An outpatient hemodialysis infection event monitoring system and monitoring method

Abstract

The application provides an outpatient hemodialysis infection event monitoring system and a monitoring method, and is applied to the technical field of data processing. The application processes blood vessel access information of a target patient and outpatient hemodialysis cannulation records of the target patient, generates a hemodialysis event influence factor, processes dialysis physiological parameter information of the target patient within a preset time, generates hemodialysis monitoring parameter information of the target patient, processes historical infection data of the target hemodialysis patient, generates a hemodialysis prediction monitoring characteristic vector, processes the hemodialysis event influence factor, the hemodialysis monitoring parameter information of the target patient, the hemodialysis prediction monitoring characteristic vector and infection examination result information of the target patient based on a target hemodialysis event early warning model, and generates hemodialysis infection event early warning information.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a monitoring system and method for monitoring infection events in outpatient hemodialysis. Background Technology

[0002] Hemodialysis is a vital treatment for end-stage renal disease patients, with a large number of them undergoing outpatient hemodialysis. However, outpatient hemodialysis patients currently face a high risk of infection, which has become a key factor affecting patient health and the effectiveness of dialysis treatment.

[0003] Traditional monitoring methods are insufficient for comprehensively, timely, and accurately monitoring patients' infection risks. Existing monitoring methods often rely on a single indicator or simple observation of patient signs, failing to comprehensively consider multiple complex factors. For example, focusing only on changes in the patient's body temperature ignores factors such as vascular access status, dialysis equipment, and the patient's own history of infection, resulting in the inability to detect potential infection risks in advance, making it difficult for patients to receive timely intervention and treatment in the early stages of infection.

[0004] There is an urgent clinical need for a more effective monitoring model for infection events in outpatient hemodialysis, in order to achieve accurate monitoring and early warning of infection risks in outpatient hemodialysis patients, reduce the incidence of infection, and ensure the safety of patients' dialysis.

[0005] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore includes information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this application is to provide an outpatient hemodialysis infection event monitoring system and method, which at least to some extent overcomes the problems existing in the prior art. By acquiring multi-faceted data of target patients, processing vascular access and outpatient hemodialysis catheterization records to generate hemodialysis event influencing factors; processing dialysis physiological parameters to obtain hemodialysis monitoring parameter information; and analyzing historical infection data to generate hemodialysis predictive monitoring feature vectors, this provides a multi-dimensional basis for subsequent infection risk assessment, enabling effective monitoring and early warning of infection risks in outpatient hemodialysis patients, and helping to promptly identify potential infection risks.

[0007] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0008] According to one aspect of this application, a method for monitoring infection events in outpatient hemodialysis is provided, comprising: acquiring dialysis physiological parameter information, vascular access information, outpatient hemodialysis catheterization records, infection test results, and historical infection data of the target hemodialysis patient within a preset time period for a target patient undergoing maintenance hemodialysis; processing the vascular access information and outpatient hemodialysis catheterization records of the target patient to generate hemodialysis event influencing factors; processing the dialysis physiological parameter information of the target patient within the preset time period to generate hemodialysis monitoring parameter information of the target patient; processing the historical infection data of the target hemodialysis patient to generate a hemodialysis predictive monitoring feature vector; and processing the hemodialysis event influencing factors, the hemodialysis monitoring parameter information, the hemodialysis predictive monitoring feature vector, and the infection test results of the target patient based on a target hemodialysis event early warning model to generate hemodialysis infection event early warning information.

