Infection risk assessment method and system for nursing in infectious disease department

Abstract

The invention relates to the technical field of medical care, and particularly discloses an infection risk assessment method and system for nursing of an infectious disease department, and the method comprises the following steps: 1, collecting patient information, an inpatient area environment, medical operation and meteorological data, and constructing a multi-dimensional data set; 2, processing the data set, and extracting key feature vectors containing seasonal feature vectors; 3, inputting the key feature vector into a pre-training hybrid model to generate an infection risk level; 4, loading a corresponding nursing scene on the VR platform according to the risk level, embedding a virtual risk point, capturing and analyzing the operation of the nursing personnel, and generating a score value; 5, if the report exceeds the threshold value, triggering early warning and pushing a guidance scheme to a VR platform. According to the method, the influence of seasonal factors is considered, basic information of a patient, environment data of an inpatient area, medical operation data and meteorological information are fused, a seasonal fluctuation rule is analyzed and extracted by using a time sequence, a hybrid prediction model is constructed, and the risk of an infectious disease department is analyzed more accurately.

Description

Technical Field

[0001] This invention relates to the field of medical and nursing technology, and in particular to a method and system for infection risk assessment in infectious disease nursing. Background Technology

[0002] In hospital infection control, current infection risk assessments largely rely on the individual experience of healthcare workers, depending on their subjective understanding of the clinical setting to determine the likelihood of a patient's infection. Due to differences in years of experience, professional background, and clinical expertise among healthcare workers, disagreements often arise regarding risk assessments for the same type of case. For example, when dealing with elderly patients with underlying medical conditions, some healthcare workers emphasize medical history assessment, while others focus on current symptoms. This results in a lack of unified standards and systematic logic in the assessment results, making it impossible to form a reproducible and traceable assessment system.

[0003] Furthermore, there is a general limitation of focusing on a single point, making it difficult to achieve comprehensive coverage of infection risk-related factors. In practice, assessments often only focus on the patient's immediate physiological state, such as explicit indicators like body temperature and white blood cell count, while neglecting multidimensional implicit factors crucial for infection control: such as whether environmental parameters like ward humidity and ventilation frequency are suitable for microbial growth, whether invasive procedures and antibiotic use increase exposure risk, and whether the patient's underlying medical history increases susceptibility. This assessment logic, which emphasizes symptoms over correlations, leads to a significant one-sidedness in risk judgment. At the same time, the climate characteristics, viral activity, and population movement patterns of different seasons can significantly alter the infection risk landscape. For example, the hot and humid environment of summer accelerates microbial reproduction, potentially increasing the risk of intestinal infections; under the low temperature and dry conditions of winter, respiratory viruses are more easily transmitted through droplets, leading to a significant increase in the incidence of diseases such as influenza. Therefore, without quantitative analysis of seasonal patterns, traditional methods struggle to predict infection peaks in advance, resulting in control measures lagging behind risk changes and failing to achieve proactive prevention and control. Summary of the Invention

[0004] The purpose of this invention is to provide an infection risk assessment method and system for infectious disease nursing, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] An infection risk assessment method for infectious disease nursing, characterized by comprising the following steps:

[0007] Step 1: Collect and integrate basic patient information, ward environment data, medical operation data, and external data including meteorological information to construct a multi-dimensional dataset;

[0008] Step 2: Standardize the multi-dimensional dataset and extract key feature vectors. The key feature vectors include seasonal feature vectors that characterize the transmission patterns of infectious pathogens. The seasonal features are used to quantify the impact of the four seasons on the risk of infectious diseases using a time series model.

[0009] Step 3: Input the key feature vector into the pre-trained hybrid infection risk prediction model to generate an infection risk score and risk level for the target patient or target ward;

[0010] Furthermore, based on the risk level, the corresponding scenario model is called from the pre-built virtual scenario library; the scenario model is preset with virtual pathogen density and distribution that match the risk level.

[0011] Step 4: Based on the risk level, dynamically generate and load the corresponding infectious disease department nursing simulation scenario in the virtual reality interactive platform, and use the virtual risk points embedded in the simulation scenario that correspond to the risk level to capture and analyze the operational behavior data of nursing staff in real time to generate an infection risk score.

[0012] Step 5: When the infection risk score exceeds the preset threshold or the operation behavior data fails to meet the standard, an early warning message is triggered, and the early warning message and the targeted operation guidance plan are pushed to the virtual reality interaction platform in real time.

