Multi-point triggering biological safety monitoring method and device, computer equipment and storage medium

Abstract

The embodiment of the invention discloses a multi-point triggering biological safety monitoring method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring diagnosis and treatment data of a patient, and performing data processing to obtain individual case data of the patient; inputting the individual case data of the patient into an early warning model of various infections and infectious diseases for infection prediction to obtain a prediction result; inputting the prediction result into a biological safety risk early warning model to distinguish the types of the suspected infection part and the disease part of the patient so as to obtain a distinguishing result; performing dynamic early warning on the distinguishing result in combination with an established incremental early warning mechanism to obtain an early warning result; and intervention information is sent to a corresponding platform based on the early warning result. By implementing the method provided by the embodiment of the invention, the defects in the aspects of data integration, risk analysis, dynamic monitoring, early warning mechanism flexibility, information processing efficiency and the like in the prior art can be overcome, and more efficient and accurate hospital biological safety monitoring is realized.

Description

Technical Field

[0001] The present invention relates to a medical information processing method, and more specifically to a multi-point triggered biosafety monitoring method, device, computer equipment and storage medium. Background Art

[0002] In modern hospital infection and infectious disease management, the application of information monitoring systems has significantly improved data processing and monitoring capabilities, but it still faces many challenges. The current technical solutions have the problem of incomplete data integration. The information between different systems and platforms cannot be effectively integrated, making it difficult to comprehensively summarize the diagnosis and treatment data of outpatients and inpatients. This not only affects the integrity of the information, but may also cause redundancy, errors and inconsistencies during data cleaning and conversion.

[0003] Existing systems usually rely on basic statistical analysis or fixed rules, and fail to deeply explore the potential complex patterns in the data, thus limiting the accurate identification of high-risk risk factors. Early warning models are often more general and lack personalized risk predictions for specific types of hospital infections or individual patients. In terms of monitoring, traditional systems mainly rely on regularly updated static data, lack dynamic monitoring of patients' real-time status, and the setting of monitoring checkpoints may not be comprehensive enough, resulting in the inability to capture risk changes in a timely manner.

[0004] In terms of early warning mechanisms, periodic warnings often fail to quickly reflect the rapid changes in patients' conditions, thus affecting response capabilities. At the same time, insufficient data integration of patients with similar risks in the region affects the prevention of cluster events and the spread of infectious diseases. Information transmission and processing in the existing system often have delays, which reduces the efficiency of medical staff's rapid response, and the lack of a real-time feedback mechanism further weakens the treatment effect. These problems jointly restrict the efficiency and accuracy of hospital infection and infectious disease management.

[0005] Therefore, it is necessary to design a new method to overcome the shortcomings of existing technologies in data integration, risk analysis, dynamic monitoring, flexibility of early warning mechanisms, and information processing efficiency, so as to achieve more efficient and accurate hospital biosafety monitoring. Summary of the invention

[0006] The purpose of the present invention is to overcome the defects of the prior art and provide a multi-point triggered biosafety monitoring method, device, computer equipment and storage medium.

[0007] To achieve the above object, the present invention adopts the following technical solution: a multi-point triggered biosafety monitoring method, comprising: Obtain the patient's diagnosis and treatment data and process the data to obtain the patient's case data; Inputting the patient case data into early warning models of multiple infections and infectious diseases to predict infection, so as to obtain prediction results; Inputting the prediction results into a biosafety risk early warning model to distinguish the types of suspected infection sites and disease sites of patients to obtain a distinction result; Performing dynamic warning on the differentiation result in combination with the established incremental warning mechanism to obtain a warning result; Send intervention information to the corresponding platform based on the early warning result.

[0008] A further technical solution is: the patient's diagnosis and treatment data is obtained and processed to obtain the patient's case data, including: Obtain the patient's diagnosis and treatment data, and convert the diagnosis and treatment data into a unified data format to obtain the patient's case data, wherein the patient's case data includes the patient's basic information and the diagnosis and treatment data generated during outpatient and hospitalization.

[0009] The further technical solution is: the training process of the early warning model of the plurality of infections and infectious diseases includes: Categorize historical case data into infection cases, infectious disease cases, and healthy patient types; The historical case data corresponding to the classified infection cases and infectious disease cases are divided according to the infection site to obtain a sample set; The sample set is used to train the BP neural network algorithm to obtain early warning models for various infections and infectious diseases.

[0010] The further technical solution is: the training process of the biosafety risk early warning model includes: Extracting the diagnosis and treatment data related to the infection site and related diseases from the historical case data, and performing data preprocessing to obtain preprocessing results; Selecting features related to the infection site in the pretreatment results and standardizing them to obtain standardized results; A logistic regression model was initially trained for each infection type in the standardized results, and the regression coefficient of each feature was calculated to assess its importance for the specific infection type; According to the regression coefficients, key features affecting various types of infection are screened out; Constructing a feature matrix according to the key features, and constructing a training set and a test set; The training set is used to train a random forest model, and multiple decision trees are constructed for classification prediction to obtain a biosafety risk early warning model.

[0011] A further technical solution is: the differentiation result includes a risk assessment value of the patient.

[0012] Its further technical solution is: the incremental early warning mechanism includes real-time data detection and recalculation of the patient's risk assessment value.

[0013] A further technical solution is: the intervention information includes outbreak warning information issued when the number of patients at the same biosafety risk during the screening period exceeds the first transmission threshold.