[0009] Another aspect of this application discloses a monitoring device for infection events in outpatient hemodialysis, characterized by comprising: an acquisition module for acquiring dialysis physiological parameter information, vascular access information, outpatient hemodialysis intubation records, infection test results, and historical infection data of a target patient undergoing maintenance hemodialysis within a preset time period; and a processing module for processing the vascular access information and outpatient hemodialysis intubation records of the target patient to generate hemodialysis event influencing factors; processing the dialysis physiological parameter information of the target patient within the preset time period to generate hemodialysis monitoring parameter information of the target patient; processing the historical infection data of the target patient to generate a hemodialysis predictive monitoring feature vector; and processing the hemodialysis event influencing factors, hemodialysis monitoring parameter information, hemodialysis predictive monitoring feature vector, and infection test results of the target patient based on a target hemodialysis event early warning model to generate hemodialysis infection event early warning information.

[0010] According to another aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a second processor, implements the above-described method for monitoring infection events during outpatient hemodialysis.

[0011] This application provides an outpatient hemodialysis infection event monitoring system and method. The server acquires multifaceted data from target patients, such as dialysis physiological parameters and vascular access information. Hemodialysis event influencing factors are generated by processing vascular access and outpatient hemodialysis catheterization records; hemodialysis physiological parameters are processed to obtain hemodialysis monitoring parameter information; and historical infection data is analyzed to generate a hemodialysis predictive monitoring feature vector. These data processing steps provide multi-dimensional basis for subsequent infection risk assessment. An infection risk score is generated by integrating various data using a model to further determine the infection risk level, and an early warning information for hemodialysis infection events is issued based on a preset threshold.

[0012] In-depth analysis of early warning information identifies the source and transmission route of the infectious pathogen, such as determining whether the infection is caused by catheter placement procedures or dialysis equipment. Based on this information, target triage information and response priority information are generated to guide subsequent targeted measures. Finally, based on the above information, early warning measures for hemodialysis infection are formulated, such as adjusting treatment plans and strengthening equipment disinfection, to reduce the risk of infection. Through comprehensive data processing and multi-step analysis and evaluation, effective monitoring and early warning of infection risks for outpatient hemodialysis patients can be achieved. This helps to promptly identify potential infection risks, provides strong support for medical staff to take appropriate measures, and is of great significance for ensuring patient dialysis safety and reducing the incidence of infection.

[0013] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0014] Figure 1 A flowchart illustrating a method for monitoring infection events in outpatient hemodialysis provided in an embodiment of this application is shown.

[0015] Figure 2 A schematic diagram of a monitoring device for infection events in outpatient hemodialysis provided in one embodiment of this application is shown. Detailed Implementation

[0016] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0017] The following is combined with Figure 1 This application describes a method for monitoring infection events during outpatient hemodialysis according to exemplary embodiments. In one embodiment, this application also provides an outpatient hemodialysis infection event monitoring system and method. Figure 1 A schematic flowchart illustrating a method for monitoring infection events in outpatient hemodialysis according to an embodiment of this application is shown. Figure 1 As shown, this method is applied to a server and includes:

[0018] S101, acquire the dialysis physiological parameters of the target patient undergoing maintenance hemodialysis within a preset time period, the vascular access information of the target patient, the outpatient hemodialysis catheterization record of the target patient, the infection test results of the target patient, and the historical infection data of the target patient undergoing hemodialysis.

[0019] In one implementation, a comprehensive monitoring system is installed in the outpatient hemodialysis center of a hospital to comprehensively monitor patients undergoing maintenance hemodialysis. Taking Mr. Li as an example, the system details how various data are acquired: the dialysis equipment is connected to the hospital's information system in real time. During each of Mr. Li's dialysis sessions, the system automatically collects multiple physiological parameters. For instance, at the start of a dialysis session, the blood flow rate is stable at 250 ml / min, the dialysate flow rate is 500 ml / min, the venous pressure is maintained at approximately 150 mmHg, and the body temperature is 36.5°C. As dialysis progresses, these parameters change in real time, and the system continuously records them. After each dialysis session, all physiological parameter data from that session, along with a timestamp accurate to the second, such as "blood flow rate, dialysate flow rate, venous pressure, and body temperature data from 10:00:00 on October 15, 2024 to 12:30:00 on October 15, 2024," are stored in a long-term monitoring database. These data not only record the current dialysis status, but also provide a basis for subsequent analysis, making it easier for medical staff to observe the changing trends of Mr. Li's various physiological indicators at different dialysis periods.

