A method and system for fungal infection risk prediction within a hospital setting

Abstract

This invention belongs to the field of hospital fungal infection detection technology, and relates to a method and system for predicting the risk of fungal infection in hospital settings. The method includes: when a sub-area within a ward is identified as a warning area based on a comprehensive analysis of the patient's vital signs and the historical operating records of diagnostic equipment, the invention first obtains the patient's vital signs to assess their probability of fungal infection and generates a multi-level risk classification accordingly; the patient's risk group and sub-area information are input into the analysis model to determine whether the sub-area is a normal area or a warning area. By combining the risk assessment of the patient with the risk determination of the sub-area, the warning area is identified. Compared to existing technologies that only focus on the patient, this invention can locate sub-areas with concentrated infection risk, allowing prevention and control resources to be allocated to the warning area, thereby improving the targeting and efficiency of prevention and control.

Description

Technical Field

[0001] This invention belongs to the field of fungal infection detection technology, specifically relating to a method and system for predicting the risk of fungal infection in hospital settings. Background Technology

[0002] Among various hospital-acquired infections, fungal infections are of particular concern due to their insidious onset, difficulty in diagnosis, and high mortality rate. In special environments such as intensive care units and transplant wards for immunocompromised patients, fungal infections have become one of the main risks threatening patients' lives. Therefore, how to effectively and accurately prevent and control fungal infections is of great significance for ensuring patient safety.

[0003] Existing prevention and control models generally have a certain degree of lag, as they mainly rely on monitoring patients' vital signs, such as body temperature, white blood cell count, or clinical symptoms. When these indicators become abnormal, the infection has often already progressed to a certain stage. At this time, adopting the above-mentioned passive response mechanism will cause the golden window for early intervention to be missed, which will not only increase the difficulty of treatment and the suffering of patients, but also lead to higher medical expenses and the risk of death. Furthermore, existing technologies neglect the comprehensive assessment of environmental risk factors. The growth and spread of fungi are closely related to environmental factors such as temperature and humidity, air cleanliness, and surface cleanliness in the ward. However, traditional prevention and control measures only focus on the patients themselves and lack the ability to dynamically and quantitatively monitor and analyze the fungal load and transmission risk in the environment. This makes it impossible to identify and eliminate potential infection risks at the source, resulting in prevention and control measures that only treat the symptoms and not the root cause.

[0004] In view of this, the present invention proposes a method and system for predicting the risk of fungal infection in hospital settings. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for predicting the risk of fungal infection in a hospital setting, so as to achieve prospective diagnosis and control of fungal infection.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for predicting the risk of fungal infection in a hospital setting, wherein when a sub-area within the ward area is determined to be a warning area based on a comprehensive analysis of the vital signs parameters of patients and the historical operation records of diagnostic equipment within the ward area, the following steps are executed: Identify the fungal infection transmission routes within the alert area; analyze the device usage time series corresponding to the patient risk groups within the alert area to determine the infection persistence trend; and generate and output corrected diagnostic results based on the transmission routes and infection persistence trends. The steps to identify a sub-region as a warning area include: obtaining the vital signs parameters of each patient within the sub-region and assessing the probability of fungal infection in each patient accordingly; Based on the probability of fungal infection, patients are classified into one of the low-risk, medium-risk, or high-risk groups to generate a multi-level risk classification; and the patient's vital signs, the patient's risk group, and the information of the sub-region where the patient is located are input into a preset analysis model to determine whether the sub-region is a normal region or a warning region.

[0007] Preferably, the vital signs parameters include at least one of body temperature, redness and swelling, sweating, and rash information; and the step of assessing the probability of fungal infection in each patient includes: The probability of fungal infection is calculated using a preset formula, and then compared with a preset standard sample set to generate a risk percentage. The risk percentage is then mapped to a risk group based on a preset risk threshold.

[0008] Preferably, the step of determining the fungal infection entry path includes: screening the historical operation records of each diagnostic device within the alert area to identify the entry and exit nodes; Specifically, if a diagnostic device does not have a shutdown record in its historical operation record, or its shutdown record does not meet the preset risk marking conditions, the diagnostic device will be identified as an incoming node. The incoming and outgoing nodes are sorted according to their recorded operation times in the historical operation to construct the incoming path; and the time interval between adjacent nodes in the incoming path is calculated.