[0013] Furthermore, the extraction of key feature vectors specifically includes:

[0014] Based on the multi-dimensional dataset, characteristics related to the source of infection, transmission route, and susceptible population are analyzed; wherein, the analysis of the transmission route includes a quantitative assessment of at least one of the routes of air, droplets, drinking water, food, biological vectors, bodily fluids, and contact; the quantitative assessment of the airborne transmission route is based on data of ward carbon dioxide concentration, per capita space area, and air circulation speed, and an airborne transmission risk index is obtained through weighted calculation.

[0015] The extraction of the seasonal feature vector specifically includes: integrating temperature, humidity and rainfall information from meteorological data, combining holiday traffic change data, and using time series models to quantitatively analyze the fluctuating impact of different seasons (spring, summer, autumn and winter) on the risk of common infectious diseases.

[0016] The quantitative analysis of the impact of different seasons involves fitting historical infection case numbers with meteorological data using the Prophet model to extract seasonal component coefficients with an annual cycle, and using these coefficients as the main components of the seasonal feature vector.

[0017] Furthermore, the mixed infection risk prediction model includes a machine learning algorithm and a time series analysis model; wherein the machine learning algorithm includes XGBoost or LSTM; and the time series analysis model includes at least one of ARIMA or Prophet.

[0018] Furthermore, the hybrid infection risk prediction model processes individual patient and static environmental characteristics through machine learning algorithms and processes the seasonal feature vector through a time series analysis model. The outputs of the two sub-models are then weighted and fused to generate an infection risk score.

[0019] Furthermore, the embedding of virtual risk points in the simulated scenario specifically includes: simulating at least one key nursing procedure such as contact with patients without washing hands, failure to wear protective equipment properly, or violation of sterile area operation rules.

[0020] Furthermore, after the early warning information is triggered, the system also includes: linking the early warning information to the hospital resource management system to provide decision-making suggestions for the reserve of protective materials and the allocation of nursing staff.

[0021] A system for infection risk assessment in infectious disease nursing, comprising:

[0022] Data acquisition and processing module: used to collect and integrate multi-dimensional data, and perform standardization processing and key feature vector extraction;

[0023] Risk prediction module: It has a built-in hybrid infection risk prediction model, which is used to calculate infection risk scores and levels based on key feature vectors;

[0024] Furthermore, the method for constructing the mixed infection risk prediction model is as follows: The XGBoost algorithm is used to train the individual clinical characteristics of patients to obtain a first predicted score S1; the ARIMA model is used to train the time-series data of historical infection rates to predict the baseline infection risk score S2 for the current period. The final infection risk score is: S_final = α*S1 + β*S2, where α and β are weighting coefficients determined by a grid search method. In a preferred embodiment, α = 0.7, β = 0.3.

[0025] The operation method of the mixed infection risk prediction model is as follows: First, the ARIMA model is used to make a macro-prediction of the infection risk of the entire ward and output the risk level (high / medium / low); then, according to the macro-risk level, different XGBoost models are selected to conduct micro-individual risk assessment of admitted patients (for example, a more sensitive model threshold is used during high-risk periods).

[0026] Furthermore, the hyperparameters of the XGBoost model are as follows: maximum tree depth (max_depth) is 3-8, and learning rate (learning_rate) is 0.01-0.3. The parameters of the ARIMA model are determined through an ACF / PACF graph, and its order (p, d, q) ranges from (1-3, 0-1, 1-3). The training data for the model comes from the electronic medical record data of XX Hospital from 2018 to 2023, containing approximately 10,000 patient admission records. The training set and test set are divided in an 8:2 ratio. The model performance is evaluated using 10-fold cross-validation; the hybrid model achieves an accuracy of 85% and a recall of 90% on the test set.

[0027] Furthermore, the mixed infection risk prediction model employs a weighted fusion approach. Specifically, the XGBoost sub-model receives individual patient characteristics and outputs an individual risk score S_ind; the ARIMA sub-model receives the monthly infection rate sequence of the ward and outputs a population baseline risk score S_pop. Through validation with historical data, the optimal weights were determined to be α = 0.7 and β = 0.3. The final risk score S_final = 0.7 * S_ind + 0.3 * S_pop.