[0020] Mr. Li used a tunneled polyester-coated central venous catheter (TCC) as his vascular access. This information was recorded in detail in the system, including the catheter placement location (e.g., right internal jugular vein) and the placement date (August 10, 2024). In addition, medical staff regularly checked the condition of the vascular access, such as whether the blood vessels were patent, and whether there were any signs of stenosis or thrombosis; these results were also promptly entered into the system. For example, during a check in October 2024, it was found that Mr. Li's blood flow velocity in his vascular access was slightly below the normal range; this information was recorded to provide a reference for subsequent assessment of infection risk. Mr. Li's outpatient hemodialysis catheterization records detailed information about each catheterization. Recently, due to treatment needs, Mr. Li underwent a catheter replacement procedure on October 5, 2024, changing from a long-term catheter (TCC) to a temporary catheter (ECC); this change in catheter type was accurately recorded. Before and after each dialysis session, medical staff evaluated the catheter, and the evaluation results were also recorded. For example, during the predialysis assessment on October 15, 2024, slight redness was found around the catheter puncture site, but there was no bleeding or discharge. These catheter assessment results provided important evidence for judging the risk of infection.

[0021] Mr. Li undergoes regular bloodborne pathogen screening, specifically the eight tests for postoperative immune system infections. His most recent test was on October 10, 2024, which showed negative results for hepatitis B surface antigen, hepatitis C antibody, HIV antibody, and syphilis-specific antibody. Past test results are also stored in the system for easy comparison and analysis. If a test result shows an abnormality, such as a change from negative to positive for hepatitis B surface antigen, the system will immediately detect this change and further assess the infection risk based on other data. The system stores Mr. Li's past infection-related data. In the past year, in March 2024, Mr. Li experienced fever symptoms due to a vascular access infection, with a temperature reaching 37.8℃. Blood culture results showed Staphylococcus aureus infection. He received antibacterial treatment at that time, using cephalosporin antibiotics for 10 days. This historical infection data, including infection time, symptoms, pathogens, treatment measures, and treatment duration, is crucial for analyzing Mr. Li's current infection risk. It helps healthcare professionals understand his infection patterns and susceptibility, thereby more accurately assessing his current infection risk.

[0022] In another implementation, acquiring the target hemodialysis event early warning model includes acquiring a training sample set and a preset hemodialysis event early warning model. The training sample set includes dialysis physiological parameters, vascular access information, outpatient hemodialysis catheterization records, infection test results, and historical infection data of other hemodialysis patients within a preset time period. Specifically, the hospital information system stores various types of data from numerous patients undergoing maintenance hemodialysis. Relevant data from 1000 patients over the past 12 months are selected as the training sample set. These data encompass dialysis physiological parameters, such as patient A's blood flow rate of 280 ml / min, dialysate flow rate of 500 ml / min, venous pressure of 160 mmHg, and body temperature of 36.8℃ during a particular dialysis session; vascular access information, such as patient B using an extracorporeal catheter (ECC) placed in the femoral vein; outpatient hemodialysis catheterization records, such as patient C undergoing two catheterizations within 30 days, with the catheter type changing from a long-term catheter (TCC) to an extracorporeal catheter (ECC); infection test results, such as patient D's positive hepatitis B surface antigen in the eight postoperative immunoglobulin tests; and historical infection data for hemodialysis patients, such as patient E's infection six months prior due to hemodialysis, exhibiting symptoms like fever and E. coli in blood cultures. Simultaneously, a logistic regression model was selected as the pre-set hemodialysis event early warning model, providing a framework for preliminary data analysis and infection risk prediction.

