Intelligent monitoring management system and method for disinfection supply area

Abstract

The invention relates to the technical field of medical informatization and intelligent monitoring, and discloses an intelligent monitoring management system and method for a disinfection supply area, and the method comprises the steps: obtaining the environment parameters of a disinfection supply center and an instrument batch source identifier, and judging the risk level; judging trend abnormity and predicting a future environment parameter value, and comparing a predicted value with a control limit value to identify an over-limit risk; quantitatively evaluating the possibility and severity of pollution cross-regional diffusion; dynamically adjusting according to the risk level change, issuing an updating instruction and recording a log; operating process parameters of sterilization equipment are collected, and process and inspection data are integrated according to sterilization batches; calculating a comprehensive quality score, judging a batch quality grade, and generating a sterilization quality evaluation result and a batch processing suggestion mark; generating a quality early warning message, determining an early warning level according to a grading rule, and carrying out grading pushing; the risk self-adaptive intelligent monitoring management of the disinfection supply area is realized.

Description

Technical Field

[0001] This invention relates to the field of medical information technology and intelligent monitoring technology, and more specifically, to an intelligent monitoring and management system and method for a disinfection supply area. Background Technology

[0003] Existing monitoring and management technologies for disinfection supply areas mainly include environmental parameter monitoring systems, medical device traceability management systems, and sterilization quality monitoring systems. Environmental parameter monitoring systems can collect basic parameters such as temperature, humidity, and differential pressure and issue alarms when limits are exceeded. However, they mostly use fixed threshold judgment methods, failing to identify gradual abnormal trends in parameters and lacking proactive predictive and early warning capabilities. Medical device traceability management systems trace back to sterilization packages, resulting in coarse granularity, and the traceability data is primarily statically recorded, lacking correlation analysis with environmental monitoring data. Sterilization quality judgment mainly relies on manual observation of chemical indicator card color changes, which is highly subjective, inconsistent, and difficult to achieve objective quantitative evaluation.

[0004] The main shortcomings of the existing technologies mentioned above are: the subsystems are independent of each other, data is not effectively integrated, and risks cannot be assessed from a global perspective; there is a lack of a differentiated monitoring mechanism based on risk levels, making it impossible to adaptively adjust monitoring intensity for high-risk treatment targets; and environmental anomaly responses are mainly passive alarms, failing to achieve proactive prediction and early intervention. Therefore, there is a need for an intelligent monitoring and management technology solution that can achieve multi-source data fusion analysis, risk-adaptive monitoring strategy adjustment, and objective quantitative evaluation of disinfection quality. Summary of the Invention

[0005] This invention provides an intelligent monitoring and management system and method for disinfection supply areas, which solves the technical problems in related technologies, such as the independence of subsystems, the lack of a differentiated monitoring mechanism based on risk level, and the inability to achieve proactive prediction and early intervention.

[0006] This invention provides an intelligent monitoring and management method for a disinfection supply area, comprising: Obtain environmental parameters and equipment batch source identification from the disinfection supply center, determine the risk level, and obtain an environmental perception dataset with risk identification; Based on the environmental perception dataset with risk labels, abnormal trends are identified and future environmental parameter values ​​are predicted. Based on the prediction results, the risk of exceeding limits is identified, and the environmental parameter trend analysis results are obtained. Based on the results of environmental parameter trend analysis, the severity of cross-regional pollution diffusion is quantitatively assessed, and the risk assessment results of cross-regional pollution diffusion are obtained. Based on the cross-regional pollution spread risk assessment results, the monitoring strategy is matched, dynamically adjusted according to changes in risk level, and update instructions are issued and logs are recorded to obtain the monitoring and control strategy. Collect sterilization equipment operation parameters according to the collection requirements of the monitoring and control strategy, integrate process and inspection data according to sterilization batches, and obtain a disinfection and sterilization quality characteristic dataset; Construct a historical quality benchmark model for similar batches, identify abnormal quality patterns and trace the associated environmental risk factors based on a disinfection and sterilization quality feature dataset, and obtain intelligent evaluation results for sterilization quality. Based on the intelligent evaluation results of sterilization quality, differentiated processing suggestions and quality early warning messages are generated. The early warning level is determined and traceability information is associated to push the messages in a graded manner, resulting in graded early warning push results with traceability information.

[0007] In a preferred embodiment, the step of obtaining the environmental awareness dataset with risk labels includes: Multiple types of environmental sensors are deployed in each functional area, and each sensor collects data according to a preset sampling period; Pressure differential monitoring points are set up at the boundary between adjacent areas to collect time-series data of pressure differential gradient between areas; Obtain the unique identification code of the medical device batch by using a barcode scanning device or radio frequency identification reader, and then query the department risk level mapping table to determine the risk level of the medical device batch. Based on the arrival timestamp of the medical device batch, environmental parameter data and differential pressure gradient data within a preset time window are extracted and integrated with the risk level identifier of the medical device batch to obtain an environmental perception dataset with risk identifier.

[0008] In a preferred embodiment, the step of obtaining the environmental parameter trend analysis results includes: The sliding window statistical method is used to extract statistical features of time series data of environmental parameters, including window mean, window standard deviation, window range and slope of linear change; Trend anomaly judgment is based on an expert rule base, which includes slope anomaly rules, fluctuation anomaly rules, continuous offset rules, and differential pressure attenuation rules. Construct a time-series prediction model for environmental parameters to predict parameter values ​​for future time periods; By comparing the predicted value sequence with the control limits, identifying the predicted time points of exceeding the limits, calculating the expected lead time of exceeding the limits, and combining the predicted exceeding information with the rule-based anomaly markers, the trend analysis results of environmental parameters are obtained.

[0009] In a preferred embodiment, the step of obtaining the cross-regional pollution diffusion risk assessment results includes: Based on the monitoring data of the switch status sensor and the door magnetic sensor, the number of times the transfer window is opened, the opening duration and the number of people passing through within the evaluation time window are statistically evaluated. After normalization, the weighted sum is used to obtain the inter-regional transfer activity intensity index. Analyze time-series data of differential pressure gradient to identify periods when differential pressure values ​​are below the safe threshold and calculate the abnormal exposure rate of differential pressure. Anomaly marker information of the upstream region is extracted from the results of environmental parameter trend analysis, and the anomaly risk score of the upstream region is obtained by summing them. The cross-regional pollution diffusion comprehensive risk score is calculated and risk levels are classified by weighting and integrating the inter-regional transmission activity intensity index, pressure difference abnormal exposure rate index, upstream regional abnormal risk score and device comprehensive risk level label.

[0010] In a preferred embodiment, the step of obtaining the monitoring and control strategy configuration includes: Based on the comprehensive risk level identification of medical device batches, the corresponding basic monitoring strategy is matched from the tiered strategy template library, which includes basic strategy templates, enhanced strategy templates, and strict control strategy templates. When the risk assessment result indicates a high risk level and the duration reaches the first duration threshold, the current strategy template will be upgraded to a higher level. When the risk level decreases and the duration reaches the second duration threshold, the strategy template will be restored to the basic configuration corresponding to the risk level of the device batch. Send policy update instructions to the edge computing gateway to update the sensor sampling period configuration, data reporting frequency configuration, and early warning threshold configuration, and record the policy change log.

[0011] In a preferred embodiment, the step of obtaining the disinfection and sterilization quality characteristic dataset includes: By establishing a data communication interface with the sterilization equipment and configuring the data acquisition requirements according to the monitoring and control strategy, the equipment parameters during the sterilization process are collected in real time. Based on the standard stages of the sterilization process, key quality characteristic parameters for each stage were extracted. Image data of chemical indicator cards are acquired through an image acquisition device, and image preprocessing and color feature analysis are performed to calculate the quantitative value of the degree of color change of the chemical indicator. Using the unique identifier of the sterilization batch as the association key, the process quality characteristic parameters of each stage, the quantitative value of the degree of chemical indicator color change, and the risk level identifier of the source of the device are integrated to obtain the disinfection and sterilization quality characteristic dataset.

