posture micro-stress path data and the micro-stress center of gravity offset data to obtain time-series posture fatigue evaluation data; Step S314: reconstruct the action posture safety boundary according to the posture microstress path data, the microstress center of gravity offset data and the time series posture fatigue assessment data to obtain the action posture safety boundary data.

[0017] The present invention obtains posture microstress path data by performing microstress path analysis on abnormal operation action characteristics based on operation posture risk evolution assessment data. The purpose of this step is to analyze in detail the transmission path of microstress under different operation postures. This analysis can reveal how stress propagates in different parts of the posture during operation, and identify the action path of microstress and its impact on the operation object or personnel. Posture microstress path data provides a deep understanding of microstress distribution, which helps to identify potential stress concentration areas and abnormal posture patterns. This analysis provides a basis for further risk assessment and optimization, so that the operation posture can be adjusted in a targeted manner, thereby reducing the negative impact of microstress on operation and improving the safety and effectiveness of operation. The center of gravity offset calculation is performed on the posture microstress path data to obtain microstress center of gravity offset data. The core of this step is to determine the offset of the center of gravity of the microstress distribution relative to the operation posture. The center of gravity offset calculation helps to identify whether there is an imbalance in the distribution of microstress during operation, thereby revealing the problem of uneven stress or offset. By calculating the offset of the center of gravity of microstress, the degree of stress concentration or dispersion and its potential impact on operational safety can be accurately understood. This data is crucial for optimizing posture design, helping to balance stress distribution during operation and reduce the problem of excessive local stress, thereby improving the comfort and safety of operation, and effectively preventing equipment damage or personnel fatigue caused by stress offset. The time-series posture fatigue assessment is performed based on the posture microstress path data and microstress center of gravity offset data to obtain the time-series posture fatigue assessment data. The purpose of this step is to evaluate how the fatigue caused by the operating posture changes over time in different time periods. By comprehensively analyzing the microstress path and center of gravity offset data, the fatigue accumulation of the posture over time is evaluated. This assessment can reveal the cumulative fatigue effect of the posture on personnel during long-term operation, thereby identifying operating postures or patterns that are prone to fatigue. Time-series posture fatigue assessment helps to optimize the operating process, design a more comfortable operating posture that does not cause excessive fatigue, thereby improving the work efficiency and safety of the operator, and reducing operating errors or health problems caused by fatigue. By integrating the posture microstress path data, microstress center of gravity offset data, and time-series posture fatigue assessment data, the action posture safety boundary is reconstructed to obtain new action posture safety boundary data. The core of this step is to adjust and optimize the safety boundary of the operating posture based on the latest micro-stress and fatigue data. By reconstructing the safety boundary, it can be ensured that the stress distribution and fatigue level are within the safe range during operation, thereby effectively preventing the risks caused by postures that exceed the safety limit. This reconstruction ensures that the operating posture is always within the verified safety range, reduces potential operating risks, and improves the safety and stability of the overall operation.Ultimately, this optimization ensures that posture during operation complies with the latest safety standards, further improving operator safety and a more comfortable working environment.

[0018] Preferably, step S33 includes the following steps: Step S331: analyzing the hidden danger triggering condition of the operation space hidden danger associated data according to the action posture safety boundary data to obtain hidden danger triggering condition data; Step S332: reconstructing the risk transmission chain of the hidden danger associated data in the operation space based on the hidden danger triggering condition data to obtain hidden danger risk transmission chain data; Step S333: performing pre-trigger behavior detection based on the hidden danger risk transmission chain data to obtain pre-trigger behavior detection data; Step S334: Pre-identify hidden danger triggering risk points based on the hidden danger risk transmission chain data and pre-trigger behavior detection data to obtain hidden danger triggering risk point pre-data.

[0019] The present invention performs hidden danger triggering condition analysis on the hidden danger associated data of the operation space according to the action posture safety boundary data to obtain hidden danger triggering condition data. The core of this step is to start from the safety boundary data and deeply analyze which conditions or situations trigger hidden dangers in the operation space. By performing a detailed analysis of the hidden danger triggering conditions, those key factors or behavior patterns can be identified, and the potential risks caused under different operation postures can be predicted. This analysis helps to understand the causes and triggering mechanisms of hidden dangers, so that more effective preventive measures can be formulated. The data of hidden danger triggering conditions can provide clear guidance for subsequent risk management and control, ensure that the safety in the operation environment is fully guaranteed, thereby reducing potential safety accidents and operational errors. Based on the hidden danger triggering condition data, the risk transmission chain of the hidden danger associated data of the operation space is reconstructed to obtain hidden danger risk transmission chain data. The purpose of this step is to reconstruct the chain of how risks propagate in the operation space, from the triggering conditions to the final risk performance. By analyzing the propagation path of the hidden danger, it is possible to understand how the risk is gradually transmitted from the initial trigger point to affect other factors or components in the operation environment. This chain reconstruction can reveal the mutual relationship between hidden dangers and their chain reactions, and help identify the potential diffusion path of risks. Through this analysis, we can strengthen protective measures in a targeted manner, slow down or prevent the spread of risks, improve the effectiveness and foresight of overall safety management, and thus more comprehensively prevent safety hazards in the operating space. Pre-trigger behavior detection is performed based on the hidden danger risk transmission chain data to obtain pre-trigger behavior detection data. The main purpose of this step is to detect early warning signals in advance by monitoring and analyzing operating behaviors before the hidden danger is actually triggered. By comparing with the risk transmission chain data, those operating behaviors or patterns with potential risks can be identified, and early signs of hidden dangers can be discovered and dealt with in a timely manner. Pre-trigger behavior detection can effectively identify potential behavioral problems before hidden dangers occur, so that appropriate preventive measures can be taken before the problem actually occurs. This early warning mechanism helps to reduce the probability of accidents, improve the safety of operations, and ensure that actions can be taken before hidden dangers are actually triggered, thereby maintaining the safety and stability of the operating environment. Pre-identification of hidden danger triggering risk points is performed based on the hidden danger risk transmission chain data and pre-trigger behavior detection data to obtain hidden danger triggering risk point pre-data. The core of this step is to synthesize the previously obtained data and identify potential risk points through analysis of the hidden danger risk transmission chain and pre-trigger behavior. Pre-identification helps to locate the key points that become the actual risk source before the hidden danger is actually triggered. Through this identification, targeted preventive measures and improvement strategies can be formulated and implemented to reduce the possibility of hidden dangers actually occurring. This pre-identification can not only effectively manage and control potential risks in operations, but also improve the overall safety management level, ensuring that potential hidden dangers in the operation process can be discovered and handled in a timely manner, thereby improving the safety and stability of the operating environment.

[0020] Preferably, step S4 comprises the following steps: Step S41: performing potential risk logic learning on the pre-data of the hidden danger triggering risk point based on a feedforward neural network algorithm to obtain potential risk logic learning data; Step S42: Designing an automated supervision process firmware according to the potential risk logic learning data to obtain a potential risk logic supervision firmware; Step S43: Embed the potential risk logic supervision firmware into the data cloud platform to perform security process supervision.