[0014] The present invention also provides a multi-point triggered biosafety monitoring device, comprising: A case data acquisition unit, used to acquire the patient's diagnosis and treatment data and perform data processing to obtain the patient's case data; A prediction unit, used for inputting the patient case data into early warning models of multiple infections and infectious diseases to perform infection prediction to obtain prediction results; A differentiation unit, used for inputting the prediction result into the biosafety risk early warning model to differentiate the types of suspected infection sites and disease sites of the patient to obtain a differentiation result; A dynamic early warning unit, used for dynamically warning the incremental early warning mechanism established in combination with the differentiation result to obtain an early warning result; A sending unit is used to send intervention information to a corresponding platform based on the early warning result.

[0015] The present invention further provides a computer device, which includes a memory and a processor. The memory stores a computer program, and the processor implements the above method when executing the computer program.

[0016] The present invention also provides a storage medium, wherein the storage medium stores a computer program, and the computer program implements the above method when executed by a processor.

[0017] Compared with the prior art, the present invention has the following beneficial effects: the present invention obtains patient diagnosis and treatment data and processes and generates individual case data, inputs multiple infection prediction models to obtain prediction results, applies the prediction results to the biosafety risk warning model, distinguishes suspected infection sites and disease types, combines the incremental warning mechanism to dynamically generate warning results, and sends intervention information to relevant platforms. This overcomes the shortcomings of the prior art in data integration, risk analysis, dynamic monitoring, flexibility of warning mechanisms, and information processing efficiency, and achieves more efficient and accurate hospital biosafety monitoring.

[0018] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0020] Figure 1 A schematic diagram of an application scenario of the multi-point triggered biosafety monitoring method provided by an embodiment of the present invention; Figure 2 A schematic diagram of a process flow of a multi-point triggered biosafety monitoring method provided by an embodiment of the present invention; Figure 3 A schematic block diagram of a multi-point triggered biosafety monitoring device provided by an embodiment of the present invention; Figure 4 A schematic block diagram of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0023] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0024] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0025] See also Figure 1 and Figure 2 , Figure 1 Schematic diagram of application scenarios of the multi-point triggered biosafety monitoring method provided in an embodiment of the present invention. Figure 2A schematic flow chart of a multi-point triggered biosafety monitoring method provided in an embodiment of the present invention. The multi-point triggered biosafety monitoring method is applied in a server. The server interacts with the terminal to automatically extract the patient's hospital infection (infectious disease) related diagnosis and treatment data from the hospital business system; the collected data is integrated into individual patient cases for subsequent analysis and monitoring; a targeted biosafety risk warning model is established based on specific problems in the clinical diagnosis and treatment process (such as surgical site infection); the model is designed around common infection risks to ensure effective management and control; the monitoring results are updated in real time according to the daily changes in the patient's condition; check whether a warning has been issued for the case; if it has been issued, further determine whether this warning is the same as the previous risk factor; if the risk is the same, confirm whether the warning has been processed: unprocessed: remind management personnel to take measures as soon as possible; processed: no longer send warnings repeatedly. The incremental warning mechanism reduces the system burden and improves the response speed and accuracy; through continuous monitoring and instant feedback, ensure that the medical team can respond quickly to potential risks; through comprehensive data integration and intelligent warning mechanisms, help hospitals effectively control infectious disease risks and improve patient safety and medical quality. , Figure 2 FIG. 1 is a flow chart of a multi-point triggering biosafety monitoring method provided by an embodiment of the present invention. Figure 2 As shown, the method includes the following steps S110 to S150.

[0026] S110, obtaining the patient's diagnosis and treatment data, and performing data processing to obtain the patient's case data.

[0027] In this embodiment, data is obtained from multiple data sources such as hospital information system (HIS), laboratory information management system (LIS), imaging information system (RIS), electronic medical record (EMR) and picture archiving and communication system (PACS). Real-time data is obtained through various interface forms (such as API, database connection, etc.).

[0028] Establish a unified data conversion module to convert data from different sources into a specific unified format; form a consistent data structure to facilitate subsequent data processing and analysis.

[0029] Patient case data includes: Basic information: patient name, age, gender, case identification ID and other basic information.

[0030] Outpatient diagnosis and treatment data: including outpatient departments, consultation time, outpatient diagnosis, test application, examination application, doctor's orders, test results and examination results, etc.

[0031] Inpatient diagnosis and treatment data: clinical data such as physical signs (temperature, pulse, respiratory rate, white blood cell concentration, etc.), surgery-related information, medication information, etc. Inpatient process data include inpatient wards, beds, hospital stays, transfer information, inpatient diagnosis, test and examination results, bacterial culture results, etc.

[0032] Collect patient infection and infectious disease-related data from various outpatient and inpatient wards to form complete patient cases; update case data regularly or in real time to ensure the timeliness and accuracy of the data and support clinical decision-making.

[0033] By integrating medical data from different sources and establishing standardized patient case information, hospitals can efficiently manage patients' infection and infectious disease-related data, thereby improving the quality and safety of medical services.

[0034] S120, inputting the patient case data into early warning models of various infections and infectious diseases to perform infection prediction to obtain prediction results.

[0035] In this embodiment, the prediction result refers to the disease type corresponding to the patient case data and the corresponding probability value.

[0036] Specifically, the training process of the early warning model for the multiple infections and infectious diseases includes: Categorize historical case data into infection cases, infectious disease cases, and healthy patient types; The historical case data corresponding to the classified infection cases and infectious disease cases are divided according to the infection site to obtain a sample set; The sample set is used to train the BP neural network algorithm to obtain early warning models for various infections and infectious diseases.