[0012] In a preferred embodiment, the specific steps for obtaining the intelligent evaluation result of sterilization quality include: Historical qualified batches were classified according to sterilization equipment type, sterilization procedure type, and instrument risk level. Statistical benchmark values ​​and standard deviations of each key quality characteristic parameter were calculated, and a historical quality benchmark model for the same batch was constructed. Extract the measured values ​​of each key quality characteristic parameter of the current batch, calculate the standardized deviation of each dimension from the historical benchmark value, and calculate the comprehensive deviation index; The abnormal quality pattern types are identified based on the combined features of the deviation dimension, including heating efficiency reduction mode, process control instability mode, and chemical indicator abnormal mode. Extract the trend analysis results of environmental parameters and the risk assessment results of cross-regional pollution diffusion for the corresponding time period of the current batch, establish the source-tracing correlation between quality anomalies and environmental risk factors, and generate anomaly source-tracing factor identifiers; Based on the comprehensive deviation, abnormal quality pattern classification, and abnormal source identification, the batch quality level is determined and differentiated processing suggestions are generated.

[0013] In a preferred embodiment, the step of pushing the graded early warning result with traceability information includes: Abnormal information is extracted from environmental parameter trend analysis results, cross-regional pollution diffusion risk assessment results, and sterilization quality intelligent evaluation results with abnormal source tracing identifiers to generate environmental early warning messages, pollution risk early warning messages, and quality early warning messages; For quality warning messages, the warning level is determined by integrating the environmental risk factors traced back to their source; Based on the warning level and push rule configuration, the warning message is pushed to different channels, and the abnormal quality mode type and the environmental risk factor identifier traced to the source are attached to the push content. Set a time window for deduplication of push notifications, and record the push time, push channel, delivery status, and source tracing information summary of the alerts.

[0014] In a preferred embodiment, the construction of the environmental parameter time series prediction model further includes: The system continuously monitors the prediction error of the prediction model. When the average prediction error within a preset number of prediction periods exceeds the error tolerance threshold, it triggers an incremental update process for the model parameters. The mini-batch gradient descent method is used to fine-tune the model parameters using measured data, and the adjustment range is controlled by a preset learning rate parameter. The updated model parameters replace the original parameters, and the timestamp and triggering reason of the model update are recorded to form a model evolution log.

[0015] This invention provides an intelligent monitoring and management system for a disinfection supply area, used to execute the aforementioned intelligent monitoring and management method for a disinfection supply area, comprising: The environmental perception data acquisition module is used to acquire environmental parameters and instrument batch source identification of the disinfection supply center, determine the risk level, and obtain an environmental perception dataset with risk identification. The environmental parameter trend analysis module, based on an environmental perception dataset with risk labels, identifies abnormal trends and predicts future environmental parameter values. Based on the prediction results, it identifies risks of exceeding limits and obtains the environmental parameter trend analysis results. The cross-regional pollution diffusion risk assessment module, based on the results of environmental parameter trend analysis, quantitatively assesses the severity of cross-regional pollution diffusion and obtains the cross-regional pollution diffusion risk assessment results. The monitoring and control strategy configuration module matches monitoring strategies based on the cross-regional pollution spread risk assessment results, dynamically adjusts them according to changes in risk level, issues update instructions and records logs to obtain the monitoring and control strategy. The disinfection and sterilization quality data acquisition module is used to collect the operating parameters of the sterilization equipment according to the acquisition requirements of the monitoring and control strategy, and integrate the process and inspection data by sterilization batch to obtain the disinfection and sterilization quality characteristic dataset. The sterilization quality evaluation module is used to build a historical quality benchmark model for similar batches. Based on the disinfection and sterilization quality feature dataset, it identifies abnormal quality patterns and traces the associated environmental risk factors to obtain intelligent sterilization quality evaluation results. The tiered early warning push module generates differentiated processing suggestions and quality early warning messages based on the intelligent evaluation results of sterilization quality, determines the early warning level and associates it with traceability information for tiered push, resulting in tiered early warning push results with traceability information.

[0016] The beneficial effects of this invention are as follows: By constructing a multi-source environmental perception and risk level autonomous judgment mechanism, and combining a trend analysis algorithm based on rule-based and adaptive model fusion with a multi-factor weighted cross-regional pollution diffusion risk assessment model, the system achieves dynamic adaptation between monitoring strategies and risk levels. The system can adaptively adjust environmental monitoring frequency, early warning thresholds, and quality control standards based on the source risk level of the equipment to be treated and the real-time assessed environmental risk status. This ensures monitoring effectiveness while avoiding excessive resource consumption, resulting in higher management efficiency and better targeted risk prevention compared to traditional fixed-strategy monitoring models.

[0017] By collecting multi-stage operational parameters and chemical indicator card image features during the sterilization process, a comprehensive sterilization quality evaluation model based on weighted fusion of sub-item scores was established, achieving an objective and quantitative evaluation of sterilization quality. This evaluation model integrates the parameter compliance of each stage of the sterilization process with the degree of color change of chemical indicators from multiple dimensions. It employs differentiated evaluation weights based on the risk level of the device, outputting a quantitative quality score and clear handling recommendations. This eliminates the subjective differences inherent in traditional manual interpretation methods, providing objective and reliable data support for sterilization quality management. Attached Figure Description

[0018] Figure 1This is a flowchart of the main process of an intelligent monitoring and management method for a disinfection supply area according to the present invention; Figure 2 This is a detailed flowchart of an intelligent monitoring and management method for a disinfection supply area according to the present invention; Figure 3 This is a block diagram of an intelligent monitoring and management system for a disinfection supply area according to the present invention. Detailed Implementation

[0019] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, some features described in the examples may be combined in other examples.

[0020] At least one embodiment of the present invention discloses an intelligent monitoring and management method for a disinfection supply area, such as... Figures 1 to 2 As shown, it includes: Step 1: Obtain environmental parameters and equipment batch source identification from the disinfection supply center, determine the risk level, and obtain an environmental perception dataset with risk identification. Specifically, the following steps are included: Step 1.1: Deployment and differentiated sampling configuration of various types of environmental sensors in each functional area; Based on the functional zoning of the disinfection supply center and the environmental control standards for each area, multiple types of environmental sensors are deployed at representative monitoring points in each area. Specifically, temperature and humidity sensors and ammonia concentration sensors are deployed in the decontamination area. The temperature and humidity sensors are used to monitor the area's temperature and relative humidity, while the ammonia concentration sensors are used to monitor potentially harmful gases generated during the cleaning process. Temperature and humidity sensors and particle counters are deployed in the inspection, packaging, and sterilization areas. The particle counters are used to monitor the concentration of suspended particles in the air to assess air cleanliness. Temperature and humidity sensors and particle counters are deployed in the sterile goods storage area, and the density of monitoring points is appropriately increased to meet higher cleanliness monitoring requirements.

[0021] Each sensor collects data according to a preset sampling cycle. The first sampling cycle for the decontamination area is five minutes, the second sampling cycle for the packaging and sterilization area is three minutes, and the third sampling cycle for the sterile item storage area is two minutes. By configuring the different sampling frequencies, precise monitoring of areas with different cleanliness requirements can be achieved.

[0022] Based on the above sensor deployment and sampling configuration, environmental parameters such as temperature, relative humidity, dust particle concentration, and ammonia concentration are collected in each area to obtain raw environmental parameter data for each area. This data is stored in a structured format with timestamp, area identifier, parameter type, and parameter value.

[0023] Step 1.2: Setting up pressure differential monitoring points at the boundary of adjacent areas and collecting pressure differential gradient data; To maintain a pressure differential gradient between the functional areas of the sterilization supply center, pressure differential monitoring points are set up at the boundaries of adjacent areas. Specifically, differential pressure transmitters are installed on both sides of the pass-through window between the decontamination area and the inspection, packaging, and sterilization area to monitor the positive pressure differential from the inspection, packaging, and sterilization area to the decontamination area; differential pressure transmitters are installed at the entrance / exit of the buffer zone between the inspection, packaging, and sterilization area and the sterile goods storage area to monitor the positive pressure differential from the sterile goods storage area to the inspection, packaging, and sterilization area; and differential pressure transmitters are installed at personnel passages between each area and the external corridor to monitor the pressure differential between the area and the outside environment. The differential pressure transmitters perform continuous measurements at a sampling frequency of one minute and trigger event recording when the pressure differential value changes by more than three Pascals. Based on the measurement data from the differential pressure transmitters, the instantaneous pressure differential value and pressure differential change events at each monitoring point are recorded to obtain the time-series data of the inter-area pressure differential gradient. This data includes fields such as monitoring point identification, timestamp, pressure differential value, and pressure differential status marker.