[0021] The present invention obtains potential risk logic learning data by performing potential risk logic learning on the pre-data of hidden danger triggering risk points based on a feedforward neural network algorithm. The main purpose of this step is to use the feedforward neural network to perform pattern recognition and logic learning on historical hidden danger data and risk points. The feedforward neural network can learn the potential logical relationship and pattern of hidden danger triggering by processing a large amount of complex data. This learning process can extract the key features and potential laws of hidden danger triggering, thereby establishing a logical model of potential risks. Through this model, the risk triggering mechanism under different conditions can be more deeply understood and predicted, providing a theoretical basis for formulating more effective safety management measures. Potential risk logic learning data not only helps to improve the accuracy of risk identification and early warning, but also provides a scientific basis for the subsequent automated supervision process firmware design and improves the overall risk management level. According to the potential risk logic learning data, the automated supervision process firmware is designed to obtain the potential risk logic supervision firmware. The core of this step is to convert the potential risk logic learning data obtained above into an operational supervision process firmware. These firmwares will contain automated safety supervision processes and rules, and can monitor and manage potential risks in real time in practical applications. By designing automated supervision firmware, we can ensure that the supervision process is efficiently and consistently applied to actual operations, reduce the need for manual intervention, and improve the accuracy and efficiency of supervision. This firmware design can not only reduce human errors, but also achieve comprehensive monitoring of complex operating environments. Through automated supervision, we can respond quickly before potential risks occur, take timely measures, and prevent hidden dangers from developing into actual problems, thereby improving the intelligence level of safety management. Embed the potential risk logic supervision firmware into the data cloud platform to perform safety process supervision. The key to this step is to integrate the designed automated supervision firmware into the actual supervision system to ensure that it can operate effectively in the actual operating environment. By embedding the firmware into the data cloud platform, real-time supervision and automated control of laboratory safety processes can be achieved. The embedded firmware will be responsible for monitoring various operating and environmental conditions in the laboratory, automatically detecting and responding to potential risks. This integration not only improves the automation level of supervision, but also ensures that laboratory operations meet safety standards. Through real-time data analysis and risk assessment, the platform can issue alarms in time and take corresponding measures to effectively protect the safety of laboratory operations. Overall, this process improves the intelligence and automation level of safety processes, making safety management more scientific, accurate and efficient.

[0022] Preferably, the present invention further provides a medical safety process supervision system based on data analysis, which is used to execute the medical safety process supervision method based on data analysis as described above, and the medical safety process supervision system based on data analysis includes: The process action decomposition module is used to monitor the whole process experiment operation process through electronic monitoring equipment to obtain the whole process experiment operation process monitoring video; based on the whole process experiment operation process monitoring video, the operation process action is decomposed to obtain the operation process action decomposition data; The operation risk evolution module is used to extract action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; to refine the abnormal operation action characteristics of the operation process action behavior parameters to obtain abnormal operation action characteristic refinement data; to evaluate the action posture risk evolution on the abnormal operation action characteristic refinement data to obtain the operation posture risk evolution evaluation data; The risk pre-analysis module is used to reconstruct the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain the action posture safety boundary data; pre-identify the hidden danger triggering risk point according to the action posture safety boundary data to obtain the hidden danger triggering risk point pre-data; The automated supervision module is used to design automated supervision process firmware based on the pre-data of risk points that trigger hidden dangers, and obtain potential risk logic supervision firmware; the potential risk logic supervision firmware is embedded in the data cloud platform to perform security process supervision.

[0023] The beneficial effect of the present invention is that the whole process of experimental operation process monitoring is carried out by electronic monitoring equipment, and every detail of the whole process of the experiment can be systematically captured. The core of this stage is to obtain a comprehensive experimental process video to ensure that all operation steps, behaviors and environmental factors are recorded. The video data provides a detailed basis for subsequent analysis, allowing a comprehensive review of the experimental process. Based on these monitoring videos, the operation process action decomposition is further performed. This process includes breaking down complex operation steps into specific actions so that each link can be analyzed more carefully. This decomposition not only helps to identify the specific details of each step of the operation, but also reveals potential operation modes and behavior laws, laying a solid foundation for subsequent parameter extraction and behavior analysis. The ultimate goal of this stage is to form structured operation process data so that subsequent analysis can be more accurate and efficient. After completing the operation process action decomposition, the next step is to extract action behavior parameters from these data. This process involves extracting key behavior parameters from the operation process, such as action speed, frequency, angle, etc. These parameters help to quantify the characteristics of each action, making the standardization and comparison of behavior a reality. The extracted parameters will then be used for the refinement of abnormal operation action features. By conducting an in-depth analysis of abnormal operations, we can identify features that are significantly different from normal behavior patterns, which indicate potential risks or operational errors. Next, we conduct an action posture risk evolution assessment on these refined data, which is to predict and evaluate the risk evolution of different operating postures in long-term use. This assessment can reveal safety hazards in operation and provide a scientific basis for improving operating specifications and optimizing the operating environment, thereby ultimately improving the safety and efficiency of the experiment. After obtaining the operation posture risk evolution assessment data, the next step is to reconstruct the action posture safety boundary based on the abnormal operation action feature refinement data. The core of this process is to determine the safety range and boundaries of each operation posture. These safety boundaries are defined based on the historical data of the action and the risk assessment results, ensuring that operations within these boundaries are relatively safe. Through precise boundary setting, the safety risks caused by improper posture during operation can be significantly reduced. Subsequently, the risk point of hidden danger triggering is pre-identified based on the constructed action posture safety boundary data. The purpose of this stage is to identify potential points that trigger risks during operation. Through the analysis of the safety boundary, those operating behaviors or environmental factors that may cause safety hazards can be discovered in advance. Pre-identification of hidden danger triggering risk points enables appropriate measures to be taken before problems occur, thereby greatly improving the safety and reliability of laboratory operations. This preventive risk management strategy not only helps protect the safety of laboratory personnel, but also reduces the probability of equipment damage and experimental failure. Based on the pre-identification data of hidden danger triggering risk points, the next step is to design an automated supervision process firmware. The focus of this stage is to develop a logical supervision firmware that can automatically monitor and manage laboratory operation risks.This firmware will design an automated supervision logic for potential risk points based on the pre-identified data, including functions such as real-time monitoring, alarm triggering and automatic adjustment. This process enables the laboratory to respond in a timely manner when abnormal operations occur, thereby reducing the need for manual intervention and improving operational safety and efficiency. Finally, the potential risk logic supervision firmware is embedded in the data cloud platform. This step involves integrating the designed firmware into the existing laboratory management system so that it can perform safety process supervision in real time. Through this integration, the laboratory can achieve comprehensive safety monitoring, automatically identify and respond to various potential risks, and ensure the standardization and safety of the experimental process. The embedded system can not only provide real-time feedback on risk information in the operation, but also provide targeted safety suggestions and measures, thereby further optimizing the laboratory's operating environment and safety management level. Therefore, the present invention is an optimization of a traditional medical safety process supervision method based on data analysis, which solves the problem that a traditional medical safety process supervision method based on data analysis has low accuracy in identifying misoperations and slow warning speed, improves the accuracy of misoperation identification, and improves the speed of warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A schematic diagram of the steps of a medical safety process supervision method based on data analysis; Figure 2 for Figure 1 Detailed implementation steps of step S2 in FIG.

[0025] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0026] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0027] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0028] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0029] To achieve this, please refer to Figure 1 to Figure 2 , a medical safety process supervision method based on data analysis, the method comprising the following steps: Step S1: monitor the whole process of the experiment operation process through electronic monitoring equipment to obtain the whole process of the experiment operation process monitoring video; decompose the operation process action based on the whole process of the experiment operation process monitoring video to obtain the operation process action decomposition data; Step S2: extracting action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; performing abnormal operation action feature refinement processing on the operation process action behavior parameters to obtain abnormal operation action feature refinement data; performing action posture risk evolution assessment on the abnormal operation action feature refinement data to obtain operation posture risk evolution assessment data; Step S3: reconstructing the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain action posture safety boundary data; performing pre-identification of hidden danger triggering risk points based on the action posture safety boundary data to obtain hidden danger triggering risk point pre-data; Step S4: Design automated supervision process firmware based on the pre-data of hidden danger triggering risk points to obtain potential risk logic supervision firmware; embed the potential risk logic supervision firmware into the data cloud platform to perform security process supervision.