[0037] In this embodiment, the types of infection cases include: According to the site of infection, it is divided into: respiratory system hospital infection, surgical site hospital infection, urinary system hospital infection, blood system hospital infection, and skin and soft tissue hospital infection.

[0038] Types of infectious disease cases include: According to the location of the disease, it is divided into: respiratory system infectious diseases, digestive system infectious diseases, reproductive system infectious diseases, and nervous system infectious diseases; Healthy patient cases: not included in the infection or infectious disease classification.

[0039] First, extract the process diagnosis and treatment data from historical cases, and perform cleaning and normalization; divide the dataset into a training set, a validation set, and a test set. Usually, 70% is used for training, 15% for validation, and 15% for testing. Set the number of nodes in the input layer of the BP neural network model to N (N is the number of features), select n (n < N, the number of hidden layers is determined according to the number of this type of features) hidden layers, which contain a number of nodes (the number of nodes is equal to the number of categories of infection sites and disease sites).

[0040] Activation function: ; Output layer: ; Among them, is the score of the i-th class, is the predicted probability of the i-th class.

[0041] Loss function: Use the cross-entropy loss function to measure the prediction error of the model: ; Among them, is the true label, is the predicted probability of the model.

[0042] Use the training set data to train the model, and update the network weights through the backpropagation algorithm. The backpropagation algorithm optimizes the loss function based on gradient descent. Use the validation set to evaluate the model performance, and adjust the hyperparameters (such as learning rate, number of hidden layer nodes, etc.) to optimize the model.

[0043] Update formula (gradient descent): ; Among them, is the learning rate, is the weight of the network.

[0044] Evaluate the performance of the model on the test set. Integrate the evaluated BP neural network model into the system to predict the infection site and disease site of patients. Continuously optimize the model according to the actual application feedback and new case data.

[0045] In actual applications, first, clean the patient case data: remove duplicates and missing values to ensure data integrity and accuracy; standardize the data for subsequent analysis.

[0046] Classify the patient case data: Classify the patients according to the infection disease site and disease site to form the smallest dataset.

[0047] Obtain the smallest dataset: According to the data source, extract the key information related to the patient's infection disease site and disease site to ensure that the data of each patient contains the necessary features.

[0048] The data features in the cleaned minimum data set are input into the BP neural network algorithm; the BP neural network is trained using historical case data, and the network parameters are adjusted to improve prediction accuracy.

[0049] Infection and Disease Assessment: Model outputs the type of suspected infection site or infectious disease site in the patient.

[0050] Determine a classification threshold to determine whether a patient meets high-risk criteria.

[0051] If the result output by the model, that is, the probability value, is greater than the first classification threshold, the next step is to determine the infection site, etc.

[0052] By cleaning and classifying patient case data and inputting it into the BP neural network, the infection and disease types of high-risk patients can be effectively identified to provide support for subsequent medical decision-making.

[0053] S130, inputting the prediction results into a biosafety risk early warning model to distinguish the types of suspected infection sites and disease sites of the patient to obtain a distinction result.

[0054] In this embodiment, the differentiation result refers to the risk value corresponding to the suspected infection site and the disease site type. Monitoring checkpoints are set during the whole process of diagnosis and treatment of patients.

[0055] Specifically, the training process of the biosafety risk early warning model includes: Extracting the diagnosis and treatment data related to the infection site and related diseases from the historical case data, and performing data preprocessing to obtain preprocessing results; Selecting features related to the infection site in the pretreatment results and standardizing them to obtain standardized results; A logistic regression model was initially trained for each infection type in the standardized results, and the regression coefficient of each feature was calculated to assess its importance for the specific infection type; According to the regression coefficients, key features affecting various types of infection are screened out; Constructing a feature matrix according to the key features, and constructing a training set and a test set; The training set is used to train a random forest model, and multiple decision trees are constructed for classification prediction to obtain a biosafety risk early warning model.

[0056] Specifically, the biosafety risk early warning model is an assessment model that collects patient infection and infectious disease case data to construct a training set. Ensure that the data is cleaned and standardized; extract important features related to the infection site and disease site from the minimum data set for model training; combine logistic regression with random forest: Logistic regression: used to initially evaluate and understand the linear relationship between features and target variables. Random forest: As an integrated learning method, it can effectively handle nonlinear relationships and improve the accuracy and robustness of the model.

[0057] Using the training set data, logistic regression and random forest algorithms are combined to establish an evaluation model for infection sites and disease sites; model parameters are optimized through cross-validation.

[0058] Output evaluation indicators of suspected patients, such as prediction accuracy, recall rate, F1 score, etc., to measure model performance.

[0059] Use the validation set: Input the validation set patient case data into the established evaluation model to test the generalization ability of the model.

[0060] Weight adjustment: Based on the results of the validation set, analyze the model performance, adjust the weights in logistic regression and random forest, and optimize the model structure.

[0061] Iterative training: Continue training the model until it reaches convergence, that is, the performance on the validation set no longer improves significantly.

[0062] Through the combination of logistic regression and random forest algorithms, an evaluation model for infection sites and disease sites can be effectively established. Through the inspection of the validation set and the adjustment of weights, the reliability and accuracy of the model are ensured, providing a scientific basis for clinical decision-making.