[0009] Preferably, the step of analyzing the equipment usage time series includes: within a preset diagnostic period, at each sampling time point at a preset sampling interval, obtaining the equipment usage intensity of each diagnostic device, so as to generate multiple equipment usage time series corresponding to each risk group in the multi-level risk classification; The equipment usage intensity is calculated based on the amount of equipment imported, the downtime, and the total usage extracted from historical operation records.

[0010] Preferably, the step of determining the persistence trend of infection includes: calculating the offset of equipment usage intensity between multiple sampling time points; and calculating its statistical mean based on the offset of equipment usage intensity to determine the trend of the offset. Based on a preset threshold range, the trend of the offset change is classified into one of three categories: abnormal upward trend, downward trend, or stable trend, and the classification result is used as the infection persistence trend.

[0011] Preferably, the step of generating and outputting the corrected diagnostic results includes: generating and outputting the corrected diagnostic results based on the probability of fungal infection, multi-level risk classification, time interval, and infection persistence trend.

[0012] A fungal infection risk prediction system for hospital settings includes: The area status determination module is used to acquire the vital signs parameters of patients and the historical operation records of diagnostic equipment within the ward area, and to determine whether a certain sub-area within the ward area is a warning area through comprehensive analysis in response to the vital signs parameters and historical operation records. The infection source tracing and trend analysis module is used to determine the alert area in response to the status determination module, identify the fungal infection transmission path within the alert area, and analyze the device usage time series corresponding to the patient risk group within the alert area to determine the infection persistence trend. The diagnosis correction module is used to generate and output corrected diagnostic results based on the transmission path and infection persistence trend determined by the infection tracing and trend analysis module.

[0013] Preferably, the regional status determination module is specifically used to: acquire the patient's vital signs parameters to assess the probability of fungal infection; Based on the probability of fungal infection, patients are classified into one of the low-risk, medium-risk, or high-risk groups to generate a multi-level risk classification; and a preset analysis model is invoked to determine whether a sub-region is a warning area by combining the risk group information.

[0014] Preferably, the infection tracing and trend analysis module is specifically used for: screening the historical operation records of diagnostic equipment within the alert area to identify incoming and outgoing nodes; and sorting the incoming and outgoing nodes according to their operation time to construct the incoming path; The system acquires the usage intensity of each diagnostic device to generate a device usage time series; and calculates the offset of device usage intensity in the device usage time series to determine the trend of the offset and thus determine the persistence trend of infection.

[0015] Preferably, the diagnostic correction module is specifically used to: output a corrected diagnostic result based on the probability of fungal infection, multi-level risk classification, time interval, and infection persistence trend. Beneficial effects

[0016] 1. This invention first obtains the patient's vital signs parameters to assess their probability of fungal infection and generates a multi-level risk classification accordingly. The patient's risk group and sub-region information are input into the analysis model to determine whether the sub-region is a normal region or a warning region. By combining the risk assessment of the patient with the risk determination of the sub-region, the warning region can be identified. Compared with existing technologies that only focus on the patient, this invention can locate sub-regions with concentrated infection risk, so that prevention and control resources can be deployed to the warning region, thereby improving the targeting and efficiency of prevention and control.

[0017] 2. This invention targets sub-regions identified as alert areas. By analyzing the historical operation records of diagnostic equipment, it determines the fungal infection transmission path, which includes the input and output nodes, and calculates the time intervals in the transmission path. By constructing the fungal infection transmission path, this invention traces the potential spread of infection, clarifies the fungal infection transmission path, input nodes, output nodes, and the continuous trend of infection, and overcomes the shortcomings of existing technologies in determining the source and mode of infection.

[0018] 3. This invention constructs a time series of device usage corresponding to multi-level risk classification within the diagnostic period and calculates dynamic indicators such as the deviation of device usage intensity and the trend of infection persistence. Then, based on the probability of fungal infection, multi-level risk classification, time interval, and infection persistence trend, it outputs the corrected diagnostic results and performs dynamic and trend analysis on infection risk, rather than static assessment. By analyzing dynamic indicators such as the deviation of device usage intensity, the evolution law of infection risk is revealed, which can intervene in the development of the patient's condition in advance. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a system module diagram of the present invention. Detailed Implementation

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Example

[0021] Please see Figure 1 This embodiment provides a method for predicting the risk of fungal infection in a hospital setting. By performing spatiotemporal correlation analysis on patient vital signs and the historical operation records of diagnostic equipment, it enables dynamic analysis and tracing of potential infection risks.