[0028] Virtual Reality Assessment Module: Used to generate and run nursing simulation scenarios, embed virtual risk points, and assess nursing staff's operational behaviors in real time;

[0029] Early warning and intervention module: used to trigger early warnings and push intervention information based on risk scores and operational assessment results.

[0030] Furthermore, the risk prediction module also includes a model optimization unit, which is used to continuously optimize the prediction accuracy by adjusting the algorithm hyperparameters and feature weights.

[0031] Furthermore, the virtual reality assessment module also includes a multi-person collaboration unit, which supports multiple nursing staff to simultaneously access the same virtual scene for collaborative drills and to compare and analyze the operation data of multiple people.

[0032] Furthermore, it includes a visualization module, which displays the spatial distribution of infection risk in the ward in the form of a heat map, and shows the seasonal variation trend of infection risk in the form of a line graph.

[0033] Furthermore, the early warning and intervention module is connected to the hospital resource management system and can automatically generate resource allocation suggestions based on risk prediction results.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] 1. This invention utilizes VR equipment to simulate nursing scenarios, monitor in real time whether nursing staff are operating in accordance with regulations, provide immediate reminders and guidance for incorrect operations, generate operation reports, support multi-person collaborative drills and online assessments, improve nursing staff's prevention and control capabilities, enhance hospital infection control efficiency, reduce infection incidence, and ensure patient safety.

[0036] 2. This invention takes into account the influence of seasonal factors, integrates basic patient information, ward environment data, medical operation data, and external data including meteorological information, uses time series analysis to extract seasonal fluctuation patterns, and constructs a hybrid prediction model to more accurately analyze the risks in the infectious disease department.

[0037] 3. This invention collects and integrates basic patient information, ward environment data, medical operation data, and external data including meteorological information to construct a multi-dimensional dataset, perform feature extraction and model training, improve the accuracy of risk assessment, and comprehensively consider multi-dimensional factors.

[0038] 4. This invention achieves intelligent assessment of infection risk through multi-dimensional data collection and machine learning algorithms, reducing the influence of subjective factors and solving the problem of traditional assessment methods relying on experience-based judgment.

[0039] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0040] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0041] Figure 1 This is a flowchart illustrating an infection risk assessment method for infectious disease nursing provided in this application. Detailed Implementation

[0042] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0043] like Figure 1As shown in the figure, this embodiment provides an infection risk assessment method for infectious disease nursing, characterized by the following steps:

[0044] Step 1: Collect and integrate basic patient information, ward environment data, medical operation data, and external data including meteorological information to construct a multi-dimensional dataset;

[0045] Furthermore, by connecting to the hospital's electronic medical record system, basic information of target patients (such as age, gender, and history of underlying diseases) and medical operation data (such as records of invasive procedures and antibiotic use) are extracted; sensor devices deployed in the wards are linked to collect ward environmental data (such as ward temperature, humidity, ventilation frequency, and disinfection frequency); external data platforms are accessed to obtain meteorological information including temperature, humidity, and rainfall, as well as data on changes in ward population flow during holidays (such as the number of visitors and patient turnover); the above data are categorized and summarized to form a multi-dimensional dataset covering patients, environment, operations, and external influences, laying the foundation for subsequent analysis.

[0046] Step 2: Standardize the multi-dimensional dataset and extract key feature vectors, including seasonal feature vectors used to characterize the transmission patterns of the infectious pathogen.

[0047] Furthermore, data standardization processing employs outlier removal (such as removing humidity and temperature data that significantly deviate from the normal range) and data normalization (mapping indicators of different magnitudes to the [0,1] interval) to eliminate data format differences and interference items.

[0048] Furthermore, the key feature vector extraction focuses on three core dimensions: source of infection, transmission route, and susceptible population. It quantitatively analyzes relevant features. For example, regarding transmission routes, it uses data such as air pollutant concentration, droplet transmission distance simulation data, and drinking water hygiene test results to numerically evaluate at least one transmission route, including air, droplets, drinking water, food, biological vectors, body fluids, and contact, and forms a transmission route feature sub-vector.

[0049] Furthermore, seasonal feature vectors are extracted. By integrating temperature, humidity, and rainfall information from meteorological data with holiday traffic data, and importing them into ARIMA or Prophet time series models, the fluctuating impact of spring, summer, autumn, and winter on the risk of common infectious diseases such as influenza and intestinal infections (e.g., the coefficient of increase in intestinal infection risk due to high humidity in summer) is analyzed to generate seasonal feature sub-vectors that characterize the transmission patterns of pathogens. The feature sub-vectors are then merged to form a complete key feature vector.