[0063] In addition, when training the model, specifically, extract the process diagnosis and treatment data in the patient case and the diagnosis and treatment data related to the infection site and disease site (for example, body temperature, white blood cell count, urine analysis results, etc.), ensure that the extracted data is complete and relevant, and select features that are closely related to the infection type. Then process missing values: you can use the mean or median to fill in, or choose an appropriate method (such as interpolation) based on the characteristics of the data. Outlier identification: use statistical methods (such as Z-score, IQR) to detect outliers and decide whether to exclude or correct them. Duplicate data processing: remove duplicate records to avoid affecting model training.

[0064] Next, select features related to the site of infection and standardize the data features to ensure that the data is in the same dimension; for example, use the Pearson correlation coefficient or other methods to check the correlation between features and select features that are highly correlated with the target variable. Use Z-score standardization or Min-Max scaling to ensure that all features are in the same dimension to improve model performance.

[0065] Next, we use libraries such as sklearn to perform logistic regression training and obtain regression coefficients; we use p-values ​​and confidence intervals to evaluate feature significance and select important features. For example, for respiratory infections, we focus on body temperature and cough; for surgical site infections, we focus on the type of surgery and wound condition.

[0066] , where P(Y = 1X) is the predicted probability of illness and is the regression coefficient of the feature. Select important features based on the significance of the regression coefficient 8. For example, if it is significant for respiratory system infection, then body temperature is a key feature; for example: body temperature, cough, respiratory rate, blood oxygen saturation, etc. have a significant effect on respiratory system infection; surgical site infection such as type of surgery, postoperative body temperature, wound secretions, etc. have a significant effect; urine white blood cell count, urine color, frequent urination, etc. have a significant effect on urinary system infection; white blood cell count, hemoglobin level, platelet count, etc. have a significant effect on blood system infection; wound redness, swelling, pain level, etc. have a significant effect on skin and soft tissue infection.

[0067] Ensure that data is randomly allocated, usually 80% as training set and 20% as test set; use Grid Search or Random Search to tune hyperparameters, optimize the number of trees, maximum depth, etc.; use the training set data to train the random forest model and build multiple decision trees. And predict each infection type. Optimize the hyperparameters of the random forest, such as the number of trees, maximum depth, etc., through cross-validation to improve model performance.

[0068] Train a random forest model. Each decision tree in the random forest The formula used for prediction is: ; where M is the number of trees, is the prediction result of the jth tree for feature X.

[0069] Indicator calculation: Use confusion matrix to calculate accuracy, precision, recall and F1 score to evaluate the performance of the two models; Cross-validation: Use K-fold cross-validation to ensure the generalization ability of the model.

[0070] Feedback mechanism: Establish a feedback mechanism to continuously update the model based on new data from actual applications.

[0071] Feature Importance Analysis: Regularly check feature importance and adjust feature set to improve model accuracy.

[0072] Through the above steps, logistic regression and random forest models can be effectively constructed and optimized to improve the prediction accuracy of disease infection. In practical applications, continuous monitoring and iterative optimization are very important to ensure the performance of the model on new data.

[0073] S140, performing dynamic warning on the differentiation result in combination with the established incremental warning mechanism to obtain a warning result; In this embodiment, the incremental warning mechanism includes real-time data detection and recalculation of the patient's risk assessment value. Incremental warning is performed based on the daily changes in a patient's condition and existing monitoring results, and warnings are issued in combination with the number of patients with similar biosafety risks in the region to avoid cluster events and infectious disease transmission events.

[0074] Specifically, data such as patient temperature, symptom changes, and laboratory test results are continuously collected; sensors and electronic medical record systems (EMR) are used to automatically update patient information to ensure the timeliness of the data.

[0075] After obtaining the latest data, the system will use the existing assessment model to recalculate the patient's risk score; algorithms such as logistic regression and random forest can be used for training and optimization based on historical data.

[0076] Regular updates: Set up a daily or regular data update strategy to incorporate new data into patient records.

[0077] Dynamic adjustment: After each data update, the early warning model is automatically adjusted and optimized to ensure that the latest patient risk status is reflected.

[0078] Repeat warning judgment: History query: The system checks whether the patient has issued an early warning.

[0079] Comparison of risk factors: If they are similar risk factors, determine whether the warning has been handled.

[0080] Unprocessed: Remind the management staff to take action as soon as possible.

[0081] Resolved: No more warnings will be issued.

[0082] Cluster event monitoring: Combined with data from other patients in the area, analyze potential cluster events and infectious disease transmission risks.

[0083] Risk trend assessment: Through data integration, real-time assessment is conducted to determine whether there is a risk aggregation trend, and early warning strategies are adjusted accordingly.

[0084] Reduce system load: Incremental warning design helps reduce the load and memory usage of system models and keep the system running efficiently.

[0085] Visual interface: Displays daily new medical data in the form of a time series diagram to facilitate quick identification by medical staff.

[0086] Highlight: When new data meets the warning conditions, the relevant patient's situation will be marked in red, making it easy to quickly determine the suspected infection date.

[0087] User-friendly interface: The intuitive and easy-to-use user interface provides clear warning information and operation instructions.

[0088] Feedback mechanism: Allows medical staff to provide feedback to continuously optimize the early warning model and system functions.