[0024] Step 1.3, Self-determination of the risk level of medical device batches based on source identification; Based on the source identification information of the batches of medical devices to be processed and returned to the decontamination area, the risk level of each batch is determined autonomously. Specifically, barcode scanning equipment or RFID readers are deployed at the medical device receiving station in the decontamination area. When a batch of medical devices arrives, the unique identification code of the batch is obtained by scanning the barcode label affixed to the medical device turnover container or reading the embedded RFID tag.

[0025] Based on the unique batch identifier of the medical device, a pre-established departmental risk classification mapping table is queried. This mapping table configures risk levels according to department type and device usage scenario. Devices used in general wards and outpatient clinics correspond to the first risk level (routine risk); devices used in intensive care units and operating rooms correspond to the second risk level (medium risk); and devices used in infectious disease departments and fever clinics correspond to the third risk level (high risk). When scanning and identification fail or cannot match the mapping table, the system automatically marks the batch as the second risk level to ensure a safety margin. Based on the above identification and mapping process, a risk level identifier for the medical device batch is obtained, which is stored in association with the unique batch identifier and arrival timestamp.

[0026] Step 1.4: Time alignment and risk identification association integration of multi-source data; Based on raw environmental parameter data, time-series data of inter-regional pressure gradients, and risk level identifiers for medical device batches, data time alignment and correlation integration are performed. Specifically, using the time stamp of a medical device batch arriving at the decontamination zone as a baseline, environmental parameter data and pressure gradient data within 15-minute time windows before and after that time stamp are extracted, and these data are correlated with the risk level identifiers of the medical device batches. For cases where multiple medical device batches exist within the same time period, the highest risk level among all batches is taken as the comprehensive risk level identifier for that time period. Based on the above data integration process, an environmental perception dataset with risk identifiers is obtained. Each record in this dataset contains complete fields such as timestamp, area identifier, various environmental parameter values, inter-regional pressure gradient values, and comprehensive risk level identifiers, providing a unified data foundation for subsequent analysis and processing.

[0027] Step 2: Based on the environmental perception dataset with risk labels, determine abnormal trends and predict future environmental parameter values. Identify the risk of exceeding limits based on the prediction results and obtain the environmental parameter trend analysis results. Specifically, the following steps are included: Step 2.1, Extraction of parameter change features based on the sliding window statistical method; Based on time-series environmental parameter data for each region in an environmental perception dataset with risk labels, a sliding window statistical method is used to extract statistical features of parameter changes. Specifically, the sliding window width is set to 30 minutes, and the window moves forward in five-minute increments. At each window position, statistical features of the temperature, humidity, and pressure difference sequences within the window are calculated, including the window mean, window standard deviation, the difference between the maximum and minimum values ​​within the window (i.e., the window range), and the slope of the linear change obtained by least squares fitting. The window mean reflects the average level of the parameter during that period, the window standard deviation reflects the degree of parameter fluctuation, the window range reflects the instantaneous change amplitude of the parameter, and the slope of the linear change reflects the overall trend direction and rate of parameter change. Based on the above sliding window statistical calculations, the sliding window statistical features of environmental parameters are obtained, and this feature data is stored in the format of window end timestamp, region identifier, parameter type, and various statistical values.

[0028] Step 2.2, trend anomaly judgment based on expert rule base; Based on the sliding window statistical characteristics of environmental parameters and a pre-configured expert rule base, rule-based trend anomaly judgment is performed. The expert rule base includes the following types of anomaly judgment rules: slope anomaly rule: when the absolute value of the linear change slope of the temperature parameter exceeds 0.5 degrees Celsius per hour or the absolute value of the linear change slope of the humidity parameter exceeds 5% per hour, it is judged as a trend anomaly, indicating that the parameter is rapidly rising or falling; fluctuation anomaly rule: when the window standard deviation of the temperature parameter exceeds 1.5 degrees Celsius or the window standard deviation of the humidity parameter exceeds 10% per hour, it is judged as a fluctuation anomaly, indicating a decrease in parameter stability; continuous offset rule: when the window mean of a parameter is continuously higher or lower than the historical average for three consecutive windows by more than two standard deviations, it is judged as a continuous offset anomaly; pressure differential decay rule: when the linear change slope of the inter-regional pressure differential value is negative and the absolute value exceeds one Pascal per hour, it is judged as a pressure differential decay trend, indicating that the pressure differential gradient may be about to fail. Based on the above rule matching judgment, rule-judged anomaly markers are obtained, which include information such as anomaly type, triggering rule identifier, and anomaly severity score.

[0029] Step 2.3, Construction and updating of environmental parameter time series prediction model; Based on historical data accumulated in the environmental perception dataset with risk labels, a time-series prediction model for environmental parameters is constructed and updated. Specifically, the system continuously counts the number of historical data records for various parameters in each region. When the historical data volume for a certain parameter in a region reaches one thousand records, the training or update process of the prediction model for that parameter is triggered. The prediction model adopts a time-series prediction algorithm based on a recurrent neural network structure. This algorithm can predict parameter values ​​at future time points by learning the temporal dependencies in historical data. During model training, historical time-series data is divided into training and validation sets in an 8:2 ratio, and the prediction error of the model on the validation set is minimized through iterative optimization. When the amount of historical data does not reach the model training threshold, the system uses a simple prediction method based on the historical average of the same period as an alternative, that is, calculating the average parameter value for the same period in the historical data as the predicted value. Based on the above prediction model or simple prediction method, the parameter values ​​at each time point within the next sixty minutes are predicted to obtain a sequence of predicted environmental parameter values.

[0030] Step 2.4, Prediction of parameter exceedance risk and comprehensive anomaly evaluation; Based on the predicted value sequence of environmental parameters and the control limits for environmental parameters, the risk of parameter exceedance is predicted. Specifically, the predicted parameter values ​​at each time point in the predicted value sequence are compared with the corresponding upper and lower control limits to identify the time points when the predicted values ​​will exceed the control limits. When a predicted exceedance time point exists, the time difference between the current time and the predicted exceedance time point is calculated as the expected exceedance lead time. The urgency of the prediction warning is determined based on the length of the expected exceedance lead time; the shorter the expected exceedance lead time, the higher the urgency. The predicted exceedance information is integrated with the rule-based anomaly marking. When the rule-based judgment shows an abnormal trend and the prediction indicates that an exceedance is imminent, the confidence and severity scores of the anomaly warning are increased. Based on the above comprehensive analysis, the environmental parameter trend analysis results are obtained, which include the current status evaluation of each parameter in each region, trend anomaly marking, predicted exceedance risk, and comprehensive anomaly score.

[0031] Furthermore, since time-series prediction models based on fixed parameters may experience increased prediction bias when environmental conditions change, an online adaptive learning mechanism can be employed to dynamically adjust the prediction model. This aims to improve the model's adaptability to environmental changes and its prediction accuracy. Specifically, the system continuously monitors the prediction error of the model. When the average prediction error over the most recent ten prediction periods exceeds a 15% error tolerance threshold, an incremental update process for the model parameters is triggered. The incremental update uses a mini-batch gradient descent method, fine-tuning the model parameters using recent measured data. The adjustment magnitude is controlled by a preset learning rate parameter to balance the model's stability and adaptability. The updated model parameters replace the original parameters for subsequent predictions, and the timestamp and triggering reason for the model update are recorded, forming a model evolution log. Based on this online adaptive learning mechanism, environmental parameter prediction results with adaptive capabilities are obtained.

[0032] Step 3: Based on the trend analysis results of environmental parameters, quantitatively assess the severity of cross-regional pollution diffusion to obtain the risk assessment results of cross-regional pollution diffusion. Specifically, the following steps are included: Step 3.1, statistical and quantitative assessment of the intensity of inter-regional transmission activities; Based on the status monitoring data of the material transfer facilities and personnel passages between the functional areas of the disinfection supply center, the intensity of inter-area transfer activities within a two-hour assessment window was statistically analyzed. Specifically, switch status sensors were installed at the transfer windows between areas to detect opening and closing events; door magnetic sensors were installed on the doors of the buffer room and airlock to detect personnel passage events. Based on the transfer window opening and closing event records, the cumulative number of times each transfer window was opened and the cumulative opening duration were statistically analyzed within the assessment time window; based on the door magnetic sensor event records, the number of times personnel passed through each passage within the assessment time window was statistically analyzed. The number of times the transfer windows were opened, the opening duration, and the number of times personnel passed through were each divided by their respective baseline values ​​for normalization, and then weighted and summed according to preset weighting coefficients to obtain the inter-area transfer activity intensity index. The higher the transfer activity intensity index, the more frequent the exchange of materials and personnel between areas, and the greater the opportunity for contamination transmission.