[0050] Step 3: Input the key feature vector into the pre-trained hybrid infection risk prediction model to generate an infection risk score and risk level for the target patient or target ward;

[0051] Furthermore, the key feature vectors extracted in step 2 are input into the pre-trained hybrid infection risk prediction model. The hybrid model, which includes a machine learning algorithm and a time series analysis model, is launched through model invocation. The machine learning algorithm is XGBoost or LSTM (used to analyze the correlation between individual patient data and environmental data), and the time series analysis model is ARIMA or Prophet (used to combine seasonal feature vectors to analyze the trend of risk changes).

[0052] Furthermore, risk calculation and level determination are performed: For target patients: the model combines their individual characteristics and seasonal influences to output an infection risk score of 0-100 points, and classifies the risk level into low risk (0-30 points), medium risk (31-70 points), and high risk (71-100 points);

[0053] Furthermore, for the target ward: the model integrates the overall environmental data of the ward, the characteristics of the patient population, and seasonal patterns to generate an overall infection risk score and level for the ward, providing a basis for regional prevention and control.

[0054] In another embodiment, the construction of the hybrid infection risk prediction model further includes feature vector concatenation and fusion; specifically, it includes the following steps:

[0055] a) Feature extraction: Extract individual feature vectors V_p from patient data, with dimension [N]; use the Prophet time series model to decompose the regional historical infection case data, extracting its seasonal component S_t and trend component T_t, which together constitute the seasonal feature vector.

[0056] V_s = [S_t, T_t], with a dimension of [2];

[0057] b) Feature fusion: The individual feature vector V_p is concatenated with the seasonal feature vector V_s to obtain the fused feature vector V_fused = [V_p, V_s], with a total dimension of [N+2].

[0058] c) Model Training and Prediction: Using the fused feature vector V_fused as input, supervised training is performed using the XGBoost algorithm. The output of the model is the final infection risk score. The fusion of the hybrid infection risk prediction model can be either weighted fusion or feature vector concatenation fusion.

[0059] Step 4: Based on the risk level, dynamically generate and load the corresponding infectious disease department nursing simulation scenario in the virtual reality interactive platform, and use the virtual risk points embedded in the simulation scenario that correspond to the risk level to capture and analyze the operational behavior data of nursing staff in real time to generate an infection risk score.

[0060] In this embodiment, the virtual reality interaction platform includes a head-mounted display and motion capture equipment.

[0061] Furthermore, based on the risk level output in step 3, a nursing assessment is performed in the virtual reality interactive platform: according to the simulation scenario loading, the platform dynamically generates an appropriate scenario based on the risk level, such as loading a respiratory infection patient nursing scenario for high-risk levels and loading a routine infectious disease nursing scenario for medium- and low-risk levels.

[0062] Furthermore, virtual risk point embedding: Based on the virtual risk points set in the scenario that match the risk level, at least one of the following key link scenarios is included: simulating scenarios of unauthorized operation such as direct contact with the patient's skin without washing hands, not wearing a protective mask properly to the face, and crossing the sterile area with hands to pick up objects during aseptic operation;

[0063] Furthermore, operational behavior monitoring and analysis: nursing staff wear VR devices (such as head-mounted displays and motion capture devices) to enter a virtual scene, and the system captures their operational behavior data in real time (such as hand movement trajectories, protective equipment wearing order, and operation duration), and compares it with the preset standard procedures; based on the compliance of the operation (such as the completeness of handwashing steps and the correctness of protective equipment wearing), an infection risk score report is generated, which includes operation score, error type, and violation frequency.

[0064] Step 5: When the infection risk score exceeds the preset threshold or the operation behavior data fails to meet the standard, an early warning message is triggered, and the early warning message and the targeted operation guidance plan are pushed to the virtual reality interaction platform in real time.

[0065] Furthermore, the infection risk scoring report and operational data generated in step 4 are subject to threshold judgment: if the score of the scoring report is lower than the preset qualified threshold (e.g., 60 points), or if the operational behavior data is found to be substandard (e.g., the number of violations of protective equipment wearing is ≥2 times), the system will immediately trigger an early warning message; the early warning message (e.g., the operational score is substandard and there is a risk of infection control) and the targeted operational guidance plan (e.g., the seven-step handwashing method video, the animation of the standard wearing of protective equipment) will be pushed to the virtual reality interactive platform in real time for nursing staff to view and correct the operation immediately; the data will be compared with the pre-stored standard operating procedure database, and when there is an error in the operation sequence or the operation exceeds the preset safety boundary, it will be judged as substandard operation behavior.