[0030] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a medical safety process supervision method based on data analysis of the present invention. In this example, the medical safety process supervision method based on data analysis includes the following steps: Step S1: monitor the whole process of the experiment operation process through electronic monitoring equipment to obtain the whole process of the experiment operation process monitoring video; decompose the operation process action based on the whole process of the experiment operation process monitoring video to obtain the operation process action decomposition data; In an embodiment of the present invention, in a laboratory or medical studio, high-precision electronic monitoring equipment, such as a 360-degree high-definition camera and an infrared sensor without blind spots, is arranged to record the experimental operation process in real time throughout the entire process. The position and angle of the monitoring equipment are precisely adjusted to ensure that every detail of the experimental operation is captured without missing any link. Every action in the experimental operation process, including hand operation, instrument use, chemical contact, etc., is recorded in real time and stored as a full-process experimental operation process monitoring video. During the monitoring video acquisition process, automatic tracking technology is applied to ensure the clarity and continuity of the target. The video data is stored in a high-capacity hard disk and redundant backup is performed in real time to avoid data loss. Then, the monitoring video is analyzed and processed using video action decomposition technology to decompose the experimental personnel's operation process into several independent action sequences. Each decomposed action is feature extracted by the inter-frame changes in the video, and the continuous action is divided into different stages by dynamic time warping (DTW) and other methods to form a time-series operation process action decomposition data. During the decomposition process, the start and end time, action amplitude, operating instrument contact point and other parameters of each action are finely marked to ensure the integrity and accuracy of the data. These data will serve as the basis for subsequent behavioral parameter extraction and provide accurate data support for security monitoring.

[0031] Step S2: extracting action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; performing abnormal operation action feature refinement processing on the operation process action behavior parameters to obtain abnormal operation action feature refinement data; performing action posture risk evolution assessment on the abnormal operation action feature refinement data to obtain operation posture risk evolution assessment data; In an embodiment of the present invention, the acquired operation process action decomposition data is used to extract the behavior parameters of each action using a high-precision behavior analysis tool, such as a three-dimensional gesture capture system (e.g., an optical tracking or inertial sensor system). These behavior parameters include, but are not limited to, the swing angle of the arm, the operation speed, the position of the hand, the duration of the action, etc. During the extraction process, the fineness of the action data is ensured, and accurate operation process action behavior parameters are generated through frame-by-frame analysis and time series comparison. These parameters are the basis for subsequent analysis and directly affect the identification and evaluation of abnormal behavior. After obtaining the operation process action behavior parameters, a multi-layer feature extraction algorithm is used to refine them, especially for abnormal operation action features. Feature refinement includes the gradual amplification of small action changes, the quantification of abnormal amplitude, and the analysis of the instability of the action. The refinement process of abnormal operation features is to establish a standard behavior model for each action parameter, amplify and summarize the features that deviate from the standard, and form a refined feature set. The refined data not only focuses on the amplitude and speed of the action, but also focuses on analyzing the impact of subtle deviations and repetition frequencies of the action on experimental safety, and obtains the refined data of abnormal operation action features. These refined data are highly interpretable and lay the foundation for subsequent risk assessment. Subsequently, the risk evolution of the action posture is evaluated for the refined data of abnormal operation action characteristics. During the evaluation, the refined abnormal feature data is compared and analyzed with the postures in the standard posture library, and the deviation risk value of the action posture is calculated using a multi-parameter risk matrix. The risk assessment is not only aimed at the danger of a single action, but also evaluates the cumulative risk of multiple action combinations through time series analysis. Through cross-analysis, high-risk postures are identified, and the risk evolution trend diagram is presented in a visual way to generate the operation posture risk evolution assessment data. The evaluation results are used to prompt the experimenter in real time and optimize the operation process to reduce the safety risks in laboratory operations.

[0032] Step S3: reconstructing the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain action posture safety boundary data; performing pre-identification of hidden danger triggering risk points based on the action posture safety boundary data to obtain hidden danger triggering risk point pre-data; In the embodiment of the present invention, according to the risk evolution assessment data of the operation posture, the safety boundary of the abnormal operation action feature refinement data is reconstructed by constructing a multi-dimensional safety parameter space of the action posture. The reconstruction process adopts constraints based on multiple parameters such as action amplitude, speed, and posture stability, and applies a constraint optimization algorithm to redefine the safety boundary of the operation action. Each safety boundary is limited by a specific numerical range, such as physical parameters such as angle, force, and action duration, to ensure that each posture is performed within a safe range. When reconstructing the boundary, the threshold of the safety boundary is dynamically adjusted in combination with the time series change analysis, especially in the process of posture evolution, the change of the safety threshold is calibrated in real time to obtain accurate action posture safety boundary data. These data will provide accurate action specifications for subsequent hidden danger identification and ensure the safety of experimental operations. Based on the action posture safety boundary data, the operation data is analyzed by the risk identification tool to identify potential hidden danger triggering risk points. In the pre-identification process, the safety threshold in the boundary data is compared point by point with the real-time posture data to capture each operation posture close to the safety boundary. A multi-point trigger judgment mechanism is adopted, combined with the posture evolution trend analysis, to predict the risk points of future triggering, and generate hidden danger triggering risk point pre-data. The data lists in detail the specific time point, posture type, trigger conditions and other information of each high-risk operation, and uses visual heat maps to illustrate risk-intensive areas, allowing experiment managers to clearly understand the location of potential safety hazards and provide a basic basis for subsequent process optimization.

[0033] Step S4: Design automated supervision process firmware based on the pre-data of hidden danger triggering risk points to obtain potential risk logic supervision firmware; embed the potential risk logic supervision firmware into the data cloud platform to perform security process supervision.

[0034] In an embodiment of the present invention, based on the pre-data of the risk point triggered by the hidden danger, the firmware design of the automated supervision process is carried out. The firmware design adopts an embedded system development tool, and the real-time monitoring and intervention of high-risk operations are realized through logic arrangement. In the firmware design process, according to the risk point position and triggering conditions in the pre-data, the corresponding logic rules are set, such as triggering alarm, operation suspension, operation guide prompts and other functions. In the design process, the hardware description language (such as VHDL or Verilog) is used to accurately encode each logic module of the firmware, and multiple simulation tests are carried out to ensure that risk actions can be accurately identified and responded to under various operating conditions. The core function of the firmware is to combine hidden danger identification with real-time feedback of the experimental operation process to realize intelligent and automated risk supervision, and finally obtain the potential risk logic supervision firmware. The completed potential risk logic supervision firmware is embedded into the data cloud platform. The embedding process loads the firmware into the existing control hardware of the laboratory through a dedicated interface circuit. After embedding, the firmware and the laboratory monitoring equipment are connected to the data to realize real-time analysis and feedback of the experimental operation. The firmware continuously monitors the experimental operation during the operation of the platform, intervenes and prompts the risk operation according to the design logic, and ensures that the laboratory operation meets the safety process standards. The embedding and operation of firmware will greatly improve the efficiency of safety supervision, significantly reduce the risk of safety accidents caused by operational errors, and realize full-process safety process supervision.

[0035] Preferably, step S1 comprises the following steps: Step S11: monitor the whole process of experimental operation process through electronic monitoring equipment to obtain a monitoring video of the whole process of experimental operation process; Step S12: performing contrast optimization processing on the whole-process experimental operation process monitoring video to obtain an operation process monitoring contrast-optimized video; Step S13: switching and integrating the operation process monitoring contrast-optimized video from multiple angles to obtain a multi-angle integrated video of the operation process monitoring; Step S14: Decompose the operation process actions based on the multi-angle integrated video of the operation process monitoring to obtain the operation process action decomposition data.