[0033] Step 3.2, Analysis of abnormal exposure to differential pressure gradient; Based on the inter-regional differential pressure gradient time-series data obtained in step 1, the abnormal exposure of differential pressure gradient within the assessment time window is analyzed and evaluated. Specifically, for each differential pressure monitoring point, a differential pressure safety threshold of five Pascals is set. When the measured differential pressure value is lower than the safety threshold, the differential pressure protection is considered to have failed, and there is a risk of reverse pollution diffusion. The differential pressure time-series data within the assessment time window is traversed to identify the periods when the differential pressure value is lower than the safety threshold, and the total duration of these periods is accumulated as the cumulative duration of differential pressure abnormality. The ratio of the cumulative duration of differential pressure abnormality to the total duration of the assessment time window is calculated to obtain the differential pressure abnormality exposure rate index. The higher the differential pressure abnormality exposure rate, the worse the protective effect of the differential pressure gradient within the assessment period, and the weaker the physical barrier for cross-regional pollution diffusion. At the same time, the time overlap between the differential pressure abnormality period and the transfer window opening event is analyzed. If a transfer window opening event occurs within the differential pressure abnormality period, the risk weight of that period is further increased.

[0034] Step 3.3, Extraction of abnormal risk indicators in the upstream area; Based on the environmental parameter trend analysis results obtained in step 2, upstream area anomaly risk indicators that are indicative of cross-regional pollution diffusion risks are extracted. Specifically, the decontamination area, as the upstream area and potential pollution source area of ​​the equipment treatment process, has a potential impact on downstream areas due to its abnormal environmental conditions. Anomaly marker information of the decontamination area is extracted from the environmental parameter trend analysis results, including whether there are abnormal temperature and humidity trends, whether ammonia concentration exceeds the standard or shows an upward trend, and whether environmental parameter exceedances are predicted. These anomalies are assigned different risk point numbers according to their severity, and the scores are accumulated to obtain an upstream area anomaly risk score. The higher the upstream area anomaly risk score, the more severe the current or impending anomaly in the upstream area, and the greater the possibility of pollution transmission to downstream areas.

[0035] Step 3.4, Calculation and classification of comprehensive risk score based on multi-factor weighted fusion; Based on the intensity index of inter-regional transmission activities, the abnormal exposure rate index of pressure difference, the abnormal risk score of upstream areas, and the comprehensive risk level label of current treatment equipment, a multi-factor weighted fusion method is used to calculate the comprehensive risk score of cross-regional pollution diffusion. Specifically, a basic risk coefficient is determined based on the comprehensive risk level of the medical device. The first risk level corresponds to a first basic coefficient of 1.0, the second risk level corresponds to a second basic coefficient of 1.5, and the third risk level corresponds to a third basic coefficient of 2.0. The intensity of inter-regional transmission activities, the abnormal exposure rate of pressure differences, and the abnormal risk score of the upstream region are multiplied by their respective preset weighting coefficients and then summed to obtain the environmental risk weighted score. The basic risk coefficient is multiplied by the environmental risk weighted score to obtain the comprehensive risk score for cross-regional pollution diffusion. Based on the range of the comprehensive risk score, the assessment results are divided into three levels: low risk, medium risk, and high risk. The specific classification criteria are as follows: a comprehensive risk score less than 30 points is considered low risk, indicating a good current state requiring no special attention; a comprehensive risk score greater than or equal to 30 points and less than 60 points is considered medium risk, indicating a certain potential hazard requiring enhanced monitoring; and a comprehensive risk score greater than or equal to 60 points is considered high risk, indicating a significant risk requiring immediate action. Based on the above assessment calculations, the cross-regional pollution diffusion risk assessment results are obtained, including the comprehensive risk score, risk level classification, and identification of major risk factors.

[0036] Furthermore, since risk assessments based solely on current point-in-time data cannot reflect the dynamic evolution of risks, risk trend analysis can be used to extend the risk assessment over time. The aim is to identify the direction and speed of risk changes, providing a basis for early intervention. Specifically, the system stores historical records of comprehensive risk scores for the most recent twenty-four assessment periods, performs trend analysis on the time-series data of risk scores, and calculates the slope and acceleration of risk score changes. When the risk score continues to rise and the slope exceeds a preset threshold, it is determined to be an upward risk trend, and an early warning of rising risk is issued even if the current risk level is still low or medium. When the risk score rises rapidly and the acceleration is positive, it is determined to be an accelerated deterioration of risk, requiring priority attention. The results of risk trend analysis serve as supplementary information to the risk assessment results, enriching the descriptive dimensions of the risk status.

[0037] Step 4: Based on the cross-regional pollution spread risk assessment results, match the monitoring strategy, dynamically adjust it according to changes in risk level, issue update instructions and record logs to obtain the monitoring and control strategy; Specifically, the following steps are included: Step 4.1: Matching the hierarchical strategy template library and determining the basic monitoring strategy; Based on the comprehensive risk level identification of medical device batches, corresponding basic monitoring strategies are matched from a pre-established tiered strategy template library. The tiered strategy template library contains three levels of strategy templates: the basic level is suitable for scenarios involving devices of the first risk level, configuring standard environmental monitoring sampling cycles, standard early warning threshold settings, and standard quality control frequencies; the enhanced level is suitable for scenarios involving devices of the second risk level, configuring shortened environmental monitoring sampling cycles, tightened early warning threshold settings, and increased quality control frequencies; and the strict control level is suitable for scenarios involving devices of the third risk level, configuring the shortest environmental monitoring sampling cycle, the most stringent early warning threshold settings, and the highest quality control frequency. Specific parameter values ​​in each level of strategy template are pre-configured based on industry standards and expert experience, and can be adjusted and optimized according to actual operational results. Based on the matching relationship between the medical device batch risk level and the strategy template, the strategy template configuration is obtained, including specific numerical settings for various monitoring parameters.

[0038] Step 4.2: Dynamic adjustment and stabilization of strategies based on risk assessment results; Based on the cross-regional pollution spread risk assessment results, the strategy template configuration is dynamically adjusted. Specifically, when the risk assessment result indicates a high-risk level, a strategy upgrade assessment process is triggered. The system checks the duration of the high-risk state. If the duration reaches 30 minutes, the current strategy template is upgraded to a higher level. If the current level is already under strict control, it remains unchanged. The strategy upgrade means adopting higher-frequency monitoring and stricter warning thresholds. When the risk assessment result indicates a low-risk or medium-risk level, a strategy recovery assessment process is triggered. The system checks the duration of the reduced risk level. If the duration reaches two hours, the current strategy template is restored to the basic configuration corresponding to the risk level of the medical device batch. This avoids wasting resources by maintaining an overly strict monitoring strategy after the risk has been eliminated. By setting a time buffer for strategy adjustment and recovery, frequent strategy switching due to short-term fluctuations in risk assessment results is avoided, ensuring strategy stability. Based on the above dynamic adjustment logic, a stable strategy adjustment decision is obtained.

[0039] Step 4.3: Monitor and control policy configuration update execution and change log recording; Based on the stable policy adjustment decision, the system executes the actual configuration update of the monitoring and control policy. Specifically, when the policy adjustment decision indicates that a change in policy configuration is required, the system sends a policy update command to the edge computing gateways deployed in each monitoring area. After receiving the command, the edge computing gateway updates its local sensor sampling period configuration, controlling the sensors to perform data acquisition according to the new period; updates its local data reporting frequency configuration, adjusting the time interval for reporting data to the central server; and updates its local early warning threshold configuration, enabling the real-time early warning judgment on the edge side to adopt the new threshold standard. After the policy configuration update is completed, the system records a policy change log, including the change timestamp, the policy level before the change, the policy level after the change, and the reason for triggering the change, for subsequent policy effect analysis and audit traceability. Based on the above configuration update process, the currently effective monitoring and control policy configuration is obtained.