[0022] Specifically, the steps include the following: Obtain vital signs parameters of each patient in the ward area: Vital signs parameters serve as the basic data for assessing the patient's health status. In this embodiment, they specifically include at least one of the following: body temperature, skin redness and swelling, abnormal sweating, and rash information, which are typical clinical indicators related to fungal infection. Based on the vital signs parameters, the probability of fungal infection in each patient is assessed through a preset quantitative scoring rule, and the likelihood of fungal infection in the patient's current state is evaluated. The measured values ​​or state descriptions of each vital sign parameter are mapped to corresponding numerical scores, and all scores are then weighted and summed. The weight coefficients are preset based on historical clinical statistical data and reflect the numerical values ​​of the correlation between different vital signs and fungal infection. The weighted sum is used to determine the probability of fungal infection for each patient. Based on this probability, patients are divided into different risk groups. The calculated probability of fungal infection is then compared with two preset risk thresholds, such as a low-risk threshold. and high risk threshold If the probability is lower than If the probability is between [missing information], the patient will be classified into the low-risk group; if the probability is between [missing information], the patient will be classified into the low-risk group. and If the probability is between [a certain value] and [a certain value], then it is classified into the medium-risk group; if the probability is higher than [a certain value], then it is classified into the medium-risk group. If the risk is high, it will be classified into the high-risk group, thus generating a multi-level risk classification that reflects the overall risk distribution of the ward area.

[0023] Input: Numerical scores corresponding to the patient's N vital signs parameters (in ).

[0024] Output: Probability of fungal infection in the patient .

[0025] The formula uses the Sigmoid function to map the weighted sum to the (0,1) interval, making it conform to the definition of probability.

[0026] In the formula, This indicates the probability of fungal infection, meaning the quantitative likelihood that a patient will be infected with fungi, and serves as the basis for subsequent risk stratification. This represents the score of the i-th vital sign parameter, which is the value converted from the original i-th vital sign parameter according to the quantitative scoring rules. denoted as the weight coefficient of the i-th item, which is a pre-defined coefficient based on historical clinical statistics, reflecting the strength of the association between the i-th physical sign parameter and fungal infection; This represents the bias term, which is the bias parameter of the model used to adjust the baseline threshold of the activation function and is an inherent part of the model. This indicates the total number of vital signs parameters, which means the total number of vital signs parameters used in this assessment.

[0027] Furthermore, based on the hospital's spatial layout plan, the ward area is divided into multiple independent sub-areas, which serve as the basic unit for spatial risk analysis; for example, it could be an independent ward, several adjacent beds managed by a specific nursing station, or a specific functional area such as a treatment room. Within a preset diagnostic cycle, historical operation records of diagnostic devices in multiple sub-regions are retrieved and integrated. Diagnostic devices may include, but are not limited to, ventilators, monitors, infusion pumps, and other diagnostic devices that come into contact with patients. The historical operation records detail the start and stop times, usage duration, sub-region, and operator information of each device, facilitating subsequent spatiotemporal correlation analysis.

[0028] Furthermore, based on the patient risk group to which the patient belongs in each sub-region, the regional risk level of that sub-region is initially determined; for example, if there is at least one patient belonging to the high-risk group in a sub-region, then the regional risk level of that sub-region is marked as high. Then, based on the preset comprehensive judgment rule set, the status of each sub-region is finally judged to determine whether it is a normal region or an alert region that needs to initiate further input path tracing procedures. The judgment rule set comprehensively considers the risk level of individual patients in the region, such as whether there are high-risk patients, and the risk clustering characteristics of patient groups, such as the number of medium-risk patients and the average infection probability of the region, to identify physical spaces with potential transmission risks. This judgment rule set integrates patient risk and regional clustering characteristics. For example, a sub-region will be identified as an alert region when it meets at least one of the following conditions to identify physical spaces with potential concentrated outbreaks or transmission risks. 1. This sub-region contains at least one patient in the high-risk group; 2. The number of patients in the medium-risk group within this sub-region exceeds the preset threshold. 3. The average probability of fungal infection among all patients in this sub-region exceeds the preset probability threshold.