[0036] In an embodiment of the present invention, high-resolution electronic monitoring equipment, such as industrial-grade high-definition cameras and multispectral sensor systems, are deployed in laboratories, operating rooms, or medical studios to monitor the entire experimental operation process in real time. The monitoring equipment is precisely calibrated to ensure that all operating areas of the laboratory are covered, including key points such as operating tables, reagent cabinets, and fume hoods. The camera is equipped with an automatic tracking function, which can automatically adjust the viewing angle and focal length according to the movements of the experimenter to obtain the clearest image. During the monitoring process, all experimental operation process videos are recorded at high frame rates and high resolutions and saved as full-process experimental operation process monitoring videos. Video data is transmitted to a dedicated storage server in real time via a wired network for multi-level redundant backup to ensure data security and reliability and avoid data loss due to equipment failure or accidental power failure. The acquired full-process experimental operation process monitoring video is subjected to contrast optimization processing to enhance the clarity of details in the video. This processing uses high dynamic range (HDR) technology and local contrast enhancement algorithms to adjust each frame in the video pixel by pixel, and enhances the sense of hierarchy of dark details and highlight areas, making key actions and objects in the experimental operation easier to identify. During the processing, a block processing method is used to optimize different areas independently to avoid image distortion caused by global adjustment. The contrast-optimized video can significantly distinguish the operation object and the background in terms of visual effect, improving the visual monitoring effect of the operation process. After the optimization process is completed, the video is re-encoded into a high-contrast version and output as a contrast-optimized video for operation process monitoring. For the contrast-optimized video for operation process monitoring, multi-angle switching and integration processing are performed using multi-camera synchronization technology and image stitching algorithm. The process first synchronizes the video streams at different angles in time, and achieves seamless connection of video switching through feature point matching and angle calibration of video frames. Then, the videos at different angles are integrated into a unified video stream to form a multi-angle integrated video for operation process monitoring with multi-view coverage. During the integration process, the real-time switching controller is used to select the best view for display to ensure that each experimental action is presented in the clearest way. The integrated video not only includes the main operation view, but also can switch to the auxiliary view at key steps, providing more comprehensive reference data for subsequent action decomposition and risk assessment. Based on the multi-angle integrated video of operation process monitoring, the action segmentation algorithm is used to decompose the experimental operation process frame by frame. During the action decomposition process, the start and end time of each operation action are extracted by detecting the dynamic changes of the human skeleton in the video. The decomposition algorithm combines time sequence detection and action trajectory analysis to automatically identify the specific operations of the experimenter, such as sampling, stirring, pouring and other detailed actions. Each action segment is marked as an independent operation unit and parameterized to record detailed information such as action type, duration, and action amplitude.The accuracy of the decomposition data directly affects the subsequent risk assessment. Therefore, each decomposition result is checked and compared multiple times to ensure the integrity and accuracy of the operation action decomposition data, providing precise data support for further refinement of laboratory safety supervision.

[0037] Preferably, step S2 comprises the following steps: Step S21: Obtaining standardized operation process data; Step S22: extracting action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; Step S23: performing abnormal operation action identification on the operation process action behavior parameters according to the standardized operation process data to obtain abnormal operation action data; Step S24: performing abnormal operation action feature refinement processing on the operation process action behavior parameters according to the abnormal operation action data to obtain abnormal operation action feature refinement data; Step S25: Perform action posture risk evolution assessment on the abnormal operation action feature refinement data to obtain operation posture risk evolution assessment data.

[0038] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes: Step S21: Obtaining standardized operation process data; In an embodiment of the present invention, the process of obtaining standardized operation flow data first extracts key action parameters from the standard operating procedure (SOP), including the order of steps, the precise duration, speed, position of each action, and the use of operating tools. The data is derived from the systematic record of common experimental operations in the laboratory, and a standardized operation flow database is constructed in combination with actual operation specifications and the training content of the experimenters. By analyzing the common parameters in different operation processes, the detailed actions in the experimental operation are formatted and stored to form a standard behavior template for the operation. In order to ensure the accuracy of the data, image contrast technology and secondary review by professionals are applied in the data collation process to ensure the accuracy of the standardized operation flow data, so as to provide a reliable basis for the subsequent abnormal operation identification.

[0039] Step S22: extracting action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; In an embodiment of the present invention, when extracting action behavior parameters from the operation process action decomposition data, an action recognition technology based on spatiotemporal analysis is used to analyze the human body motion trajectory and posture changes in the video frame by frame. Specific extraction parameters include detailed information such as movement amplitude, speed, joint angle, posture stability, relative positions between various parts of the body, etc. The extraction process dynamically tracks the key nodes of the experimenter during the operation process through joint point positioning technology to obtain detailed behavioral characteristics of the operation action. The linear segmentation method of the motion trajectory is used to divide the continuous action into multiple time segments, each of which corresponds to an independent operation action. The extracted parameter data is subjected to statistical calculation and filtering processing to eliminate noise and unreasonable data, ensure the integrity and accuracy of the obtained operation process action behavior parameters, and provide specific data support for the next step of abnormal action recognition.

[0040] Step S23: performing abnormal operation action identification on the operation process action behavior parameters according to the standardized operation process data to obtain abnormal operation action data; In an embodiment of the present invention, based on the standardized operation process data, the extracted operation process action behavior parameters are compared and analyzed one by one to identify abnormal operation actions that do not conform to the standard specifications. During the comparison process, the difference analysis method is used to compare the differences between the actual operation parameters and the standardized parameters one by one, such as excessive movement amplitude, abnormally fast or slow speed, unstable posture, etc. For each action that does not meet the standards, it is marked as an abnormal operation action, and the specific characteristics of the abnormality are recorded in detail, including the type of abnormality, degree of deviation and duration. The identified abnormal operation action data further annotates the time point when the abnormality occurred, the type of action and its corresponding parameter deviation value. This data directly reflects the deviant behavior in the experimental operation, laying the foundation for subsequent feature refinement processing.

[0041] Step S24: performing abnormal operation action feature refinement processing on the operation process action behavior parameters according to the abnormal operation action data to obtain abnormal operation action feature refinement data; In an embodiment of the present invention, the operation process action behavior parameters are analyzed in detail based on the abnormal operation action data, and the feature refinement data of the abnormal action is further extracted. During the processing, a fine-grained action decomposition technology is used to conduct an in-depth analysis of each abnormal operation action, and the overall action is refined into a variety of features at the micro level, such as subtle posture adjustments, speed changes, force deviations, etc. High-precision motion tracking tools are used to obtain the specific features of abnormal actions through frame-by-frame analysis and local feature extraction, such as excessive stretching of local joints, abnormal rotation trajectories, etc. The feature refinement data of each abnormal action contains a detailed numerical description and a corresponding action trajectory diagram, forming a comprehensive characterization of the abnormal action, providing key detail information for subsequent risk assessment.

[0042] Step S25: Perform action posture risk evolution assessment on the abnormal operation action feature refinement data to obtain operation posture risk evolution assessment data.

[0043] In an embodiment of the present invention, the action posture risk evolution assessment is performed on the refined data of abnormal operation action characteristics, and the characteristics of the abnormal action are compared with the existing safety specifications by constructing a risk assessment matrix. During the assessment process, the impact of abnormal characteristics on operational safety is analyzed item by item, such as the risk of tool falling off due to unstable posture, the risk of reagent overflow caused by speed deviation, etc. A hierarchical risk assessment method is adopted to calculate the comprehensive risk score of each action posture according to the severity and frequency of occurrence of abnormal characteristics. The risk assessment tool automatically generates a risk evolution curve, which shows the risk accumulation effect and potential evolution trend of different posture characteristics, and outputs it as operation posture risk evolution assessment data. The assessment results can provide a scientific basis for the improvement of laboratory safety operation specifications, effectively warn of high-risk actions, and reduce safety hazards during the experiment.

[0044] Preferably, step S24 comprises the following steps: Step S241: calculating the abnormal operation action dynamic deviation of the operation process action behavior parameters according to the abnormal operation action data to obtain the abnormal operation action dynamic deviation data; Step S242: performing a multi-dimensional parameter heterogeneous comparison on the operation process action behavior parameters based on the abnormal operation action dynamic deviation data to obtain multi-dimensional parameter heterogeneous data; Step S243: performing microscopic motion scale analysis according to the abnormal operation motion dynamic deviation data and the multi-dimensional parameter heterogeneous data to obtain multi-dimensional microscopic motion scale data; Step S244: performing a subtle motion feature deep analysis on the multi-dimensional micro-motion scale data to obtain multi-dimensional micro-motion scale deep analysis data; Step S245: performing abnormal operation action feature refinement processing according to the multi-dimensional micro-action scale depth analysis data to obtain abnormal operation action feature refinement data.