[0040] Furthermore, since the preset hierarchical strategy templates may not fully adapt to the specific needs of all real-world scenarios, a strategy parameter fine-tuning mechanism can be adopted to allow for small-scale adjustments based on the templates, aiming to improve the flexibility and applicability of strategy configuration. Specifically, based on the strategy template configuration, the system supports setting adjustment coefficients for specific regions or specific parameters; when historical monitoring data of a certain region shows that the environmental parameters of that region fluctuate significantly, a warning threshold tightening coefficient can be set for that region, making the actual warning threshold of that region more stringent than the template configuration value; when historical operating data of a certain type of sterilization equipment shows that the equipment's performance is stable, a monitoring frequency relaxation coefficient can be set for that equipment, reducing unnecessary monitoring costs while ensuring safety; the setting of adjustment coefficients requires authorized approval, and the basis for adjustment and the effective time range must be recorded.

[0041] Step 5: Collect the sterilization equipment operation process parameters according to the collection requirements of the monitoring and control strategy, integrate the process and inspection data according to sterilization batches, and obtain the disinfection and sterilization quality characteristic dataset; Specifically, the following steps are included: Step 5.1: Real-time collection of operating parameters of the sterilization equipment; Based on the data communication interfaces established with various sterilization equipment in the disinfection supply center, and in accordance with the data acquisition requirements determined by the monitoring and control strategy, equipment parameters during the sterilization process are collected in real time. Specifically, for pre-vacuum pressure steam sterilizers, the collected parameters include sterilization chamber temperature, sterilization chamber pressure, pre-vacuum pulse count, pre-vacuum pressure value, sterilization holding time, drying time, program stage identifier, and equipment alarm code; for low-temperature plasma sterilizers, the collected parameters include sterilization chamber pressure, hydrogen peroxide injection volume, plasma excitation power, sterilization cycle time, program stage identifier, and equipment alarm code. The data acquisition interval is determined according to the strategy configuration; under the strict control strategy, the highest acquisition frequency is used to obtain more refined process data. Based on the above data acquisition process, sterilization process parameter records are obtained, which are stored in the format of sterilization batch identifier, timestamp sequence, and parameter value sequence.

[0042] Step 5.2: Extract key quality characteristic parameters according to the standard stage. Based on the recorded sterilization process parameters, key quality characteristic parameters for each stage are extracted according to the standard stages of the sterilization procedure. For the pre-vacuum pressure steam sterilization procedure, it is divided into pre-vacuum stage, heating stage, sterilization stage, exhaust stage, and drying stage. In the pre-vacuum stage, the actual value of the pre-vacuum pulse count and the set value, as well as the vacuum level reached for each pulse, are extracted. In the heating stage, the time from program start to reaching the set sterilization temperature (heating time) and the average slope of the temperature rise curve (heating rate) are extracted. In the sterilization stage, the duration of temperature maintenance (constant temperature time), the maximum temperature fluctuation within the sterilization stage (temperature stability), and the maximum pressure fluctuation within the sterilization stage (pressure stability) are extracted. In the drying stage, the drying time and the item temperature at the end of the program are extracted. For the low-temperature plasma sterilization procedure, it is divided into vacuum stage, injection stage, diffusion stage, plasma stage, and ventilation stage, and key characteristic parameters for each stage are extracted. Based on the above feature extraction process, the quality characteristics of the staged sterilization process are obtained.

[0043] Step 5.3, Image acquisition and color change feature analysis of chemical indicator cards; Based on the image acquisition device installed at the discharge port of the sterilizer, image data of the chemical indicator card on the outer packaging is acquired after each batch of sterilization is completed, and image analysis is performed to extract color change features. Specifically, when the sterilizer completes a sterilization cycle and opens to discharge the material, the image acquisition device automatically captures a picture of the chemical indicator card located outside the package, obtaining raw image data. The raw image is preprocessed, including grayscale correction to compensate for differences in ambient brightness, perspective correction to compensate for shooting angle deviations, and region segmentation to extract the color-changing indicator area of ​​the chemical indicator card. Color feature analysis is performed on the extracted color-changing indicator area, converting the image from a red-green-blue color space to a hue-saturation-brightness color space, and calculating the average hue, average saturation, and average brightness values ​​of the color-changing area as color feature vectors. These color feature vectors are compared with a pre-established standard color-changing sample feature library, which includes the definition of feature vector ranges for qualified and unqualified color-changing states. Based on the Euclidean distance between the color feature vectors and the standard qualified color-changing features, a quantitative value for the degree of chemical indicator color change is calculated; a higher quantitative value indicates that the color change better meets the standard requirements. Based on the above image analysis process, the quantitative value for the degree of chemical indicator color change and the color-changing state determination result are obtained.

[0044] Step 5.4: Integration and correlation of sterilization batch quality data; Based on the phased sterilization process quality characteristics and the quantified values ​​of chemical indicator color change, data is integrated and correlated according to sterilization batches. Specifically, using the unique identifier of the sterilization batch as the association key, information such as the process quality characteristic parameters of each stage of the batch, the quantified values ​​of chemical indicator color change, sterilization start time, sterilization end time, sterilizer equipment identifier, and sterilization program type are integrated into a complete batch quality record. The instrument source risk level identifier corresponding to the batch is extracted from the data in step 1 and associated with the batch quality record, providing a basis for subsequent differentiated quality evaluation. Based on the above data integration process, a disinfection and sterilization quality characteristic dataset is obtained. This dataset is organized by sterilization batch and contains complete process characteristic and inspection characteristic information.

[0045] Furthermore, since the acquisition and recognition of chemical indicator card images may be affected by factors such as changes in ambient light and indicator card positional shifts, leading to a decrease in recognition accuracy, a multi-frame image fusion and adaptive threshold adjustment method can be used to improve recognition robustness. The aim is to reduce the interference of environmental factors on the chemical indicator recognition results. Specifically, the image acquisition device continuously captures multiple frames of images during each acquisition, and the color feature vectors of the multiple frames are averaged to reduce the impact of random noise in a single frame. The system periodically acquires images of a standard reference color card under the current environmental conditions, calculates the deviation between the actual acquired reference color card color features and the standard value, and uses this deviation as an environmental compensation coefficient to correct the color features of the chemical indicator card. When the environmental compensation coefficient exceeds a reasonable range, the system prompts that the status of the acquisition device needs to be checked or the installation position adjusted.

[0046] Step 6: Construct a historical quality benchmark model for similar batches, identify abnormal quality patterns based on the disinfection and sterilization quality feature dataset, trace the associated environmental risk factors, and obtain intelligent evaluation results of sterilization quality. Specifically, the following steps are included: Step 6.1: Construct a historical quality benchmark model for similar batches; Based on a historically accumulated dataset of disinfection and sterilization quality characteristics, historical batches are categorized according to three dimensions: sterilization equipment type, sterilization procedure type, and instrument risk level. A historical quality benchmark model is constructed for each category. Specifically, for scenarios where pre-vacuum pressure steam sterilizers process instruments of the second risk level, all qualified batch records for this equipment type, procedure type, and risk level are selected from historical data, requiring a minimum of one hundred batches to ensure statistical validity. For the selected historical batches, statistical benchmark values ​​for each key quality characteristic parameter are calculated, including the mode of pre-vacuum pulse count, median heating time, median isothermal time, 75th quartile of temperature fluctuation, 75th quartile of pressure stability, median drying time, and median chemical indicator color change quantification value. The standard deviation of each characteristic parameter is also calculated for subsequent deviation assessment. The benchmark values ​​and standard deviations of each characteristic parameter are organized into a historical quality benchmark model for that category. For categories with insufficient historical samples, benchmark models from similar categories are used as substitutes and marked as reference benchmarks. Based on the above construction process, a historical quality benchmark model library covering various sterilization scenarios is obtained.