[0089] In the hospital infection monitoring system, the incremental early warning mechanism dynamically updates the patient's health data and adjusts the patient's risk assessment in a timely manner to make early warnings and avoid cluster events or the spread of infectious diseases. The key lies in real-time monitoring and incremental update processing of data to gradually optimize the early warning results. This mechanism mainly includes the following steps: Real-time data detection and dynamic early warning: In the system, patient data is collected in real time, including but not limited to the following: Body temperature, symptom changes (such as respiratory rate, heart rate), laboratory test results (such as white blood cell count, C-reactive protein, etc.), and other biological indicators (obtained through bedside monitoring equipment and sensors) The latest health information of patients is automatically updated through sensors and electronic medical record systems (EMR) to ensure the timeliness of data. The daily changes in the patient's status will be automatically collected by the system and input into the early warning mechanism in real time, forming the basis for dynamic monitoring. The timeliness of data is the core of the incremental early warning mechanism, which ensures that any changes in physical signs or test indicators can be quickly captured and evaluated.

[0090] Recalculation of risk assessment values: The system reassesses the patient's risk based on the new data generated by the patient every day. After the data is collected, the system uses a preset assessment model to calculate the patient's current risk score. These models can be based on the following algorithms: Logistic regression: used to deal with binary classification problems, such as determining whether a patient is at risk of infection.

[0091] Random Forest: Combines multiple decision tree models to provide more stable predictions.

[0092] Advanced algorithms like Support Vector Machines (SVM) or Neural Networks: These can handle complex, multi-dimensional data.

[0093] After each data update, the system recalculates the patient's risk score using the existing risk model. This recalculation of the risk assessment value is based on daily dynamic changes, ensuring that the risk assessment reflects the patient's latest health status in a timely manner.

[0094] Incremental early warning: The core of incremental early warning is that the system does not have to start data analysis from scratch every time, but updates the existing evaluation results based on the daily incremental data. This can significantly reduce the computing load and improve the system's response speed. The specific approach is: Incremental data update: The system only needs to process some of the newly collected data and combine it with the patient’s previous health information to calculate a new risk score.

[0095] Historical data accumulation: The system retains the health history of all patients and updates them incrementally with new data without having to reprocess all the old data each time.

[0096] In this way, the system can make a "cumulative judgment" based on the new data, that is, combine it with previous trends and data changes to form a continuous evaluation of current risks. Incremental early warning avoids redundant steps in data processing and improves the real-time and efficiency of the system.

[0097] Regional biosafety risk warning: In addition to monitoring data from a single patient, the system also combines similar data from other patients in the area, such as patients in the same ward or department. If multiple patients in the same area have similar symptoms or biosafety risks, such as multiple patients with fever or signs of infection with the same pathogen at the same time, the system will analyze the potential risk of this cluster event. At this time, the system will: Analyze cluster trends: Use algorithms to identify whether there is a trend of spread of a certain type of infectious disease, especially the risk of infection by the same pathogen.

[0098] Compare data of patients in the area: Compare the patient's data with other patients in the same area to determine whether there is a possibility of large-scale infection.

[0099] Issue a regional warning: When the system detects that the number of patients in a specific area reaches a certain threshold (such as multiple patients with similar symptoms), a regional warning will be issued to prompt relevant departments or managers to take prevention and control measures.

[0100] Regular data updates and dynamic adjustments: In daily operation, the system will update patient data every day or every few hours according to the preset regular update strategy. This mechanism ensures that the system continuously obtains the latest patient data and uses it as a basis for risk assessment. The specific steps include: Regular data integration: Integrate newly collected data with the patient’s previous medical history and monitoring results to form a completely new case for the patient.

[0101] Dynamically adjust the early warning model: After each data update, the early warning model will automatically readjust to adapt to the latest risk changes. For example, if certain indicators of the patient (such as body temperature or infection indicators) suddenly rise, the system will automatically increase the risk level.

[0102] Repeat warning judgment: In order to avoid repeated warnings and reduce interference, the system automatically checks the patient's history to avoid issuing the same warning multiple times for the same problem. The logic is as follows: History query: Before issuing a new warning, the system will first check whether the patient has already triggered an alert due to the same risk factors.

[0103] Risk factor comparison: If the current risk factor is the same as the previous warning factor, the system will check whether the previous warning has been processed.

[0104] Unhandled: If the risk has not been handled, the system will remind the relevant managers to take action as soon as possible.

[0105] Resolved: If the issue has been resolved, the system will avoid issuing repeated alerts.

[0106] This mechanism ensures the effectiveness of early warning and reduces unnecessary alarms that interfere with the judgment of medical staff.

[0107] Cluster event monitoring and risk trend assessment: While processing a single patient, the system also monitors the patient data of the entire hospital or region in real time to identify potential cluster events. The specific process includes: Regional data integration: By integrating real-time data from all patients, the system can assess whether there is a clustering trend of similar infection cases, such as multiple patients in the same ward with elevated temperatures or multiple patients in the same department with respiratory symptoms.

[0108] Risk trend assessment: By dynamically analyzing the data of the entire area, the system can determine whether there is a risk of clustered incidents in a certain area, and adjust the early warning strategy accordingly and issue an alarm in advance.

[0109] System load optimization and visualization interface: An important purpose of the incremental early warning mechanism design is to reduce the system load by processing only the newly added data, reducing the computational complexity and memory usage of the model. This ensures that the system can still operate efficiently when large-scale data is input.

[0110] In terms of user experience, the system will use a visual interface to display the daily added data in the form of time series and trend charts to help medical staff intuitively identify risks. The interface design also includes: Highlight: When new data meets the warning conditions, the system will mark the relevant patient's condition in red to quickly guide medical staff to pay attention and handle it.

[0111] User-friendly design: The intuitive user interface allows medical staff to quickly view warning information, and simple operation instructions ensure that post-warning processing can be carried out quickly.