[0045] In the embodiment of the present invention, according to the abnormal operation action data, the abnormal operation action dynamic deviation calculation is performed on the operation process action behavior parameters. First, it is necessary to analyze the time series data of each abnormal operation action frame by frame, and obtain its dynamic deviation value by extracting the trajectory and dynamic changes of the key motion nodes. In the dynamic deviation calculation process, a high-precision motion trajectory capture tool is used to calculate the differences between the abnormal action and the standard operation action in multiple dimensions such as time, position, speed, and angle. For example, during the execution of a certain abnormal action, the motion trajectory of the wrist joint deviates from the normal range. The deviation is quantified into a specific value and classified as the abnormal operation action dynamic deviation data. Each deviation data is detailed with the time point of the abnormal occurrence, the deviation amplitude and the related motion trajectory to accurately reflect the deviation degree of the abnormal operation in the dynamic dimension, and provide basic data support for subsequent multidimensional parameter analysis. Based on the abnormal operation action dynamic deviation data, the operation process action behavior parameters are subjected to multidimensional parameter heterogeneous comparison, and by constructing a multidimensional parameter heterogeneous matrix, various parameters are refined according to multiple dimensions such as time, space, force, and speed. The comparison process uses three-dimensional motion reconstruction technology to reproduce the actual action as a spatial stereo model, and compares the dynamic deviation value of each action with the multi-dimensional parameter data one-to-one to ensure the comprehensiveness and accuracy of the analysis. In particular, during the comparison process, the heterogeneous comparison algorithm is used to cross-validate the parameters between different dimensions. For example, the real deviation trajectory of abnormal actions is evaluated through the linkage analysis of spatial position and speed. The multi-dimensional parameter heterogeneous data finally formed contains detailed parameter comparison results, marking the abnormal correlation between the parameters of each dimension, laying the foundation for micro-scale analysis. According to the dynamic deviation data of abnormal operation actions and multi-dimensional parameter heterogeneous data, micro-motion scale analysis is carried out. Specifically, the macro-motion is decomposed into smaller-scale micro-motion units by subdividing the spatial position, force distribution, dynamic changes and other details of abnormal actions from multiple angles. Complex actions are divided into smaller scales through frame-by-frame refinement and decomposition, such as the movement changes of specific parts such as fingertips, wrists, and elbows. Each micro-motion unit is described by its independent dynamic parameters, such as slight deviations in the motion path, slight changes in the direction of force application, etc., forming multi-dimensional micro-motion scale data. This data can clearly present the deviations and features of abnormal operations at the micro level, providing an accurate data basis for subsequent in-depth analysis. The multi-dimensional micro-motion scale data is deeply analyzed for subtle motion features, and each detail of the micro-motion is further analyzed through a multi-scale decomposition algorithm to discover the hidden abnormal features. The analysis process uses multi-level motion feature extraction technology, combined with mechanical analysis and dynamic tracking, to gradually analyze the subtle features of abnormal movements at the micro scale, such as overstretching of small joints, unconscious micro-tremors, etc. The deep analysis tool forms a detailed motion feature spectrum by superimposing multi-dimensional information, accurately recording the detailed features of each micro-motion.The multi-dimensional micro-motion scale deep analysis data finally obtained contains the fine parameters, feature maps and detailed descriptions of differences from standard operations for each action, providing deep information support for the refinement of abnormal operation features. According to the multi-dimensional micro-motion scale deep analysis data, the abnormal operation action features are refined. By classifying and summarizing the analyzed micro-motion features and mapping them, the system identifies and extracts the key features of abnormal operations. The refinement process uses feature mapping tools to compare the parsed data with the standard database of abnormal operations, identify abnormal features one by one, and enhance their annotations, such as feature name, abnormal degree, related operation nodes, etc. Each refinement step includes precise positioning and parameterized description of features to ensure that the identification and annotation of abnormal action features have sufficient accuracy and depth. The final output of the abnormal operation action feature refinement data is a comprehensive feature characterization of abnormal actions in the operation process, providing a key basis for subsequent risk assessment and improvement of operation specifications.

[0046] Preferably, step S25 comprises the following steps: Step S251: performing posture stress distribution analysis on the abnormal operation action feature refinement data to obtain posture stress distribution data; Step S252: performing microseismic force decoupling processing on the abnormal operation action feature refinement data according to the posture stress distribution data to obtain posture action microseismic force decoupling data; Step S253: performing displacement residual analysis on the posture action microseismic force decoupling data to obtain microseismic force displacement residual data; Step S254: Perform risk evolution assessment of the action posture based on the microseismic force decoupling data of the posture action and the microseismic force displacement residual data to obtain the risk evolution assessment data of the operation posture.

[0047] In an embodiment of the present invention, the abnormal operation action feature refinement data is used to perform posture stress distribution analysis. In a specific implementation, firstly, a high-precision stress sensing device and a three-dimensional motion capture system are used to measure the posture stress distribution of the abnormal operation action. By gradually analyzing the stress conditions of different joints and limb parts during the action, the stress distribution map of each key action node is obtained. During the stress analysis process, a stress field reconstruction tool is used to visualize the captured force data to form a detailed posture stress distribution data map, indicating the magnitude and direction of the force of each part. For example, when an abnormal action is detected, if the waist is concentrated and the stress value is abnormally high, the posture stress distribution data will mark the waist as a high-risk stress area. These data can accurately display the stress distribution characteristics of each posture, laying the foundation for subsequent microseismic decoupling processing. According to the posture stress distribution data, the microseismic force decoupling processing is performed on the abnormal operation action feature refinement data, and the microvibration generated during the posture action is captured by a high-frequency microseismic sensing device, and the stress and microseismic signals are decoupled and separated in combination with the mechanical analysis tool. During the decoupling process, multi-channel filtering and vibration stratification technology are applied to decompose the complex microseismic signals into different frequency bands, gradually eliminate external interference factors, and extract the real force performance of microseismic. In this process, special attention is paid to the detailed changes of microseismic signals, and accurate microseismic force decoupling data of postures and movements is obtained through a refined force decoupling algorithm. This data records the force decoupling characteristics of each posture in detail, distinguishes the microseismic influence between different force sources, and provides accurate basic data for subsequent displacement residual analysis. Displacement residual analysis is performed on the microseismic force decoupling data of postures and movements, and a high-precision displacement measurement system is used to track the displacement of postures and movements after microseismic force frame by frame. During the analysis process, the displacement trajectory of each posture and movement is first quantified to obtain the residual value. By calculating the displacement difference between the actual movement and the standard movement under the microseismic state, the abnormal movement deviation point is identified. The displacement residual analysis tool can identify subtle displacement deviations at micron-level accuracy, especially slight dislocations or unexpected displacement changes at joints, which will be captured and quantified as residual data. The generated microseismic force-displacement residual data contains the complete displacement trajectory and residual details of each stage, providing a comprehensive analysis of displacement risk characteristics for action posture risk assessment. Based on the microseismic force decoupling data of posture movements and the microseismic force-displacement residual data, the risk evolution of action postures is evaluated. During the evaluation process, the decoupled microseismic force characteristics and displacement residuals are comprehensively analyzed using the mechanical risk assessment platform, and the safety of each action posture is dynamically evaluated through a multi-dimensional risk analysis algorithm. The evaluation system integrates and calculates multiple parameters such as the force characteristics, microseismic response, and displacement deviation of the posture to form a risk evolution trend chart to evaluate the risk evolution process of each posture under continuous operation.The evolutionary assessment results can clearly point out the high-risk links of posture movements and their potential safety hazards, such as the increased risk caused by stress accumulation and abnormal displacement caused by a certain movement under specific operating conditions. The final operating posture risk evolutionary assessment data provides a scientific basis for laboratory safety supervision and operation optimization.