[0029] Furthermore, for all sub-areas identified as alert zones, an entry path tracing procedure is initiated to identify potential infection transmission chains; this is accomplished by screening the historical operation records of each diagnostic device within the alert zone during the diagnostic cycle, and tracing its flow process in time and space. Specifically, for any diagnostic device within the alert area, its previous historical operation record is traced: if the last location of the device was in the "source sub-region" containing high-risk patients, and the time interval from the device being deactivated in the "source sub-region" to its activation in the current alert area is less than a preset time threshold, then a high-risk device transfer event is determined to have occurred. In this case, diagnostic devices are identified as a key component of the fungal infection transmission path. Their usage events in the "source sub-region" are marked as transmission nodes, and their usage events in the current alert area are marked as output nodes. The spatiotemporal flow from the transmission node to the output node constitutes a specific fungal infection transmission path. By conducting this investigation on all relevant devices, one or more fungal infection transmission paths pointing to the alert area can be constructed.

[0030] Furthermore, the incoming and outgoing nodes in the identified fungal infection transmission pathways are precisely sorted according to their operation time, and the time interval between them is calculated to analyze the dynamic development of infection risk. At the same time, within the diagnostic cycle, a device usage time series is constructed using a preset sampling interval as the unit. Within the time slice corresponding to each sampling time point, the total usage time of each diagnostic device is counted, and this duration is used as the baseline value of device usage intensity at that time point. Based on the patient risk group to which the patient using the device belongs, the baseline value of device usage intensity is adjusted so that the intensity value can more accurately reflect the risk level. This adjustment process follows a set of preset adjustment rules. If a device has been used by a high-risk patient in the time slice, its baseline value of device usage intensity will be multiplied by a risk weight coefficient greater than 1, thereby generating the adjusted device usage intensity. This process is performed on all devices, ultimately generating multiple adjusted device usage intensity time series corresponding to each patient risk group in the multi-level risk classification.

[0031] Furthermore, before outputting the corrected diagnostic results, a deep trend analysis is performed on the generated corrected equipment usage intensity time series; by calculating the change in corrected equipment usage intensity between multiple consecutive sampling time points, i.e. the equipment usage intensity offset, the instantaneous fluctuation of risk is reflected. Then, the infection persistence trend is calculated by analyzing the modified time series of device usage intensity over a recent period and calculating the overall slope of change or the average of the first difference. The result is compared with the preset trend judgment threshold range, and the change trend is classified into one of three categories: abnormal upward trend, downward trend, or stable trend. This classification result is taken as the infection persistence trend. The duration of infection is then calculated, from the identification of the entry node of the first fungal infection entry path to the current analysis time point, totaling the cumulative duration. Finally, the analysis results of multiple dimensions, including fungal infection probability, multi-level risk classification, time interval, infection trend, equipment usage intensity deviation, infection persistence trend, and infection duration, are integrated to generate a structured and corrected diagnostic result. This result clearly identifies the warning area and shows the potential fungal infection entry path, risk development speed, and dynamic change trend. Example

[0032] Please see Figure 2 This embodiment provides a fungal infection risk prediction system for hospital settings. By comprehensively analyzing the vital signs of patients in the ward area and the historical operation records of diagnostic equipment, it dynamically identifies potential warning areas, further traces the fungal infection entry path, analyzes the infection persistence trend, and generates a corrected diagnostic result.

[0033] In its implementation, this system can be deployed on hospital servers, workstations, or cloud platforms. It interacts with the Hospital Information System (HIS), Electronic Medical Record (EMR) system, and Internet of Things (IoT) gateway through standard data interfaces to obtain patient information and device status. The system specifically includes the following modules: The regional status determination module continuously monitors the infection risk status within the ward area and determines whether a certain sub-area is a warning area; it obtains the vital signs parameters of each patient in the specified sub-area through the data interface. The vital signs parameters may include at least one of the following: body temperature, skin redness and swelling, abnormal sweating, and rash information. At the same time, it obtains the historical operation records of each diagnostic device in the sub-area. Then, based on the acquired vital signs parameters, the probability of fungal infection for each patient is assessed; using a preset calculation formula, the patient's various vital signs parameters are quantified and an initial risk score is calculated; then, this score is compared and analyzed with a preset standard sample set to generate a risk percentage; based on this risk percentage and referring to a preset risk cutoff value, the patient is divided into a low-risk group, a medium-risk group, or a high-risk group, thereby completing the multi-level risk classification of the patient. The division method described here is only a preferred implementation method and does not constitute a limitation of the present invention. Finally, the patient's vital signs, the patient's risk group, and the sub-region information of the patient are used as inputs and provided to a preset analysis model. For example, a machine learning classification model trained on historical data performs comprehensive analysis on the above input information and outputs a judgment result to determine whether the sub-region is a normal region or a warning region that requires special attention. When the sub-region is determined to be a warning region, the operation of subsequent modules will be triggered.