[0066] The extraction of key feature vectors specifically includes:

[0067] Based on the aforementioned multi-dimensional dataset, characteristics related to the source of infection, transmission routes, and susceptible populations are analyzed; wherein, the analysis of transmission routes includes a quantitative assessment of at least one of the following routes: airborne, droplet-borne, waterborne, foodborne, biological vector-borne, bodily fluid-borne, and contact-borne.

[0068] The extraction of the seasonal feature vector specifically includes: integrating temperature, humidity and rainfall information from meteorological data, combining holiday traffic change data, and using time series models to quantitatively analyze the fluctuating impact of different seasons (spring, summer, autumn and winter) on the risk of common infectious diseases.

[0069] The mixed infection risk prediction model includes a machine learning algorithm and a time series analysis model; wherein the machine learning algorithm includes XGBoost or LSTM; and the time series analysis model includes at least one of ARIMA or Prophet.

[0070] The embedding of virtual risk points in the simulated scenario specifically includes: simulating at least one key nursing procedure such as contact with patients without washing hands, failure to wear protective equipment properly, or violation of sterile area operation rules.

[0071] After the early warning information is triggered, the system will also link the early warning information to the hospital resource management system to provide decision-making suggestions for the reserve of protective materials and the allocation of nursing staff.

[0072] Specifically, the early warning information will be synchronized with the hospital resource management system, and decision-making suggestions will be generated based on the risk level. For example, for high-risk wards, a material and manpower allocation plan will be pushed to increase the reserve of N95 masks by 20% and add one infection control nurse, so as to assist managers in optimizing the allocation of prevention and control resources.

[0073] A system for infection risk assessment in infectious disease nursing, comprising:

[0074] Data acquisition and processing module: used to collect and integrate multi-dimensional data, and perform standardization processing and key feature vector extraction;

[0075] Furthermore, after the system is launched, the data acquisition and processing module first collects and integrates multi-dimensional data: it connects to the hospital's electronic medical record system to obtain basic patient information (such as age and underlying diseases) and medical operation records (such as the number of invasive treatments); it links with ward sensors to collect environmental data (such as ward humidity and ventilation frequency); and it connects to the meteorological platform and holiday database to import data on temperature, humidity, rainfall, and changes in population flow. Subsequently, the module automatically performs standardization processing on the data (such as outlier removal and data normalization) and extracts key feature vectors (including infectious source correlation features, quantitative indicators of transmission routes, and seasonal feature vectors) according to preset rules, providing a data foundation for risk prediction.

[0076] Risk prediction module: It has a built-in hybrid infection risk prediction model, which is used to calculate infection risk scores and levels based on key feature vectors;

[0077] Furthermore, the risk prediction module calls a pre-trained hybrid infection risk prediction model, inputting the extracted key feature vectors into the model: for a single patient, the model combines their individual data with seasonal characteristics, calculating an infection risk score using XGBoost or LSTM algorithms, and outputting a low / medium / high risk level; for the entire ward, the model combines ARIMA or Prophet time series analysis results to generate an overall ward risk level. Simultaneously, the model optimization unit can monitor prediction accuracy in real time, and medical staff can continuously optimize model performance by adjusting algorithm hyperparameters (such as learning rate and tree depth) or feature weights in the background.

[0078] Virtual Reality Assessment Module: Used to generate and run nursing simulation scenarios, embed virtual risk points, and assess nursing staff's operational behaviors in real time;

[0079] Furthermore, based on the risk level, a virtual reality assessment module automatically generates suitable nursing simulation scenarios: for high-risk scenarios, high-frequency risk points such as improper wearing of N95 masks and hands crossing sterile areas during aseptic procedures are embedded in the scenario; for low-risk scenarios, the focus is on basic procedures such as routine handwashing and the removal of protective equipment after patient contact. Nursing staff wear head-mounted displays, activate motion capture devices, and connect to the virtual scenario. The system then captures their operational behavior data in real time (such as hand movement trajectories and the order in which protective equipment is worn).