[0048] Preferably, step S3 comprises the following steps: Step S31: reconstructing the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain the action posture safety boundary data; Step S32: performing operation space hidden danger association analysis according to the action posture safety boundary data to obtain operation space hidden danger association data; Step S33: pre-identify hidden danger triggering risk points on the hidden danger associated data of the operation space according to the action posture safety boundary data to obtain hidden danger triggering risk point pre-data.

[0049] In an embodiment of the present invention, based on the risk evolution assessment data of the operation posture, the abnormal operation action feature refinement data is used to reconstruct the action posture safety boundary. During the implementation process, the refinement features of the abnormal operation action are first mapped to the spatial coordinate system through a three-dimensional posture reconstruction tool. A stress analyzer and a force distribution measuring instrument are used to perform multi-angle stress and dynamic force analysis on the operation posture, and the safety boundary of the posture is gradually constructed by locating the high-risk stress area of ​​each posture. During the posture safety boundary reconstruction process, factors such as force limit, posture balance point, and inertial offset are included in the calculation to accurately adjust the tolerance range of the action posture. By continuously optimizing the parameters of the posture boundary, the threshold of the safety boundary is determined and finally verified. The obtained action posture safety boundary data clearly defines the safety area range of each posture, providing a reliable safety baseline for subsequent hidden danger analysis. The operation space hidden danger association analysis is performed based on the action posture safety boundary data. During implementation, the space hidden danger detection equipment is used to perform a full range of hidden danger scanning on the experimental operation area. By combining the posture safety boundary data, the physical environment of the operation space is refined and analyzed, focusing on identifying the hidden danger points that overlap or are close to the safety boundary. The operation space hidden danger association analysis tool compares boundary data with environmental data in multiple dimensions to generate a risk map of potential hidden dangers. For example, during the analysis process, if the edge area of ​​the test bench is found to overlap with the posture safety boundary, the area will be marked as a high-risk hidden danger association area. This analysis process is detailed to the triggering conditions and specific locations of each potential hidden danger, providing detailed data support for the accurate identification of high-risk areas, and finally generating operation space hidden danger association data. According to the action posture safety boundary data and the operation space hidden danger association data, the experimental operation area is pre-identified for hidden danger triggering risk points. During implementation, the risk point pre-identification device is used in combination with the real-time operation monitoring system to conduct a comprehensive detection of hidden danger trigger points during the operation process. In this process, the risk identification tool locates the potential trigger points of each hidden danger by comparing the safety boundary data and the hidden danger association data, and evaluates the risk level of each trigger point through the dynamic risk weight calculation method. The identified risk points will be marked and recorded. For example, in the experiment, the area where the hand action touches the dangerous equipment is identified and the trigger point is positioned as a high-risk area for priority supervision. The pre-identification process includes the time series analysis and multi-angle review of the trigger point to ensure the accuracy of risk point positioning. The final output of the hidden danger trigger risk point pre-data provides a substantial basis for early prediction of laboratory safety monitoring and hidden danger warning.

[0050] Preferably, step S31 includes the following steps: Step S311: performing operation posture microstress path analysis on abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain posture microstress path data; Step S312: performing gravity center offset calculation on posture microstress path data to obtain microstress gravity center offset data; Step S313: performing time-series posture fatigue evaluation according to the posture micro-stress path data and the micro-stress center of gravity offset data to obtain time-series posture fatigue evaluation data; Step S314: reconstruct the action posture safety boundary according to the posture microstress path data, the microstress center of gravity offset data and the time series posture fatigue assessment data to obtain the action posture safety boundary data.

[0051] In the embodiment of the present invention, based on the operation posture risk evolution assessment data, the abnormal operation action feature refinement data is analyzed for the operation posture micro-stress path. In the specific implementation, a high-precision stress sensing device is used to monitor the micro-stress changes of each joint and muscle group of the experimenter in real time during the operation. The stress sensing data is recorded by the stress path analyzer to record the distribution and dynamic change trend of micro-stress and generate a detailed stress path diagram. The analysis process includes the evaluation of the stress accumulation and dispersion of the starting, vertex and end points of each posture, gradually constructing the stress path model of each posture, and identifying the high-risk stress path in the posture in combination with the operation risk evolution data. The posture micro-stress path data finally obtained clarifies the conduction and distribution characteristics of the micro-stress of each posture in the operation, laying the foundation for the subsequent center of gravity offset and fatigue assessment. The center of gravity offset is calculated for the posture micro-stress path data. In the implementation process, the center of gravity calculation tool is first used to extract the key nodes in the posture micro-stress path data. Through moment analysis and dynamic balance measurement, the center of gravity position of each posture under different stress paths is accurately calculated. During the calculation process, the multi-axis center of gravity analysis method is used to fit and track the center of gravity offset trajectory of each posture, and the center of gravity movement amplitude and direction of the posture at different operation moments are clarified. The center of gravity changes in the high stress path area are evaluated in particular, and the risk points where the center of gravity deviates from the safe area are identified. For example, in high-frequency operation postures, if the center of gravity offset exceeds the set threshold, it is marked as a potential posture instability state. The output of microstress center of gravity offset data provides key center of gravity offset reference information for determining the posture safety boundary. According to the posture microstress path data and microstress center of gravity offset data, the time-series posture fatigue assessment is performed. During the implementation process, the fatigue assessment gradually collects and analyzes the muscle fatigue index of the continuous operation posture through the fatigue monitoring device. The stress-time integral method is used to quantify the fatigue accumulation of each posture in combination with the time-series changes of the microstress path and the center of gravity offset. During the assessment process, special attention is paid to the high stress area and the posture with abnormal center of gravity offset, and the cumulative trend of its stress load over time is calculated. The time-series posture fatigue assessment tool compares these cumulative trends with the operation safety standards to generate fatigue level assessment results. For example, during the experiment, if the fatigue accumulation index of a certain posture reaches a critical value, it indicates that the posture has a high risk of operational fatigue. The formation of time-series posture fatigue assessment data provides a dual reference of time and stress for the precise adjustment of the safety boundary. The action posture safety boundary is reconstructed based on the posture microstress path data, microstress center of gravity offset data and time-series posture fatigue assessment data. During the implementation process, the stress path, center of gravity offset and fatigue assessment data of the posture are first comprehensively analyzed to determine the risk level of the posture under different operating conditions. The safety boundary reconstruction tool is used to dynamically adjust the safety boundary of each posture.During the adjustment process, the safety boundary constraints and safety parameters are reconstructed for high-load areas in the stress path, locations where the center of gravity offset exceeds the threshold, and moments of high fatigue. Through reconstruction, the high-risk areas and safe areas of posture operations are clearly divided, and the operation tolerance and warning range are set. Finally, the reconstructed posture safety boundary is verified in practice to ensure that it can effectively avoid posture risks during the experiment. The obtained action posture safety boundary data contains detailed posture risk boundary parameters, which provides a comprehensive safety benchmark for the subsequent identification of hidden risk points.

[0052] Preferably, step S33 includes the following steps: Step S331: analyzing the hidden danger triggering condition of the operation space hidden danger associated data according to the action posture safety boundary data to obtain hidden danger triggering condition data; Step S332: reconstructing the risk transmission chain of the hidden danger associated data in the operation space based on the hidden danger triggering condition data to obtain hidden danger risk transmission chain data; Step S333: performing pre-trigger behavior detection based on the hidden danger risk transmission chain data to obtain pre-trigger behavior detection data; Step S334: Pre-identify hidden danger triggering risk points based on the hidden danger risk transmission chain data and pre-trigger behavior detection data to obtain hidden danger triggering risk point pre-data.