[0112] Feedback Mechanism: The system is equipped with a feedback mechanism that allows medical staff to provide feedback and processing results. Feedback can not only help optimize the early warning model, but also form a closed loop of processing to ensure that each early warning event has a corresponding response measure. This feedback mechanism can also serve as an important basis for adjusting and optimizing the model, and continuously improve the accuracy of the system's early warning.

[0113] The design of the entire incremental early warning mechanism is based on the effective combination of real-time data collection, dynamic risk assessment, regional cluster monitoring, system load optimization and feedback mechanism. It can not only effectively respond to the dynamic changes of individual patients, but also monitor cluster events in the region, issue early warnings in a timely manner, and ensure the biosafety of medical institutions.

[0114] S150: Send intervention information to a corresponding platform based on the early warning result.

[0115] In this embodiment, the intervention information includes outbreak warning information issued when the number of patients at the same type of biosafety risk during the screening period exceeds a first transmission threshold.

[0116] Based on monitoring data and preset risk models, early warning information is evaluated and generated. This includes risk warnings for individual patients and warnings for all staff based on regional data. Early warning information is generated and the warning results are pushed to the intervention module and relevant medical staff. Early warning information is also passed to the optimization module for subsequent adjustments and optimizations.

[0117] Based on the early warning information, specific intervention plans are generated and pushed. These include treatment suggestions, inspection and treatment plans, isolation measures, etc. for the infected site or infectious disease type. Based on the detailed early warning information provided by the early warning module, intervention measures are implemented and the intervention effects are recorded. The results of these intervention measures are fed back to the optimization module for further improvement of the system.

[0118] The method of this embodiment adopts the following technologies: Data Integration Technology: Data extraction: Extract data from multiple sources such as hospital information system (HIS), electronic medical record system (EMR), laboratory information system (LIS), etc.

[0119] Data cleaning: Remove duplicate, missing, and erroneous data to ensure data quality.

[0120] Data transformation (ETL): standardize data in different formats to meet subsequent analysis needs.

[0121] Data fusion: Integrate multi-source data from different systems to form a unified patient data set to support subsequent data analysis and decision-making.

[0122] Real-time data stream processing technology: Stream computing framework: uses stream processing technologies such as Apache Kafka and Apache Flink to achieve rapid processing of real-time data.

[0123] Event-driven architecture: trigger specific events based on changes in patient status and respond to real-time data streams in a timely manner.

[0124] Dynamic risk assessment algorithm: monitors patient data in real time, identifies abnormalities and risk changes, and provides real-time feedback.

[0125] Intelligent early warning model construction: Model selection: Use classification models (such as decision trees, support vector machines), regression models (such as linear regression), and ensemble learning models (such as random forests, XGBoost) to identify high-risk factors.

[0126] Combination of historical and real-time data: Use historical data to train models and combine real-time data for dynamic prediction and early warning.

[0127] Rule-based decision support: Determine the type of infection based on early warning information, and provide medical staff with operational suggestions and intervention plans to improve response efficiency.

[0128] Incremental early warning implementation: Incremental learning: The system updates data regularly or daily, and adjusts risk assessment and early warning models through incremental learning methods to ensure the accuracy and timeliness of the models.

[0129] Real-time updating algorithms: Designing algorithms to support models to automatically update when new data arrives, maintaining sensitivity to risks.

[0130] Performance optimization: System operation data analysis: Collect system operation data during use to evaluate the effectiveness of early warning effects and intervention measures.

[0131] Application of optimization algorithms: Genetic algorithms, particle swarm optimization and other optimization techniques are used to improve early warning models and intervention strategies to improve the overall performance and response speed of the system.

[0132] The method of this embodiment realizes real-time monitoring and risk assessment of the patient's health status, which will help medical institutions better respond to public health emergencies and improve the quality and efficiency of medical services.

[0133] The above-mentioned multi-point trigger biosafety monitoring method obtains patient diagnosis and treatment data and processes it to generate individual case data, inputs multiple infection prediction models to obtain prediction results, applies the prediction results to the biosafety risk warning model, distinguishes suspected infection sites and disease types, combines the incremental warning mechanism to dynamically generate warning results, and sends intervention information to relevant platforms. This overcomes the shortcomings of existing technologies in data integration, risk analysis, dynamic monitoring, flexibility of warning mechanisms, and information processing efficiency, and achieves more efficient and accurate hospital biosafety monitoring.

[0134] Figure 3 FIG. 3 is a schematic block diagram of a multi-point trigger biosafety monitoring device 300 provided in an embodiment of the present invention. Figure 3 As shown, corresponding to the above multi-point trigger biosafety monitoring method, the present invention also provides a multi-point trigger biosafety monitoring device 300. The multi-point trigger biosafety monitoring device 300 includes a unit for executing the above multi-point trigger biosafety monitoring method, and the device can be configured in a server. Figure 3 The multi-point triggered biosafety monitoring device 300 includes a case data acquisition unit 301, a prediction unit 302, a distinction unit 303, a dynamic warning unit 304 and a sending unit 305.

[0135] The case data acquisition unit 301 is used to acquire the patient's diagnosis and treatment data and perform data processing to obtain the patient's case data; the prediction unit 302 is used to input the patient's case data into early warning models of various infections and infectious diseases to predict infection and obtain prediction results; the differentiation unit 303 is used to input the prediction results into the biosafety risk early warning model to differentiate between the types of suspected infection sites and disease sites of patients and obtain differentiation results; the dynamic early warning unit 304 is used to dynamically warn the incremental early warning mechanism established in combination with the differentiation results to obtain early warning results; the sending unit 305 is used to send intervention information to the corresponding platform based on the early warning results.