[0023] The number of each data feature in the training sample set is counted, and a sampling ratio is generated based on these counts. Dialysis physiological parameters include four main features: blood flow, dialysate flow, venous pressure, and body temperature; vascular access information includes three features: catheter type and catheter location; outpatient hemodialysis catheterization records include four features: number of catheter insertions and changes in catheter type; infection test results involve eight features from the eight postoperative immune markers; and historical infection data for hemodialysis patients includes five features: number of infections and type of infectious pathogen. Based on the number of these features, and considering the importance of different features and data distribution, a sampling ratio is generated according to the proportion of each feature to the total number of features. Assuming that the number of dialysis physiological parameter features accounts for 20% of the total number of features, its sampling ratio is set to 0.2.

[0024] The training sample set is sampled based on a sampling ratio to generate a predetermined number of sampled features. From the data of 1000 patients, dialysis physiological parameter data of 200 patients are randomly selected according to a sampling ratio of 0.2 for dialysis physiological parameter information; vascular access data of a corresponding number of patients are selected according to a sampling ratio for vascular access information, and so on, sampling is performed for each data type, ultimately generating a predetermined number of sampled features. These sampled features constitute a new dataset containing representative data extracted from the original training sample set.

[0025] Based on any data feature and each sampling feature, the data is processed to divide it into a response group and a poor response group. Each group contains a predetermined number of data samples, and at least one data sample carries identification information. For example, the data feature of "body temperature" in dialysis physiological parameters is selected and processed with each sampling feature. Using 37.3℃ as the cutoff, patient data with a body temperature of 37.3℃ or higher and accompanied by other infection-related signs (such as positive blood culture, abnormal white blood cell count, etc.) are classified into the poor response group; patient data with a body temperature below 37.3℃ and no other obvious signs of infection are classified into the response group. Each group contains 200 data samples, and a patient ID is added to each data sample as identification information for subsequent tracking and analysis.

[0026] The pre-set hemodialysis event early warning model is processed based on the response group and the poor response group to generate a trained hemodialysis event early warning model and training results. If the data samples containing labeled information in the training results are factors characterizing the risk of hemodialysis infection, then the trained hemodialysis event early warning model is used as the target hemodialysis event early warning model. The model continuously adjusts its parameters (such as regression coefficients) by analyzing the relationship between various features and infection risk in the two sets of data. During training, the model learns the association between factors such as elevated body temperature, frequent changes in catheter type, and positive blood pathogen screening and the risk of hemodialysis infection. After training, the trained hemodialysis event early warning model and training results are obtained. The training results include the model's evaluation of the relationship between various data features and infection risk. The data samples with labeled information in the training results are examined to determine which factors have a significant impact on the risk of hemodialysis infection. If it is found that data samples containing labeled information in the training results, such as "body temperature higher than 37.3℃", "catheter type changes more than twice in three months", and "positive for one of the eight indicators in the surgical immunization test", can indeed effectively characterize factors of hemodialysis infection risk, then the trained logistic regression model is determined as the target hemodialysis event early warning model. This model can be used to more accurately predict and warn of infection risks in other outpatient hemodialysis patients.

[0027] S102 processes the vascular access information and outpatient hemodialysis catheterization records of the target patient to generate hemodialysis event influencing factors.

[0028] In one implementation, the outpatient hemodialysis catheterization records of the target patient are processed to obtain information on changes in catheter type and catheter evaluation results. Mr. Li's outpatient hemodialysis catheterization record details the relevant information for each catheterization. Initially, Mr. Li used a long-term catheter (TCC), inserted on August 10, 2024, in the right internal jugular vein. On October 5, 2024, due to treatment needs, Mr. Li's catheter was replaced with a temporary catheter (ECC). This change in catheter type was accurately recorded by the system, becoming the catheter type change information. Before and after each dialysis session, medical staff evaluate Mr. Li's catheter and record the evaluation results in the system. For example, during the pre-dialysis evaluation on October 15, 2024, slight redness was observed around the catheter puncture site, but there was no bleeding, discharge, or hematoma, and the catheter was well fixed. This evaluation information constitutes the catheter evaluation results.