[0053] In an embodiment of the present invention, according to the action posture safety boundary data, the operation space hidden danger associated data is analyzed for hidden danger triggering conditions, the purpose is to clarify the specific conditions for the occurrence of hidden dangers. During implementation, the risk area and boundary parameters marked in the action posture safety boundary data are first extracted, and the hidden danger conditions of each operation position in the laboratory are analyzed one by one in combination with the operation space hidden danger associated data. A physical interactive analysis device is used to scan and cross-compare the laboratory equipment, instrument placement and personnel action paths, focusing on analyzing the triggering factors of potential hidden dangers in different operation scenarios, such as the intersection of unstable posture and high stress area. Through the comparison of these cross data, the hidden danger triggering condition data is finally output, and the specific triggering environment, associated actions and critical parameters of each hidden danger are clarified, providing detailed triggering basis for the next step of hidden danger transmission chain reconstruction. Based on the hidden danger triggering condition data, the risk transmission chain of the operation space hidden danger associated data is reconstructed. During specific implementation, the risk association analysis tool is used to connect and analyze the triggering conditions of each hidden danger with the chain risk event to reconstruct the transmission path of the hidden danger risk. During the analysis, the specific location of each hidden danger triggering condition in the operation space is first located, and then a multi-dimensional risk transmission path reconstruction device is used to associate the physical, chemical, and biological factors in the laboratory environment one by one to form a logical chain of risk transmission. For example, in the chemical operation area, if the triggering condition of reagent splashing occurs, its risk transmission chain includes multiple links such as contact risk, chemical reaction risk, and equipment damage risk. During the reconstruction process, the transmission time, impact range, and degree of impact of each risk point are quantified and arranged to obtain the hidden danger risk transmission chain data, clarifying the risk transmission path and key nodes. Pre-trigger behavior detection is carried out based on the hidden danger risk transmission chain data, aiming to identify the precursor behavior of hidden dangers in advance. In specific implementation, the high-frequency risk chain nodes in the operation process are monitored and analyzed in real time through the pre-trigger detection system. The system synchronously collects on-site actions, environmental changes, and material reactions through multiple sensors, and compares these data with the preset conditions in the hidden danger risk transmission chain. During the detection, special attention is paid to the high-risk transmission nodes in the hidden danger risk transmission chain data, such as abnormal vibration of equipment, rapid temperature changes, and abnormal movement amplitude of operators. Through threshold alarms and dynamic comparisons, pre-trigger behaviors that cause risk transmission are identified in real time. The formation of pre-trigger behavior detection data provides a basis for early intervention in hidden danger prevention and control, ensuring that appropriate emergency measures are taken before the hidden danger is officially triggered. Based on the hidden danger risk transmission chain data and pre-trigger behavior detection data, pre-identification of hidden danger triggering risk points is carried out, with the goal of clarifying the pre-risk points of hidden dangers and actively controlling them. During the implementation process, the risk point identification system is used to conduct a comprehensive analysis of the transmission chain data and detection data, and the risk cross-comparison technology is used to deeply identify the pre-conditions of each hidden danger risk point.During the identification process, the key abnormal behaviors in the pre-trigger behavior detection data are paired with the various risk points listed in the hidden danger risk transmission chain for analysis, focusing on identifying the node where the abnormality first occurs in the chain and judging the degree to which the node stimulates subsequent risks. For example, if an operating device frequently overheats in a low-temperature environment, the node is marked as a high-priority pre-risk point. The final output of the hidden danger trigger risk point pre-data clarifies the location and characteristics of the high-risk pre-nodes, providing comprehensive risk precursor information for the formulation of prevention and control strategies.

[0054] Preferably, step S4 comprises the following steps: Step S41: performing potential risk logic learning on the pre-data of the hidden danger triggering risk point based on a feedforward neural network algorithm to obtain potential risk logic learning data; Step S42: Designing an automated supervision process firmware according to the potential risk logic learning data to obtain a potential risk logic supervision firmware; Step S43: Embed the potential risk logic supervision firmware into the data cloud platform to perform security process supervision.

[0055] In the embodiment of the present invention, the potential risk logic learning is performed on the pre-data of the hidden danger triggering risk point based on the feedforward neural network algorithm to identify and understand the deep logical relationship of the hidden danger triggering. During the implementation process, the risk point data is mapped by multi-level nonlinear mapping using the feedforward neural network, and the learning of the hidden danger triggering mechanism is realized through the step-by-step calculation of the input layer, the hidden layer and the output layer. First, the pre-data of the hidden danger triggering risk point is input into the input layer of the neural network. After the weighted calculation of multiple hidden layers and the processing of the activation function, the network gradually learns the nonlinear relationship between the hidden danger triggering risk point and other potential factors. In the process, the network weight is adjusted through the continuous iteration of the training data, so that the output layer can accurately map the logical relationship between the hidden danger triggering condition and the potential risk. The potential risk logic learning data finally output contains the characteristic parameters of each hidden danger point and its associated potential risk logic, providing data support for the design of the subsequent supervision firmware. The purpose of designing the firmware of the automated supervision process according to the potential risk logic learning data is to transform the risk logic learning results into an executable supervision mechanism. During implementation, the risk triggering conditions, parameter thresholds and logical relationships in the potential risk logic learning data are embedded into the firmware control instructions through the firmware design tool. During the design process, the risk logic is mapped to specific supervision actions using logic controllers and event response modules. For example, for specific hidden danger trigger points, such as chemical spills or temperature anomalies, response instructions such as automatic alarms, operation suspension, or emergency isolation are set. By writing firmware code, various risk logic conditions are bound to sensors and controllers in the laboratory, enabling them to perceive and respond to potential risk situations in real time. The completed potential risk logic supervision firmware not only has the ability to identify risks, but also can automatically execute preset safety control operations when specific conditions are detected, thereby realizing the automated management of laboratory risks. The potential risk logic supervision firmware is embedded in the data cloud platform to perform safety process supervision. During the implementation process, the firmware is first connected to the hardware control unit of the laboratory supervision platform through a standardized interface to ensure that the firmware logic is compatible with the existing control architecture of the platform. Then, the potential risk logic supervision firmware is written into the embedded processor of the platform through the firmware burning device, so that the firmware establishes data interaction and instruction execution channels with the sensor system, alarm system, and automatic control equipment of the platform. After embedding, the firmware monitors the status changes of key risk points in the laboratory in real time in the background, and dispatches various emergency control measures in real time according to the potential risk logic. For example, when a risk trigger condition is detected, the firmware automatically instructs to shut down related equipment, start emergency ventilation, or activate protective devices to prevent the further development of hidden dangers. Through the embedding and execution of firmware, the laboratory safety process is fully automated and supervised, greatly improving the laboratory's risk prevention and control capabilities.

[0056] Preferably, the present invention further provides a medical safety process supervision system based on data analysis, which is used to execute the medical safety process supervision method based on data analysis as described above, and the medical safety process supervision system based on data analysis includes: The process action decomposition module is used to monitor the whole process experiment operation process through electronic monitoring equipment to obtain the whole process experiment operation process monitoring video; based on the whole process experiment operation process monitoring video, the operation process action is decomposed to obtain the operation process action decomposition data; The operation risk evolution module is used to extract action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; to refine the abnormal operation action characteristics of the operation process action behavior parameters to obtain abnormal operation action characteristic refinement data; to evaluate the action posture risk evolution on the abnormal operation action characteristic refinement data to obtain the operation posture risk evolution evaluation data; The risk pre-analysis module is used to reconstruct the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain the action posture safety boundary data; pre-identify the hidden danger triggering risk point according to the action posture safety boundary data to obtain the hidden danger triggering risk point pre-data; The automated supervision module is used to design automated supervision process firmware based on the pre-data of risk points that trigger hidden dangers, and obtain potential risk logic supervision firmware; the potential risk logic supervision firmware is embedded in the data cloud platform to perform security process supervision.