[0136] In one embodiment, the case data acquisition unit 301 is used to acquire the patient's diagnosis and treatment data and convert the diagnosis and treatment data into a unified data format to obtain the patient's case data, wherein the patient's case data includes the patient's basic information and the diagnosis and treatment data generated during outpatient and hospitalization.

[0137] In one embodiment, the multi-point triggered biosafety monitoring device 300 also includes a first training unit, which is used to classify historical case data into infection cases, infectious disease cases and healthy patient types; divide the historical case data corresponding to the classified infection cases and infectious disease cases according to the infection site to obtain a sample set; and use the sample set to train the BP neural network algorithm to obtain early warning models for multiple infections and infectious diseases.

[0138] In one embodiment, the multi-point triggered biosafety monitoring device 300 also includes a second training unit, which is used to extract the diagnosis and treatment data related to the infection site and related diseases in the historical case data, and perform data preprocessing to obtain preprocessing results; select the features related to the infection site in the preprocessing results, and standardize them to obtain standardized results; use a logistic regression model to perform preliminary training on each infection type in the standardized results, calculate the regression coefficient of each feature, and evaluate its importance to the specific infection type; screen out the key features that affect each type of infection based on the regression coefficient; construct a feature matrix based on the key features, and construct a training set and a test set; use the training set to train a random forest model, and construct multiple decision trees for classification prediction to obtain a biosafety risk warning model.

[0139] It should be noted that technicians in the relevant field can clearly understand that the specific implementation process of the above-mentioned multi-point triggered biosafety monitoring device 300 and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and brevity of description, it will not be repeated here.

[0140] The multi-point trigger biosafety monitoring device 300 can be implemented in the form of a computer program. The computer program can be implemented in the form of a computer program. Figure 4 Runs on the computer device shown.

[0141] See also Figure 4 , Figure 4 5 is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device 500 may be a server, wherein the server may be an independent server or a server cluster composed of multiple servers.

[0142] See also Figure 4 The computer device 500 includes a processor 502 , a memory and a network interface 505 connected via a system bus 501 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .

[0143] The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions, and when the program instructions are executed, the processor 502 can execute a multi-point trigger biosafety monitoring method.

[0144] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .

[0145] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute a multi-point triggered biosafety monitoring method.

[0146] The network interface 505 is used to communicate with other devices over the network. Figure 4 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0147] The processor 502 is used to run the computer program 5032 stored in the memory to implement the following steps: Obtain the patient's diagnosis and treatment data and perform data processing to obtain the patient's individual case data; input the patient's individual case data into early warning models of various infections and infectious diseases to predict infection and obtain prediction results; input the prediction results into a biosafety risk early warning model to distinguish between the patient's suspected infection sites and disease site types to obtain distinction results; conduct dynamic early warning on the distinction results in combination with the established incremental early warning mechanism to obtain early warning results; and send intervention information to the corresponding platform based on the early warning results.

[0148] Wherein, the differentiation result includes the patient's risk assessment value.

[0149] The incremental early warning mechanism includes real-time data detection and recalculation of the patient's risk assessment value.

[0150] The intervention information includes outbreak warning information issued when the number of patients at the same biosafety risk during the screening period exceeds the first transmission threshold.

[0151] In one embodiment, when the processor 502 implements the step of acquiring the patient's diagnosis and treatment data and performing data processing to obtain the patient's case data, the processor 502 specifically implements the following steps: Obtain the patient's diagnosis and treatment data, and convert the diagnosis and treatment data into a unified data format to obtain the patient's case data, wherein the patient's case data includes the patient's basic information and the diagnosis and treatment data generated during outpatient and hospitalization.

[0152] In one embodiment, the processor 502 implements the following steps when implementing the training step of the early warning model of the multiple infections and infectious diseases: The historical case data are classified into infection cases, infectious disease cases and healthy patient types; the historical case data corresponding to the classified infection cases and infectious disease cases are divided according to the infection site to obtain a sample set; the BP neural network algorithm is trained using the sample set to obtain early warning models for multiple infections and infectious diseases.

[0153] In one embodiment, when implementing the training step of the biosafety risk early warning model, the processor 502 specifically implements the following steps: Extract the diagnosis and treatment data related to the infection site and related diseases from the historical case data, and perform data preprocessing to obtain preprocessing results; select the features related to the infection site in the preprocessing results, and standardize them to obtain standardized results; use a logistic regression model to perform preliminary training on each infection type in the standardized results, calculate the regression coefficient of each feature, and evaluate its importance to the specific infection type; screen out the key features that affect each type of infection based on the regression coefficient; construct a feature matrix based on the key features, and construct a training set and a test set; use the training set to train a random forest model, and construct multiple decision trees for classification prediction to obtain a biosafety risk early warning model.

[0154] It should be understood that in the embodiment of the present application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0155] It can be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiment of the above method.

[0156] Therefore, the present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program, wherein when the computer program is executed by a processor, the processor executes the following steps: Obtain the patient's diagnosis and treatment data and perform data processing to obtain the patient's individual case data; input the patient's individual case data into early warning models of various infections and infectious diseases to predict infection and obtain prediction results; input the prediction results into a biosafety risk early warning model to distinguish between the patient's suspected infection sites and disease site types to obtain distinction results; conduct dynamic early warning on the distinction results in combination with the established incremental early warning mechanism to obtain early warning results; and send intervention information to the corresponding platform based on the early warning results.