[0029] Information on changes in catheter type was processed to generate catheter type data for different time periods. From August 10th to October 4th, 2024, Mr. Li's catheter type was a long-term catheter (TCC); from October 5th, 2024, the catheter type changed to a temporary catheter (ECC). This processing clearly presents Mr. Li's catheter type at different time periods, facilitating subsequent analysis of the impact of different catheter types on infection risk. Catheter assessment results were processed to generate hematoma and bleeding information. In Mr. Li's catheter assessment on October 15th, 2024, no hematoma was clearly recorded, so the hematoma information was "no hematoma." Simultaneously, the assessment results showed no bleeding, therefore the bleeding information was "no bleeding." This information will serve as one of the important bases for assessing infection risk.

[0030] Information on catheter type, hematoma status, and bleeding status at different time points is processed to generate impact degree information. Among this, imaging quality information is used to characterize the impact weights corresponding to different catheter type changes and catheter assessment results. Changing the catheter type from long-term catheterization (TCC) to temporary catheterization (ECC) may increase the risk of infection, so this change is assigned a higher impact weight; conversely, the absence of hematoma and bleeding has a relatively smaller impact on the risk of infection, and is assigned a lower impact weight. For example, based on certain algorithms or expert experience, the impact weight for changing the catheter type (TCC to ECC) is set to 0.8, the impact weight for the absence of hematoma is set to 0.1, and the impact weight for the absence of bleeding is set to 0.1. This weighted information constitutes the impact degree information, used to quantify the degree of influence of different situations on the risk of hemodialysis infection.

[0031] The vascular access information and imaging data of the target patient were processed to generate a hemodialysis event influencing factor. In assessing Mr. Li's risk of hemodialysis infection, the hemodialysis event influencing factor was a key indicator, integrating multiple factors related to infection risk. Calculating this factor required considering factors such as changes in catheter type, hematoma status, bleeding, and vascular access location information, assigning a corresponding weight to each factor. Let the hemodialysis event influencing factor be I; the weight of the change in catheter type be ω1, set at 0.8; the weight of the hematoma status be ω2, set at 0.1; the weight of the bleeding status be ω3, set at 0.1; and the weight of the vascular access location information be ω4, assuming the vascular access location is the right internal jugular vein, with a weight of ω4 of 0.3.

[0032] Regarding the change in catheter type, Mr. Li changed from a long-term catheter (TCC) to a temporary catheter (ECC). This change has a significant impact on the risk of infection. We use x1 to represent the degree of impact of the change in catheter type, where x1 = 1 when the change occurs. Regarding hematoma, Mr. Li's catheter assessment result was no hematoma. We use x2 to represent the degree of impact of the hematoma situation, where x2 = 0 when there is no hematoma. Similarly, regarding bleeding, Mr. Li had no bleeding, so the degree of impact of bleeding is x3 = 0. Regarding the vascular access location, Mr. Li's catheter was placed in the right internal jugular vein, so we set the degree of impact of the vascular access location at this time to x4 = 1.

[0033] According to the formula I=ω1x1+ω2x2+ω3x3+ω4x4, substitute the above values ​​into the calculation:

[0034] I=0.8×1+0.1×0+0.1×0+0.3×1=1.1.

[0035] Mr. Li's current hemodialysis event impact factor is 1.1. This impact factor will be combined with other information (such as dialysis physiological parameters, infection test results, etc.) in subsequent assessments to evaluate Mr. Li's risk of hemodialysis infection. The higher the impact factor, the greater the potential risk of Mr. Li developing a hemodialysis infection. Based on this factor, Mr. Li's infection status can be more scientifically assessed, and appropriate measures can be taken in a timely manner, such as strengthening monitoring or adjusting the treatment plan.