[0080] Early warning and intervention module: used to trigger early warnings and push intervention information based on risk scores and operational assessment results.

[0081] The risk prediction module also includes a model optimization unit, which is used to continuously optimize the prediction accuracy by adjusting the algorithm hyperparameters and feature weights.

[0082] Furthermore, the early warning and intervention module automatically initiates a response when two types of triggering conditions occur: First, if the risk prediction score exceeds a preset threshold (e.g., high risk, score ≥80 points), the module immediately sends an SMS warning to the responsible medical staff and marks the risk warning in the electronic medical record system; Second, if the operational behavior in the virtual assessment does not meet the standards (e.g., handwashing steps are missing ≥2 steps), the system pops up an instant reminder on the virtual reality interactive platform and simultaneously displays a video of the correct operation (e.g., an animation of the seven-step handwashing method).

[0083] The virtual reality assessment module also includes a multi-person collaboration unit, which supports multiple nursing staff to simultaneously access the same virtual scene for collaborative drills and to compare and analyze the operation data of multiple people.

[0084] The multi-user collaboration unit is implemented based on a client-server architecture. Specifically, it includes the following:

[0085] State synchronization server: As the authoritative host, it maintains a unified virtual scene state, including the position and posture of all user avatars and the state of interactive objects in the scene.

[0086] Client synchronization module: Runs on each virtual reality interaction device, used to send local user operation data (via UDP / TCP protocol) to the state synchronization server, and receive global state update data from the server, thereby refreshing the local rendering screen.

[0087] Synchronization mechanism: The server broadcasts state snapshots to all clients at fixed time intervals (e.g., 10-40 times per second) and smooths the client display through interpolation and extrapolation algorithms to offset the effects of network latency.

[0088] In another embodiment, the multi-user collaborative unit employs a frame synchronization mechanism. The server is responsible for collecting the operation instructions generated by each VR client within each logical frame and packaging them into an instruction set, which is then broadcast to all clients. Each client executes all instructions in the instruction set locally in the same order, thereby ensuring the consistency of the virtual environment state across multiple clients.

[0089] The multi-user collaboration unit also includes an operation comparison and analysis submodule. Its workflow is as follows: During the collaborative exercise, the server records the key operation event sequence of all users with a unified timestamp; after the exercise, the operation event sequence of each user is matched with the standard operation process sequence by the longest common subsequence to identify operation sequence errors; at the same time, the completion time of each step is calculated and compared with the standard time consumption, and finally a visual comparison and analysis report is generated.

[0090] Furthermore, if multi-person collaborative drills are conducted, multiple nursing staff can be invited to simultaneously access the same virtual scenario (such as simulating a scenario where multiple patients are cared for simultaneously in a ward) through a multi-person collaborative unit. The system records the degree of cooperation and compliance of each person's operation and generates team operation evaluation data.

[0091] It also includes a visualization module, which displays the spatial distribution of infection risk in the ward in the form of a heat map and the seasonal trend of infection risk in the form of a line graph.

[0092] Furthermore, the visualization module updates data in real time, marking the infection risk distribution of each ward on the ward floor plan in the form of a heat map (red indicates high-risk areas), and displaying the seasonal change trend of ward risk over the past 3 months (such as the peak risk during the winter flu season) using a line graph, making it easier for medical staff to intuitively grasp the key points of prevention and control. After the nursing assessment, the system generates an individual / team operation assessment report, recording the error type (such as exceeding the time limit for wearing protective equipment), the frequency of errors, and improvement suggestions. Medical staff can conduct targeted training based on the report to improve their prevention and control capabilities.

[0093] The early warning and intervention module is connected to the hospital resource management system and can automatically generate resource allocation suggestions based on risk prediction results.

[0094] At the same time, the module works in conjunction with the hospital resource management system to generate resource allocation suggestions based on the risk level: if the ward is high-risk, it will automatically output a plan to increase the reserve of N95 masks by 30% and add 2 infection control nurses. Medical and nursing management personnel can directly refer to the plan to implement resource adjustments.

[0095] In this implementation, the system operation data is reviewed regularly (e.g., monthly): the data acquisition and processing module summarizes historical risk prediction results and nursing assessment data to analyze the correlation between seasonal characteristics and infection risk (e.g., the correlation between high humidity in summer and intestinal infection); the risk prediction module iteratively trains the model based on new data to further improve prediction accuracy; the virtual reality assessment module updates virtual risk points and scenarios according to the latest infection control standards (e.g., requirements for protection against novel pathogens) to ensure that the system always meets actual clinical needs.