[0034] The infection tracing and trend analysis module is activated after the regional status determination module identifies a sub-region as a warning area; it determines the fungal infection entry path and analyzes the infection persistence trend, screens the historical operation records of each diagnostic device within the warning area, and determines the fungal infection entry path; in this process, it is committed to identifying the entry node and the output node; Specifically, when a diagnostic device has no shutdown records in its historical operation records, or its shutdown records do not meet the preset risk marking conditions, the diagnostic device is identified as a potential entry node. The output node can be a patient who has had contact with the device and was subsequently classified into a medium-risk or high-risk group, or the device that has been removed from the area. Then, all identified entry and output nodes are sorted according to the operation time recorded in their historical operation records to construct one or more fungal infection entry paths. After the path is constructed, the time interval between adjacent nodes in the path is further calculated to facilitate subsequent risk assessment. Device usage time series corresponding to each patient risk group in the multi-level risk classification are constructed to analyze the infection persistence trend. Within a preset diagnostic cycle, the device usage intensity of each diagnostic device is obtained at each sampling time point at a preset sampling interval. The device usage intensity can be calculated based on data extracted from historical operation records, including device import volume, downtime, and total usage. This generates multiple device usage time series reflecting the interaction between patients and devices in different patient risk groups. Then, based on these time series, the infection persistence trend is determined, and the offset of device usage intensity between multiple consecutive sampling time points is calculated. Then, the statistical mean of the intensity offset of the device is calculated to determine the trend of the offset. Finally, based on the preset threshold range, the trend of the offset is classified into one of three categories: abnormal upward trend, downward trend, or stable trend. This classification result is used as the infection persistence trend for the diagnostic correction module.

[0035] The diagnosis correction module, as the decision output unit of the entire system, outputs the corrected diagnosis results based on the analysis results of the preceding modules; it also collects and comprehensively analyzes all key information generated by other modules; specifically including: the fungal infection probability and multi-level risk classification results of each patient generated by the regional status determination module, as well as the time interval in the fungal infection transmission path determined by the infection source tracing and trend analysis module and the infection persistence trend obtained from the analysis. Based on the aforementioned multi-dimensional input information, a structured and revised diagnostic result is generated. This is no longer an isolated judgment for a single patient, but a macro-level risk profile of the entire alert area. Specifically, it includes: clearly identifying the current alert area, listing the specific fungal infection transmission routes, highlighting diagnostic equipment as key transmission nodes, identifying the most severely affected high-risk patient groups, and providing early warnings about future risk evolution based on infection trend information. This revised diagnostic result can be formatted into reports, alerts, or visualizations and pushed to professionals in the hospital infection control department to provide data support for them to take precise intervention measures.

[0036] Through the collaborative work of the above modules, the system in this embodiment can dynamically, multidimensionally, and proactively predict and analyze the risk of fungal infection in hospital settings. This upgrades the traditional passive diagnosis based on individual symptoms to a systematic proactive early warning based on the interaction of environment, equipment, and people, enabling early detection of potential infection transmission chains and prediction of their development trends.

Claims

1. A method for predicting the risk of fungal infection in a hospital setting, characterized in that, When a sub-area within the ward area is designated as a warning zone based on a comprehensive analysis of patients' vital signs and the historical operational records of diagnostic equipment, the following actions are taken: Determine the fungal infection transmission route within the alert area; analyze the device usage time series corresponding to the patient risk groups within the alert area to determine the infection persistence trend; And based on the transmission path and the ongoing infection trend, generate and output corrected diagnostic results; The steps to identify a sub-region as a warning area include: obtaining the vital signs parameters of each patient within the sub-region and assessing the probability of fungal infection in each patient accordingly; Based on the probability of fungal infection, patients are classified into one of the low-risk, medium-risk, or high-risk groups to generate a multi-level risk classification; and the patient's vital signs, the patient's risk group, and the information of the sub-region where the patient is located are input into a preset analysis model to determine whether the sub-region is a normal region or a warning region.