[0057] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0058] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A medical safety process supervision method based on data analysis, It is characterized in that The following steps are involved: Step S1: Monitor the whole process of experimental operation process through electronic monitoring equipment to obtain a monitoring video of the whole process of experimental operation process; Decompose the operation process based on the monitoring video of the whole process experiment operation process to obtain the operation process action decomposition data; Step S2: extracting action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; performing abnormal operation action feature refinement processing on the operation process action behavior parameters to obtain abnormal operation action feature refinement data; Performing action posture risk evolution assessment on the abnormal operation action feature refinement data to obtain operation posture risk evolution assessment data; Step S3: reconstructing the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain the action posture safety boundary data; Pre-identify the hidden danger triggering risk points based on the action posture safety boundary data to obtain the hidden danger triggering risk point pre-data; Step S4: Design an automated supervision process firmware based on the pre-data of the hidden danger triggering risk point to obtain a potential risk logic supervision firmware; Embed potential risk logic supervision firmware into the data cloud platform to perform security process supervision.

2. According to the medical safety process supervision method based on data analysis according to claim 1, It is characterized in that Step S1 includes the following steps: Step S11: monitor the whole process of experimental operation process through electronic monitoring equipment to obtain a monitoring video of the whole process of experimental operation process; Step S12: performing contrast optimization processing on the whole-process experimental operation process monitoring video to obtain an operation process monitoring contrast-optimized video; Step S13: switching and integrating the operation process monitoring contrast-optimized video from multiple angles to obtain a multi-angle integrated video of the operation process monitoring; Step S14: Decompose the operation process actions based on the multi-angle integrated video of the operation process monitoring to obtain the operation process action decomposition data.

3. According to the medical safety process supervision method based on data analysis according to claim 1, It is characterized in that Step S2 includes the following steps: Step S21: Obtaining standardized operation process data; Step S22: extracting action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; Step S23: performing abnormal operation action identification on the operation process action behavior parameters according to the standardized operation process data to obtain abnormal operation action data; Step S24: performing abnormal operation action feature refinement processing on the operation process action behavior parameters according to the abnormal operation action data to obtain abnormal operation action feature refinement data; Step S25: Perform action posture risk evolution assessment on the abnormal operation action feature refinement data to obtain operation posture risk evolution assessment data.

4. According to the medical safety process supervision method based on data analysis according to claim 3, It is characterized in that Step S24 includes the following steps: Step S241: calculating the abnormal operation action dynamic deviation of the operation process action behavior parameters according to the abnormal operation action data to obtain the abnormal operation action dynamic deviation data; Step S242: performing a multi-dimensional parameter heterogeneous comparison on the operation process action behavior parameters based on the abnormal operation action dynamic deviation data to obtain multi-dimensional parameter heterogeneous data; Step S243: performing microscopic motion scale analysis according to the abnormal operation motion dynamic deviation data and the multi-dimensional parameter heterogeneous data to obtain multi-dimensional microscopic motion scale data; Step S244: performing a subtle motion feature deep analysis on the multi-dimensional micro-motion scale data to obtain multi-dimensional micro-motion scale deep analysis data; Step S245: performing abnormal operation action feature refinement processing according to the multi-dimensional micro-action scale depth analysis data to obtain abnormal operation action feature refinement data.

5. According to the medical safety process supervision method based on data analysis according to claim 3, It is characterized in that Step S25 includes the following steps: Step S251: performing posture stress distribution analysis on the abnormal operation action feature refinement data to obtain posture stress distribution data; Step S252: performing microseismic force decoupling processing on the abnormal operation action feature refinement data according to the posture stress distribution data to obtain posture action microseismic force decoupling data; Step S253: performing displacement residual analysis on the posture action microseismic force decoupling data to obtain microseismic force displacement residual data; Step S254: Perform risk evolution assessment of the action posture based on the microseismic force decoupling data of the posture action and the microseismic force displacement residual data to obtain the risk evolution assessment data of the operation posture.

6. According to the medical safety process supervision method based on data analysis according to claim 1, It is characterized in that Step S3 includes the following steps: Step S31: reconstructing the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain the action posture safety boundary data; Step S32: performing operation space hidden danger association analysis according to the action posture safety boundary data to obtain operation space hidden danger association data; Step S33: pre-identify hidden danger triggering risk points on the hidden danger associated data of the operation space according to the action posture safety boundary data to obtain hidden danger triggering risk point pre-data.

7. According to the medical safety process supervision method based on data analysis according to claim 6, It is characterized in that Step S31 includes the following steps: Step S311: performing operation posture microstress path analysis on abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain posture microstress path data; Step S312: performing gravity center offset calculation on posture microstress path data to obtain microstress gravity center offset data; Step S313: performing time-series posture fatigue evaluation according to the posture micro-stress path data and the micro-stress center of gravity offset data to obtain time-series posture fatigue evaluation data; Step S314: reconstruct the action posture safety boundary according to the posture microstress path data, the microstress center of gravity offset data and the time series posture fatigue assessment data to obtain the action posture safety boundary data.

8. According to the medical safety process supervision method based on data analysis according to claim 6, It is characterized in that Step S33 includes the following steps: Step S331: analyzing the hidden danger triggering condition of the operation space hidden danger associated data according to the action posture safety boundary data to obtain hidden danger triggering condition data; Step S332: reconstructing the risk transmission chain of the hidden danger associated data in the operation space based on the hidden danger triggering condition data to obtain hidden danger risk transmission chain data; Step S333: performing pre-trigger behavior detection based on the hidden danger risk transmission chain data to obtain pre-trigger behavior detection data; Step S334: Pre-identify hidden danger triggering risk points based on the hidden danger risk transmission chain data and pre-trigger behavior detection data to obtain hidden danger triggering risk point pre-data.

9. The medical safety process supervision method based on data analysis according to claim 1, It is characterized in that Step S4 includes the following steps: Step S41: performing potential risk logic learning on the pre-data of the hidden danger triggering risk point based on a feedforward neural network algorithm to obtain potential risk logic learning data; Step S42: Designing an automated supervision process firmware according to the potential risk logic learning data to obtain a potential risk logic supervision firmware; Step S43: Embed the potential risk logic supervision firmware into the data cloud platform to perform security process supervision.

10. A medical safety process supervision system based on data analysis, It is characterized in that Used to execute the medical safety process supervision method based on data analysis as claimed in claim 1, the medical safety process supervision system based on data analysis includes: The process action decomposition module is used to monitor the whole process experiment operation process through electronic monitoring equipment to obtain the whole process experiment operation process monitoring video; based on the whole process experiment operation process monitoring video, the operation process action is decomposed to obtain the operation process action decomposition data; The operation risk evolution module is used to extract action behavior parameters from the operation process action decomposition data to obtain the operation process action behavior parameters; to refine the abnormal operation action characteristics of the operation process action behavior parameters to obtain abnormal operation action characteristic refinement data; to evaluate the action posture risk evolution on the abnormal operation action characteristic refinement data to obtain the operation posture risk evolution evaluation data; The risk pre-analysis module is used to reconstruct the action posture safety boundary of the abnormal operation action feature refinement data based on the operation posture risk evolution assessment data to obtain the action posture safety boundary data; pre-identify the hidden danger triggering risk point according to the action posture safety boundary data to obtain the hidden danger triggering risk point pre-data; The automated supervision module is used to design automated supervision process firmware based on the pre-data of risk points that trigger hidden dangers, and obtain potential risk logic supervision firmware; the potential risk logic supervision firmware is embedded in the data cloud platform to perform security process supervision.

Citation Information

Patent Citations

  • Experimental operation monitoring and early warning method, device and system

    CN116959216A

  • Medical big data full-process intelligent supervision method

    CN117219250A

  • Abnormality analysis method and system based on medical data

    CN118035921A

  • Road construction safety monitoring method and system

    CN118094396A

  • Intelligent monitoring system for evaluating safety of gynecological minimally invasive surgery

    CN118197650A

Cited By

  • Laboratory intelligent comprehensive management method and system

    CN119359265A

  • Chemical safety production patrol control system and method

    CN120297743A

  • Postoperative follow-up visit system and method for maxillofacial malformation orthognathic operation

    CN120340733A