[0036] S103 processes the dialysis physiological parameter information of the target patient within a preset time period to generate the hemodialysis monitoring parameter information of the target patient.

[0037] In one implementation, the dialysis physiological parameters of the target patient within a preset time period are processed to generate data for each dialysis session and a corresponding timestamp. When Mr. Li undergoes hemodialysis at the hospital, the dialysis equipment is connected to the hospital's information system in real time. During a dialysis session, the system automatically collects various physiological parameters from the start. For example, at the start of dialysis, the blood flow rate is 250 ml / min, the dialysate flow rate is 500 ml / min, the venous pressure is 150 mmHg, and the body temperature is 36.5℃. The time is 9:00 AM on November 1, 2024. These data, along with the time accurate to the second, "November 1, 2024, 09:00:00," are recorded. As dialysis progresses, at regular intervals (e.g., every minute), the system records a new set of parameter data and the corresponding time. For example, at 9:01 AM, the blood flow rate changes to 245 ml / min, and other parameters also change accordingly; all this data is completely recorded. After dialysis, all the data collected during this dialysis session, such as "November 1, 2024, 09:00:00 - blood flow 250 ml / min, dialysate flow 500 ml / min, venous pressure 150 mmHg, body temperature 36.5℃; November 1, 2024, 09:01:00 - blood flow 245 ml / min...", were organized into data for each dialysis session and corresponding to their respective timestamps, and stored in the long-term monitoring database.

[0038] Noise filtering and normalization are performed on each dialysis session's data to generate a sequence of dialysis parameters for the target patient. The raw data may contain noise interference, affecting its accuracy and subsequent analysis. For example, occasional fluctuations in the equipment may cause outliers in certain parameters. The system uses specific algorithms to filter noise from each dialysis session's data. For instance, a moving average method is used to process blood flow data, removing outliers that significantly deviate from the normal fluctuation range.

[0039] After noise filtering, normalization is still needed to facilitate unified analysis because different parameters have different dimensions and numerical ranges. Taking blood flow and venous pressure as examples, blood flow is generally 100-300 ml / min, and venous pressure is 50-300 mmHg; their numerical ranges and dimensions are different. A normalization formula (such as...) is used... Where X is the original data, X max and X minThis process converts parameters such as blood flow and venous pressure into a 0-1 range, representing the maximum and minimum values ​​of the parameter over a certain period. After this processing, all parameters from Mr. Li's dialysis sessions are transformed into data on a uniform scale, forming a sequence of dialysis parameters for the target patient. For example, the processed data for a particular dialysis session might appear as "[0.5 (normalized blood flow), 0.6 (normalized dialysate flow), 0.4 (normalized venous pressure), 0.3 (normalized body temperature)]", arranged chronologically by dialysis time, thus forming the parameter sequence for Mr. Li's dialysis session.

[0040] Based on preset threshold ranges for each parameter and the monitoring objectives, the system dynamically analyzes and correlates the dialysis parameter sequence information of the target patient, generating hemodialysis monitoring parameter information for the target patient. Preset threshold ranges are set for each parameter based on clinical experience and medical research. For example, the threshold range for normal blood flow is set at 200-280 ml / min, and the threshold range for venous pressure is set at 80-200 mmHg. Simultaneously, based on the monitoring objective of assessing the risk of infection during Mr. Li's dialysis process, the system performs dynamic and correlation analyses on his dialysis parameter sequence information.

[0041] In terms of dynamic analysis, the system observes the trends of parameters over time in real time. For example, if Mr. Li's blood flow, which was initially stable within the normal range during dialysis, suddenly and continuously drops below 200 ml / min for a period of time, this is considered an abnormal fluctuation, and the system will record this change. In terms of correlation analysis, the system studies the relationships between different parameters. For example, when venous pressure increases, it observes whether blood flow decreases accordingly, or whether there is a correlation between changes in body temperature and other parameters. If it finds that venous pressure increases while blood flow continues to decrease, and body temperature shows an upward trend, this suggests some potential problems.