[0157] Wherein, the differentiation result includes the patient's risk assessment value.

[0158] The incremental early warning mechanism includes real-time data detection and recalculation of the patient's risk assessment value.

[0159] The intervention information includes outbreak warning information issued when the number of patients at the same biosafety risk during the screening period exceeds the first transmission threshold.

[0160] In one embodiment, when the processor executes the computer program to implement the step of obtaining the patient's diagnosis and treatment data and performing data processing to obtain the patient's case data, the processor specifically implements the following steps: Obtain the patient's diagnosis and treatment data, and convert the diagnosis and treatment data into a unified data format to obtain the patient's case data, wherein the patient's case data includes the patient's basic information and the diagnosis and treatment data generated during outpatient and hospitalization.

[0161] In one embodiment, when the processor executes the computer program to implement the training steps of the early warning models of the multiple infections and infectious diseases, the processor specifically implements the following steps: The historical case data are classified into infection cases, infectious disease cases and healthy patient types; the historical case data corresponding to the classified infection cases and infectious disease cases are divided according to the infection site to obtain a sample set; the BP neural network algorithm is trained using the sample set to obtain early warning models for multiple infections and infectious diseases.

[0162] In one embodiment, when the processor executes the computer program to implement the training step of the biosafety risk early warning model, the processor specifically implements the following steps: Extract the diagnosis and treatment data related to the infection site and related diseases from the historical case data, and perform data preprocessing to obtain preprocessing results; select the features related to the infection site in the preprocessing results, and standardize them to obtain standardized results; use a logistic regression model to perform preliminary training on each infection type in the standardized results, calculate the regression coefficient of each feature, and evaluate its importance to the specific infection type; screen out the key features that affect each type of infection based on the regression coefficient; construct a feature matrix based on the key features, and construct a training set and a test set; use the training set to train a random forest model, and construct multiple decision trees for classification prediction to obtain a biosafety risk early warning model.

[0163] The storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, etc., which are computer-readable storage media that can store program codes.

[0164] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0165] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0166] The steps in the method of the embodiment of the present invention can be adjusted in order, combined and deleted according to actual needs. The units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0167] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, terminal, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention.

[0168] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A multi-point triggered biosafety monitoring method, characterized in that: include: Obtain the patient's diagnosis and treatment data and process the data to obtain the patient's case data; Inputting the patient case data into early warning models of multiple infections and infectious diseases to predict infection, so as to obtain prediction results; Inputting the prediction results into a biosafety risk early warning model to distinguish the types of suspected infection sites and disease sites of patients to obtain a distinction result; Performing dynamic warning on the differentiation result in combination with the established incremental warning mechanism to obtain a warning result; Send intervention information to the corresponding platform based on the early warning result.

2. The multi-point triggered biosafety monitoring method according to claim 1, characterized in that: The obtaining of the patient's diagnosis and treatment data and processing of the data to obtain the patient's case data includes: Obtain the patient's diagnosis and treatment data, and convert the diagnosis and treatment data into a unified data format to obtain the patient's case data, wherein the patient's case data includes the patient's basic information and the diagnosis and treatment data generated during outpatient and hospitalization.

3. The multi-point triggered biosafety monitoring method according to claim 1, characterized in that: The training process of the early warning model for multiple infections and infectious diseases includes: Categorize historical case data into infection cases, infectious disease cases, and healthy patient types; The historical case data corresponding to the classified infection cases and infectious disease cases are divided according to the infection site to obtain a sample set; The sample set is used to train the BP neural network algorithm to obtain early warning models for various infections and infectious diseases.

4. The multi-point triggered biosafety monitoring method according to claim 3, characterized in that: The training process of the biosafety risk early warning model includes: Extracting the diagnosis and treatment data related to the infection site and related diseases from the historical case data, and performing data preprocessing to obtain preprocessing results; Selecting features related to the infection site in the pretreatment results and standardizing them to obtain standardized results; A logistic regression model was initially trained for each infection type in the standardized results, and the regression coefficient of each feature was calculated to assess its importance for the specific infection type; According to the regression coefficients, key features affecting various types of infection are screened out; Constructing a feature matrix according to the key features, and constructing a training set and a test set; The training set is used to train a random forest model, and multiple decision trees are constructed for classification prediction to obtain a biosafety risk early warning model.

5. The multi-point triggered biosafety monitoring method according to claim 1, characterized in that: The differentiation result includes a risk assessment value for the patient.

6. The multi-point triggered biosafety monitoring method according to claim 5, characterized in that: The incremental early warning mechanism includes real-time data detection and recalculation of the patient's risk assessment value.

7. The multi-point triggered biosafety monitoring method according to claim 1, characterized in that: The intervention information includes outbreak warning information issued when the number of patients at the same biosafety risk during the screening period exceeds the first transmission threshold.

8. A multi-point triggered biosafety monitoring device, characterized in that: include: A case data acquisition unit, used to acquire the patient's diagnosis and treatment data and perform data processing to obtain the patient's case data; A prediction unit, used for inputting the patient case data into early warning models of multiple infections and infectious diseases to perform infection prediction to obtain prediction results; A differentiation unit, used for inputting the prediction result into the biosafety risk early warning model to differentiate the types of suspected infection sites and disease sites of the patient to obtain a differentiation result; A dynamic early warning unit, used for dynamically warning the incremental early warning mechanism established in combination with the differentiation result to obtain an early warning result; A sending unit is used to send intervention information to a corresponding platform based on the early warning result.

9. A computer device, characterized in that: The computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program.

10. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images