[0096] The foregoing description illustrates and describes several preferred embodiments of this application. However, as previously stated, it should be understood that this application is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the inventive concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of this application should be within the protection scope of the appended claims.

Claims

1. A method for infection risk assessment in infectious disease ward nursing, characterized in that, Includes the following steps: Step 1: Collect and integrate basic patient information, ward environment data, medical operation data, and external data including meteorological information to construct a multi-dimensional dataset; Step 2: Standardize the multi-dimensional dataset and extract key feature vectors, including seasonal feature vectors used to characterize the transmission patterns of the infectious pathogen. Step 3: Input the key feature vector into the pre-trained hybrid infection risk prediction model to generate an infection risk score and risk level for the target patient or target ward; Step 4: Based on the risk level, dynamically generate and load the corresponding infectious disease department nursing simulation scenario in the virtual reality interactive platform, and use the virtual risk points embedded in the simulation scenario that correspond to the risk level to capture and analyze the operational behavior data of nursing staff in real time to generate an infection risk score. Step 5: When the infection risk score exceeds the preset threshold or the operation behavior data fails to meet the standard, an early warning message is triggered, and the early warning message and the targeted operation guidance plan are pushed to the virtual reality interaction platform in real time.

2. The infection risk assessment method for infectious disease nursing according to claim 1, characterized in that, The extraction of key feature vectors specifically includes: Based on the aforementioned multi-dimensional dataset, characteristics related to the source of infection, transmission routes, and susceptible populations are analyzed; wherein, the analysis of transmission routes includes a quantitative assessment of at least one of the following routes: airborne, droplet-borne, waterborne, foodborne, biological vector-borne, bodily fluid-borne, and contact-borne. The extraction of the seasonal feature vector specifically includes: integrating temperature, humidity and rainfall information from meteorological data, combining holiday traffic change data, and using time series models to quantitatively analyze the fluctuating impact of different seasons (spring, summer, autumn and winter) on the risk of common infectious diseases.

3. The infection risk assessment method for infectious disease nursing according to claim 1, characterized in that, The mixed infection risk prediction model includes a machine learning algorithm and a time series analysis model; wherein the machine learning algorithm includes XGBoost or LSTM; and the time series analysis model includes at least one of ARIMA or Prophet.

4. The infection risk assessment method for infectious disease nursing according to claim 1, characterized in that, The embedding of virtual risk points in the simulated scenario specifically includes: simulating at least one key nursing procedure such as contact with patients without washing hands, failure to wear protective equipment properly, or violation of sterile area operation rules.

5. The infection risk assessment method for infectious disease nursing according to claim 1, characterized in that, After the early warning information is triggered, the system will also link the early warning information to the hospital resource management system to provide decision-making suggestions for the reserve of protective materials and the allocation of nursing staff.

6. An infection risk assessment system for implementing the method of any one of claims 1-5, characterized in that, include: Data acquisition and processing module: used to collect and integrate multi-dimensional data, and perform standardization processing and key feature vector extraction; Risk prediction module: It has a built-in hybrid infection risk prediction model, which is used to calculate infection risk scores and levels based on key feature vectors; Virtual Reality Assessment Module: Used to generate and run nursing simulation scenarios, embed virtual risk points, and assess nursing staff's operational behaviors in real time; Early warning and intervention module: used to trigger early warnings and push intervention information based on risk scores and operational assessment results.

7. The system according to claim 6, characterized in that, The risk prediction module also includes a model optimization unit, which is used to continuously optimize the prediction accuracy by adjusting the algorithm hyperparameters and feature weights.

8. The system according to claim 6, characterized in that, The virtual reality assessment module also includes a multi-person collaboration unit, which supports multiple nursing staff to simultaneously access the same virtual scene for collaborative drills and to compare and analyze the operation data of multiple people.

9. The system according to claim 6, characterized in that, It also includes a visualization module, which displays the spatial distribution of infection risk in the ward in the form of a heat map and the seasonal trend of infection risk in the form of a line graph.

10. The system according to claim 6, characterized in that, The early warning and intervention module is connected to the hospital resource management system and can automatically generate resource allocation suggestions based on risk prediction results.

Images