[0047] Step 6.2: Extraction of multi-dimensional quality feature vectors and calculation of deviation for the current batch; Based on the disinfection and sterilization quality characteristic dataset of the current batch, multi-dimensional quality feature vectors are extracted, and the deviation from historical benchmarks is calculated. Specifically, according to the equipment type, program type, and risk level identifier of the current batch, the corresponding benchmark model is matched from the historical quality benchmark model library; the measured values ​​of each key quality characteristic parameter of the current batch are extracted, including the number of pre-vacuum pulses, heating time, isothermal time, temperature fluctuation, pressure stability, drying time, and chemical indicator color change quantification value, and organized into the quality feature vector of the current batch; for each dimension in the feature vector, the difference between the current measured value and the historical benchmark value is calculated, and divided by the historical standard deviation of that dimension to obtain the standardized deviation; the absolute value of the standardized deviation reflects the degree to which the current batch deviates from the historical normal level in that dimension, and the larger the absolute value, the more obvious the deviation; the square root of the sum of the squares of the standardized deviations of each dimension is used to obtain the comprehensive deviation index, which comprehensively reflects the overall difference between the current batch and the historical benchmark; based on the above calculation process, the deviation of each dimension and the comprehensive deviation of the current batch are obtained.

[0048] Step 6.3, Abnormal quality pattern identification and classification; Based on the deviation of each dimension of the current batch, abnormal quality patterns are identified and classified. Specifically, a single-dimensional deviation abnormality threshold is set to twice the standard deviation. When the absolute value of the standardized deviation of a certain dimension exceeds this threshold, it is determined that there is a significant deviation in that dimension. A comprehensive deviation abnormality threshold is set. When the comprehensive deviation exceeds this threshold, the overall quality pattern of the current batch is determined to be abnormal. Based on the combined characteristics of the deviation dimensions, the abnormal quality pattern type is identified: when both the heating time and the isothermal time are excessively long, it is identified as a heating efficiency decline pattern, indicating that the equipment heating system may have performance degradation; when both temperature fluctuation and pressure stability are significantly deviated, it is identified as a process control instability pattern, indicating that the equipment control system may have a fault; when the chemical indicator color change quantification value is significantly low but the process parameter deviation is not significant, it is identified as a chemical indicator abnormality pattern, indicating that there may be a quality problem with the indicator card or improper placement; when only a few dimensions deviate slightly and the comprehensive deviation does not exceed the threshold, it is identified as a normal fluctuation pattern. Based on the above identification process, the abnormal quality pattern classification results are obtained.

[0049] Step 6.4, Source tracing and correlation analysis of environmental risk factors; Based on the classification results of abnormal quality patterns, the environmental risk factors corresponding to the current batch and time period are traced and associated to establish the correlation between quality anomalies and environmental factors. Specifically, the sterilization start and end times of the current batch are extracted. The environmental parameter trend analysis results from step 2 are used to query whether there are environmental anomaly markers during this time period, including abnormal temperature and humidity trends, pressure differential decay trends, etc. The risk level of this time period is queried from the cross-regional pollution diffusion risk assessment results from step 3 to determine whether it is in a medium-risk or high-risk state. When an unstable process control pattern is identified and there are significant fluctuations in environmental parameters during this time period, the correlation between quality anomalies and environmental fluctuations is established, and environmental factors are marked as possible influencing factors. When a heating efficiency decline pattern is identified and multiple batches of the equipment have the same pattern recently, the correlation between quality anomalies and equipment aging is established, and equipment factors are marked as the main influencing factors. When a chemical indicator anomaly pattern is identified and the environmental humidity exceeds the standard during this time period, the correlation between quality anomalies and environmental humidity is established, and environmental humidity is marked as a possible influencing factor. Based on the above source tracing analysis, the anomaly source factor identifiers are obtained, including the associated environmental risk events, equipment status information, and possible cause inferences.

[0050] Step 6.5, Intelligent quality assessment result generation and differential processing suggestions; Based on the current batch's overall deviation, abnormal quality pattern classification, and abnormal source identification factors, intelligent quality evaluation results and differentiated processing suggestions are generated. Specifically, when the overall deviation exceeds the severe anomaly threshold, or a severe deviation in a key dimension is identified, or the chemical indicator color change quantification value is lower than the acceptable threshold, the batch is deemed unqualified, and a mandatory reprocessing recommendation is generated, specifying the specific reasons for the unqualification and the possible influencing factors traced back to the source. When the overall deviation is within the slightly abnormal range, and all key dimensions are within the acceptable range, the batch is deemed qualified but requires attention, and a recommendation to strengthen subsequent monitoring is generated, specifying the specific dimensions of the deviation and the recommended focus. When the overall deviation is within the normal range, and all dimensions are close to the historical benchmark, the batch is deemed excellent. For batches identified as being related to equipment factors, equipment maintenance and inspection prompts are added to the processing recommendations. For batches identified as being related to environmental factors, environmental control enhancement prompts are added to the processing recommendations. Based on the above evaluation process, a sterilization quality intelligent evaluation result with an anomaly traceability identifier is obtained, including complete information such as overall deviation, details of deviation in each dimension, anomaly pattern type, traceability factor identifier, quality level determination, and differentiated processing recommendations.

[0051] Furthermore, since the historical quality benchmark model is built based on historical data over a fixed period, the historical benchmark may no longer be applicable to the current state after equipment maintenance or process parameter adjustments. A dynamic benchmark model update mechanism can be used to maintain the timeliness of the benchmark, aiming to ensure that quality evaluation is always based on the latest normal operating conditions for comparison. Specifically, the system continuously monitors the quality data of the most recent 30 batches for each category. When the pass rate of the most recent 30 batches remains above 95% and the average comprehensive deviation is lower than the historical benchmark, the equipment status is determined to be optimized, triggering the benchmark model update process. During the benchmark model update, the most recent qualified batch data is included in the benchmark calculation, and the statistical benchmark values ​​and standard deviations of each characteristic parameter are recalculated. The updated benchmark model replaces the original model, and the update timestamp and triggering reason are recorded. For cases where the benchmark value changes by more than 20% before and after the update, a benchmark change prompt is generated for quality management personnel to review and confirm.

[0052] Step 7: Generate differentiated processing suggestions and quality early warning messages based on the intelligent evaluation results of sterilization quality, determine the early warning level and associate it with traceability information for hierarchical push, and obtain hierarchical early warning push results with traceability information; Specifically, the following steps are included: Step 7.1: Generating and intelligently classifying early warning messages that integrate traceability information; Based on various analysis and assessment results and source tracing information, early warning messages are generated and intelligently classified. Specifically, environmental early warning messages are generated by extracting parameter information marked as having abnormal trends or predicted to exceed limits from environmental parameter trend analysis results. The message content includes the type of abnormal parameter, its location, current value, predicted trend, and recommended measures. Pollution risk early warning messages are generated by extracting assessment information with medium or high risk levels from cross-regional pollution diffusion risk assessment results. The message content includes the risk level, main risk factors, affected areas, and recommended measures. Batch information judged as unqualified or requiring attention is extracted from sterilization quality intelligent evaluation results with abnormal source tracing labels. Quality early warning messages are generated by extracting batch information, comprehensive deviation, abnormal pattern type, source tracing factor label, and differentiated treatment recommendations. For quality early warning messages, the warning level is intelligently determined by integrating traced environmental risk factors: when a batch is determined to be non-compliant and its traceability is linked to a high-risk environmental event, it is designated as an emergency warning, indicating a strong correlation between the quality problem and environmental anomaly requiring immediate action; when a batch is determined to be non-compliant but not linked to obvious environmental risks, or when a batch is determined to be of concern and linked to a medium-risk environmental event, it is designated as a warning, indicating the need for timely attention and handling; when a batch is determined to be of concern but not linked to environmental risks, or only has a slight deviation, it is designated as a reminder warning, indicating the need for recording and tracking but not immediate intervention. For environmental and pollution risk warnings, the warning level is determined according to preset grading rules. Based on the above warning generation process, various warning message sets, their level labels, and traceability information are obtained.