2. The method for predicting the risk of fungal infection in a hospital setting according to claim 1, characterized in that, Vital signs parameters include at least one of the following: body temperature, redness and swelling, sweating, and rash information; Furthermore, the steps for assessing the probability of fungal infection in each patient include: calculating the probability of fungal infection using a preset calculation formula, comparing the probability of fungal infection with a preset standard sample set to generate a risk percentage, and then mapping the risk percentage to a risk group based on a preset risk threshold.

3. The method for predicting the risk of fungal infection in a hospital setting according to claim 1, characterized in that, The steps to determine the entry path of fungal infection include: screening the historical operation records of each diagnostic device within the alert area to identify the entry and exit nodes; Specifically, if a diagnostic device does not have a shutdown record in its historical operation record, or its shutdown record does not meet the preset risk marking conditions, the diagnostic device will be identified as an incoming node. The incoming and outgoing nodes are sorted according to their recorded operation times in the historical operation to construct the incoming path; and the time interval between adjacent nodes in the incoming path is calculated.

4. The method for predicting the risk of fungal infection in a hospital setting according to claim 1, characterized in that, The steps involved in using time series analysis equipment include: Within a preset diagnostic cycle, at a preset sampling interval at each sampling time point, the equipment usage intensity of each diagnostic device is obtained to generate multiple equipment usage time series corresponding to each risk group in the multi-level risk classification. The equipment usage intensity is calculated based on the amount of equipment imported, the downtime, and the total usage extracted from historical operation records.

5. The method for predicting the risk of fungal infection in a hospital setting according to claim 4, characterized in that, The steps to determine the persistence trend of infection include: calculating the magnitude of the offset in device usage intensity between multiple sampling time points; Based on the offset magnitude of equipment usage intensity, its statistical mean is calculated to determine the trend of offset change; and according to the preset threshold range, the trend of offset change is classified into one of three types: abnormal upward trend, downward trend, or stable trend, and the classification result is used as the infection persistence trend.

6. The method for predicting the risk of fungal infection in a hospital setting according to claim 5, characterized in that, The steps to generate and output the corrected diagnostic results; This includes generating and outputting corrected diagnostic results based on fungal infection probability, multi-level risk classification, time interval, and infection persistence trend.

7. A fungal infection risk prediction system for use in hospital settings, characterized in that, include: The area status determination module is used to acquire the vital signs parameters of patients and the historical operation records of diagnostic equipment within the ward area, and to determine whether a certain sub-area within the ward area is a warning area through comprehensive analysis in response to the vital signs parameters and historical operation records. The infection source tracing and trend analysis module is used to determine the alert area in response to the status determination module, identify the fungal infection transmission path within the alert area, and analyze the device usage time series corresponding to the patient risk group within the alert area to determine the infection persistence trend. The diagnosis correction module is used to generate and output corrected diagnostic results based on the transmission path and infection persistence trend determined by the infection tracing and trend analysis module.

8. A fungal infection risk prediction system for hospital settings according to claim 7, characterized in that, The regional status determination module is specifically used to: acquire the patient's vital signs parameters to assess the probability of fungal infection; Based on the probability of fungal infection, patients are divided into one of the low-risk, medium-risk, or high-risk groups to generate a multi-level risk classification. It also invokes a preset analysis model to determine whether a sub-region is a warning zone by combining risk group information.

9. A fungal infection risk prediction system for hospital settings according to claim 7, characterized in that, The infection tracing and trend analysis module is specifically used to: screen the historical operation records of diagnostic equipment within the alert area to identify incoming and outgoing nodes; and sort the incoming and outgoing nodes according to their operation time to construct the incoming path. The system acquires the usage intensity of each diagnostic device to generate a device usage time series; and calculates the offset of device usage intensity in the device usage time series to determine the trend of the offset and thus determine the persistence trend of infection.

10. A fungal infection risk prediction system for hospital settings according to claim 7, characterized in that, The diagnostic correction module is specifically used for: Based on the probability of fungal infection, multi-level risk classification, time interval, and infection persistence trend, the corrected diagnostic results are output.

Images