[0053] Step 7.2: Early warning messages related to traceability information are pushed out through different channels. Based on the level markings, traceability information of various warning messages, and the configured preset push rules, perform the multi-channel push of warning messages. Specifically, the push targets and push channels for warnings of each level are defined in the warning push rule configuration: Tips-level warnings are pushed to the monitoring large screen and the duty workstations in the sterile supply center and are displayed in the form of pop-up windows or scrolling messages, and the message content includes basic abnormal information; Warnings-level warnings are not only pushed to the monitoring large screen and workstations but also to the mobile terminals of relevant post responsible persons and are delivered through in-app notifications or WeChat enterprise account messages, and the message content adds a brief description of traceability factors; Emergency-level warnings, in addition to the above channels, also trigger the audible and visual alarm devices at the sterile supply center site and send SMS notifications to the center head and the hospital infection management department, and the message content includes the complete traceability analysis results and associated environmental risk event information, facilitating the recipients to quickly understand the root cause of the problem and take targeted measures. For quality warning messages, the abnormal quality mode type and the identified environmental risk factors traced are appended to the push content, enabling the recipients to understand the possible causes of quality problems; For warning messages of the same type, a ten-minute push deduplication time window is set, and warnings with the same content are not pushed repeatedly within the deduplication time window to avoid message bombing. Based on the above push execution process, a hierarchical warning push result with traceability information is obtained, and information such as the push time, push channel, push target, delivery status, and traceability information summary of each warning is recorded.

[0054] Step 7.3, construction of a multi-dimensional visual monitoring interface; Based on various monitoring data, analysis results, and warning information, construct a multi-dimensional visual monitoring interface. Specifically, the visual monitoring interface includes the following main components: The overall view component of the regional environmental status displays the positional relationships of the functional areas in the sterile supply center in the form of a floor plan layout, and the current temperature, humidity, differential pressure, and other key parameter values are displayed at the positions of each area, and the parameter status (normal, warning, alarm) is represented by color markings, and clicking on the area can expand the parameter trend curve of that area; The risk assessment heat map component displays the distribution of pollution diffusion risk levels between regions in the form of a heat map, and the depth of the color represents the level of risk, and it supports switching to view the risk distribution changes at each historical time point; The operation monitoring component of sterilization equipment displays the current operating status, the program information being executed, and the quality scores of the most recent batches of each sterilizer in the form of equipment cards, and clicking on the equipment card can view the detailed process parameter curve; The quality statistical analysis component displays the number of sterilization batches and the qualification rate of each recent day in the form of bar charts, the change trend of the comprehensive quality score in the form of line charts, and the distribution of unqualified reasons in the form of pie charts; The warning message list component displays the most recent warning messages in reverse chronological order, supports filtering by level and type, and clicking on the message can view the details and processing records. Each visual component supports custom layout and refresh frequency settings to adapt to the viewing habits of different post personnel. Based on the above visual construction process, a visual monitoring interface is obtained.

[0055] Furthermore, since a fixed-layout visual interface may not meet the differentiated information needs of various usage scenarios, a role-based view configuration method can be adopted to provide customized information displays for personnel in different positions, aiming to improve the relevance and efficiency of information display. Specifically, the system predefines several role-based view configurations: the quality administrator view focuses on displaying quality statistical analysis data and details of non-conforming batches; the equipment administrator view focuses on displaying equipment operating status and maintenance warnings; and the on-duty personnel view focuses on displaying real-time environmental parameters and current warning messages. After logging into the system, users automatically load the corresponding default view configuration based on their role, while also supporting users to customize and adjust component layouts and parameters of interest within their authorized scope. Users' personalized configurations are saved in their user configuration files and automatically restored upon the next login.

[0056] A smart monitoring and management system for a disinfection supply area is used to execute the aforementioned smart monitoring and management method for a disinfection supply area, such as... Figure 3 As shown, it includes: The environmental perception data acquisition module is used to acquire environmental parameters and instrument batch source identification of the disinfection supply center, determine the risk level, and obtain an environmental perception dataset with risk identification. The environmental parameter trend analysis module, based on an environmental perception dataset with risk labels, identifies abnormal trends and predicts future environmental parameter values. Based on the prediction results, it identifies risks of exceeding limits and obtains the environmental parameter trend analysis results. The cross-regional pollution diffusion risk assessment module, based on the results of environmental parameter trend analysis, quantitatively assesses the severity of cross-regional pollution diffusion and obtains the cross-regional pollution diffusion risk assessment results. The monitoring and control strategy configuration module matches monitoring strategies based on the cross-regional pollution spread risk assessment results, dynamically adjusts them according to changes in risk level, issues update instructions and records logs to obtain the monitoring and control strategy. The disinfection and sterilization quality data acquisition module is used to collect the operating parameters of the sterilization equipment according to the acquisition requirements of the monitoring and control strategy, and integrate the process and inspection data by sterilization batch to obtain the disinfection and sterilization quality characteristic dataset. The sterilization quality evaluation module is used to build a historical quality benchmark model for similar batches. Based on the disinfection and sterilization quality feature dataset, it identifies abnormal quality patterns and traces the associated environmental risk factors to obtain intelligent sterilization quality evaluation results. The tiered early warning push module generates differentiated processing suggestions and quality early warning messages based on the intelligent evaluation results of sterilization quality, determines the early warning level and associates it with traceability information for tiered push, resulting in tiered early warning push results with traceability information.

[0057] In one embodiment of the present invention, a specific example is provided: A tertiary-level Class A general hospital's disinfection supply center has adopted the intelligent monitoring and management system of this invention. The center processes approximately 3,000 sets of reusable medical devices daily, and is divided into three functional areas according to national standards: a decontamination area, an inspection, packaging, and sterilization area, and a sterile item storage area. The center is equipped with three pre-vacuum pressure steam sterilizers and two low-temperature plasma sterilizers, serving the instrument disinfection and sterilization needs of more than fifty clinical departments throughout the hospital.

[0058] During the system deployment phase, four temperature and humidity sensors and two ammonia concentration sensors were deployed in the decontamination zone; six temperature and humidity sensors and two dust particle counters were deployed in the packaging and sterilization zone; and eight temperature and humidity sensors and three dust particle counters were deployed in the sterile goods storage zone. Six differential pressure sensors and eight door magnetic sensors were installed at the transfer windows and buffer zones at the boundaries of each zone. Data communication interfaces were established with the five sterilization units, and a chemical indicator card image acquisition device was installed at the discharge port of each sterilizer.

[0059] Table 1 shows an example of environmental monitoring data collected by the system for a certain period of time: Table 1: Example of environmental monitoring data for a certain period;

[0060] Table 2 shows an example of sterilization process data collected by the system for a certain batch: Table 2: Example of sterilization process data for a certain batch;

[0061] By applying the method of this invention, the disinfection supply center has achieved continuous automatic monitoring of environmental parameters around the clock, eliminating monitoring gaps at night and on holidays; it has achieved risk-based classification management based on the source of medical devices, automatically triggering enhanced monitoring modes for devices sent by the infectious disease department; it has achieved objective quantitative evaluation of sterilization quality, eliminating subjective differences in manual interpretation of chemical indicator cards; and it has achieved timely early warning and graded push notifications for abnormal situations, ensuring that relevant personnel can be informed and handle abnormalities as soon as possible.

[0062] The embodiments of the present invention have been described above. However, the embodiments are not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make more equivalent embodiments under the guidance of the present embodiments, and all of them are within the protection scope of the present embodiments.

Claims

1. A method for intelligent monitoring and management of a disinfection supply area, characterized in that, Includes the following steps: Obtain environmental parameters and equipment batch source identification from the disinfection supply center, determine the risk level, and obtain an environmental perception dataset with risk identification; Based on the environmental perception dataset with risk labels, abnormal trends are identified and future environmental parameter values ​​are predicted. Based on the prediction results, the risk of exceeding limits is identified, and the environmental parameter trend analysis results are obtained. Based on the results of environmental parameter trend analysis, the severity of cross-regional pollution diffusion is quantitatively assessed, and the risk assessment results of cross-regional pollution diffusion are obtained. Based on the cross-regional pollution spread risk assessment results, the monitoring strategy is matched, dynamically adjusted according to changes in risk level, and update instructions are issued and logs are recorded to obtain the monitoring and control strategy. Collect sterilization equipment operation parameters according to the collection requirements of the monitoring and control strategy, integrate process and inspection data according to sterilization batches, and obtain a disinfection and sterilization quality characteristic dataset; Construct a historical quality benchmark model for similar batches, identify abnormal quality patterns and trace the associated environmental risk factors based on a disinfection and sterilization quality feature dataset, and obtain intelligent evaluation results for sterilization quality. Based on the intelligent evaluation results of sterilization quality, differentiated processing suggestions and quality early warning messages are generated. The early warning level is determined and traceability information is associated to push the messages in a graded manner, resulting in graded early warning push results with traceability information.

2. The intelligent monitoring and management method for a disinfection supply area according to claim 1, characterized in that, The steps for obtaining the environmental awareness dataset with risk labels include: Multiple types of environmental sensors are deployed in each functional area, and each sensor collects data according to a preset sampling period; Pressure differential monitoring points are set up at the boundary between adjacent areas to collect time-series data of pressure differential gradient between areas; Obtain the unique identification code of the medical device batch by using a barcode scanning device or radio frequency identification reader, and then query the department risk level mapping table to determine the risk level of the medical device batch. Based on the arrival timestamp of the medical device batch, environmental parameter data and differential pressure gradient data within a preset time window are extracted and integrated with the risk level identifier of the medical device batch to obtain an environmental perception dataset with risk identifier.

3. The intelligent monitoring and management method for a disinfection supply area according to claim 1, characterized in that, The steps for obtaining the environmental parameter trend analysis results include: The sliding window statistical method is used to extract statistical features of time series data of environmental parameters, including window mean, window standard deviation, window range and slope of linear change; Trend anomaly judgment is based on an expert rule base, which includes slope anomaly rules, fluctuation anomaly rules, continuous offset rules, and differential pressure attenuation rules. Construct a time-series prediction model for environmental parameters to predict parameter values ​​for future time periods; By comparing the predicted value sequence with the control limits, identifying the predicted time points of exceeding the limits, calculating the expected lead time of exceeding the limits, and combining the predicted exceeding information with the rule-based anomaly markers, the trend analysis results of environmental parameters are obtained.

4. The intelligent monitoring and management method for a disinfection supply area according to claim 1, characterized in that, The steps for obtaining the cross-regional pollution diffusion risk assessment results include: Based on the monitoring data of the switch status sensor and the door magnetic sensor, the number of times the transfer window is opened, the opening duration and the number of people passing through within the evaluation time window are statistically evaluated. After normalization, the weighted sum is used to obtain the inter-regional transfer activity intensity index. Analyze time-series data of differential pressure gradient to identify periods when differential pressure values ​​are below the safe threshold and calculate the abnormal exposure rate of differential pressure. Anomaly marker information of the upstream region is extracted from the results of environmental parameter trend analysis, and the anomaly risk score of the upstream region is obtained by summing them. The cross-regional pollution diffusion comprehensive risk score is calculated and risk levels are classified by weighting and integrating the inter-regional transmission activity intensity index, pressure difference abnormal exposure rate index, upstream regional abnormal risk score and device comprehensive risk level label.

5. The intelligent monitoring and management method for a disinfection supply area according to claim 1, characterized in that, The steps for obtaining the monitoring and control strategy configuration include: Based on the comprehensive risk level identification of medical device batches, the corresponding basic monitoring strategy is matched from the tiered strategy template library, which includes basic strategy templates, enhanced strategy templates, and strict control strategy templates. When the risk assessment result indicates a high risk level and the duration reaches the first duration threshold, the current strategy template will be upgraded to a higher level. When the risk level decreases and the duration reaches the second duration threshold, the strategy template will be restored to the basic configuration corresponding to the risk level of the device batch. Send policy update instructions to the edge computing gateway to update the sensor sampling period configuration, data reporting frequency configuration, and early warning threshold configuration, and record the policy change log.

6. The intelligent monitoring and management method for a disinfection supply area according to claim 1, characterized in that, The steps for obtaining the disinfection and sterilization quality feature dataset include: By establishing a data communication interface with the sterilization equipment and configuring the data acquisition requirements according to the monitoring and control strategy, the equipment parameters during the sterilization process are collected in real time. Based on the standard stages of the sterilization process, key quality characteristic parameters for each stage were extracted. Image data of chemical indicator cards are acquired through an image acquisition device, and image preprocessing and color feature analysis are performed to calculate the quantitative value of the degree of color change of the chemical indicator. Using the unique identifier of the sterilization batch as the association key, the process quality characteristic parameters of each stage, the quantitative value of the degree of chemical indicator color change, and the risk level identifier of the source of the device are integrated to obtain the disinfection and sterilization quality characteristic dataset.

7. The intelligent monitoring and management method for a disinfection supply area according to claim 1, characterized in that, The specific steps for obtaining the intelligent evaluation results of sterilization quality include: Historical qualified batches were classified according to sterilization equipment type, sterilization procedure type, and instrument risk level. Statistical benchmark values ​​and standard deviations of each key quality characteristic parameter were calculated, and a historical quality benchmark model for the same batch was constructed. Extract the measured values ​​of each key quality characteristic parameter of the current batch, calculate the standardized deviation of each dimension from the historical benchmark value, and calculate the comprehensive deviation index; The abnormal quality pattern types are identified based on the combined features of the deviation dimension, including heating efficiency reduction mode, process control instability mode, and chemical indicator abnormal mode. Extract the trend analysis results of environmental parameters and the risk assessment results of cross-regional pollution diffusion for the corresponding time period of the current batch, establish the source-tracing correlation between quality anomalies and environmental risk factors, and generate anomaly source-tracing factor identifiers; Based on the comprehensive deviation, abnormal quality pattern classification, and abnormal source identification, the batch quality level is determined and differentiated processing suggestions are generated.

8. The intelligent monitoring and management method for a disinfection supply area according to claim 1, characterized in that, The steps for pushing tiered early warning results with traceability information include: Abnormal information is extracted from environmental parameter trend analysis results, cross-regional pollution diffusion risk assessment results, and sterilization quality intelligent evaluation results with abnormal source tracing identifiers to generate environmental early warning messages, pollution risk early warning messages, and quality early warning messages; For quality warning messages, the warning level is determined by integrating the environmental risk factors traced back to their source; Based on the warning level and push rule configuration, the warning message is pushed to different channels, and the abnormal quality mode type and the environmental risk factor identifier traced to the source are attached to the push content. Set a time window for deduplication of push notifications, and record the push time, push channel, delivery status, and source tracing information summary of the alerts.

9. The intelligent monitoring and management method for a disinfection supply area according to claim 3, characterized in that, The construction of the environmental parameter time series prediction model also includes: The system continuously monitors the prediction error of the prediction model. When the average prediction error within a preset number of prediction periods exceeds the error tolerance threshold, it triggers an incremental update process for the model parameters. The mini-batch gradient descent method is used to fine-tune the model parameters using measured data, and the adjustment range is controlled by a preset learning rate parameter. The updated model parameters replace the original parameters, and the timestamp and triggering reason of the model update are recorded to form a model evolution log.

10. An intelligent monitoring and management system for a disinfection supply area, characterized in that, A method for intelligent monitoring and management of a disinfection supply area as described in any one of claims 1-9, comprising: The environmental perception data acquisition module is used to acquire environmental parameters and instrument batch source identification of the disinfection supply center, determine the risk level, and obtain an environmental perception dataset with risk identification. The environmental parameter trend analysis module, based on an environmental perception dataset with risk labels, identifies abnormal trends and predicts future environmental parameter values. Based on the prediction results, it identifies risks of exceeding limits and obtains the environmental parameter trend analysis results. The cross-regional pollution diffusion risk assessment module, based on the results of environmental parameter trend analysis, quantitatively assesses the severity of cross-regional pollution diffusion and obtains the cross-regional pollution diffusion risk assessment results. The monitoring and control strategy configuration module matches monitoring strategies based on the cross-regional pollution spread risk assessment results, dynamically adjusts them according to changes in risk level, issues update instructions and records logs to obtain the monitoring and control strategy. The disinfection and sterilization quality data acquisition module is used to collect the operating parameters of the sterilization equipment according to the acquisition requirements of the monitoring and control strategy, and integrate the process and inspection data by sterilization batch to obtain the disinfection and sterilization quality characteristic dataset. The sterilization quality evaluation module is used to build a historical quality benchmark model for similar batches. Based on the disinfection and sterilization quality feature dataset, it identifies abnormal quality patterns and traces the associated environmental risk factors to obtain intelligent sterilization quality evaluation results. The tiered early warning push module generates differentiated processing suggestions and quality early warning messages based on the intelligent evaluation results of sterilization quality, determines the early warning level and associates it with traceability information for tiered push, resulting in tiered early warning push results with traceability information.

Images