[0042] Based on the results of combined dynamic and correlation analyses, hemodialysis monitoring parameters for the target patient are generated. If Mr. Li's dialysis parameters fluctuate within the normal threshold range, and the correlation between the parameters follows normal patterns, the hemodialysis monitoring parameters may display as "normal." However, if parameters exceed the threshold range, exhibit abnormal fluctuations, or show abnormal correlations, the hemodialysis monitoring parameters will record these abnormalities in detail, such as "blood flow is below the normal threshold at [specific time period], and venous pressure shows an abnormal negative correlation with blood flow." This hemodialysis monitoring parameter information will be used to subsequently assess Mr. Li's dialysis status and infection risk, helping doctors to promptly identify potential problems and take appropriate measures.

[0043] S104 processes historical infection data of target hemodialysis patients to generate a hemodialysis prediction and monitoring feature vector.

[0044] In one implementation, physiological parameters and a target historical time period are acquired for each dialysis session. Mr. Li undergoes regular hemodialysis at the hospital, and the hospital's monitoring system automatically records detailed physiological parameters for each session. For example, during dialysis on November 1, 2024, the initial blood flow rate was 250 ml / min, the dialysate flow rate was 500 ml / min, the venous pressure was 150 mmHg, and the body temperature was 36.5°C. As dialysis progresses, the system records a new set of parameter data at regular intervals (e.g., 1 minute). Simultaneously, to analyze the changing trends of Mr. Li's dialysis parameters, the system selects his dialysis data from the past three months as the target historical time period. During these three months, Mr. Li underwent 12 dialysis sessions, and the physiological parameters and corresponding times for each session were fully recorded in the system.

[0045] The system processes the physiological parameters and target historical time periods for each dialysis session to generate deviation values ​​between the parameters and historical data. Taking the dialysis session on November 1, 2024, as an example, the system compares each parameter during this dialysis session with the parameters at the corresponding time points within the target historical time period. For instance, if the blood flow rate was 240 ml / min at the start of dialysis (approximately 15 minutes after the start), while the average blood flow rate for the same dialysis session (approximately 15 minutes after the start) over the past three months was 255 ml / min, then the deviation value between this blood flow rate and historical data is 240 - 255 = -15 ml / min. Using the same method, the deviation values ​​of other parameters such as dialysate flow rate, venous pressure, and body temperature at various time points are calculated compared to historical data. This comprehensively reflects the changes in each parameter during this dialysis session compared to historical conditions.

[0046] The fluctuation range and degree of abnormality of parameter deviation values ​​are obtained based on the deviation values ​​between each dialysis parameter and historical data. Further analysis is performed on the parameter deviation values ​​obtained from each dialysis session. Taking blood flow rate as an example, during the dialysis session on November 1, 2024, the deviation value of blood flow rate fluctuated between -20 ml / min and 10 ml / min from start to finish. This range is the fluctuation range of blood flow rate deviation value for this dialysis session. To assess the degree of abnormality, the system sets a normal fluctuation range threshold based on statistical analysis of historical data. Assuming that, based on past experience, the normal fluctuation range of blood flow rate deviation value is between -10 ml / min and 10 ml / min, then the blood flow rate deviation value in this dialysis session exceeded the lower limit of the normal range, indicating that the change in blood flow rate has a certain degree of abnormality. The same method is used to calculate the fluctuation range and assess the degree of abnormality for other parameters, such as dialysate flow rate, venous pressure, and body temperature.

[0047] The system uses correlation and regression analysis to process the fluctuation range and abnormality of parameter deviation values, generating the correlation between various dialysis parameters and the predicted range of each parameter. Correlation analysis reveals the intrinsic relationships between different parameters. For example, analysis showed that when Mr. Li's venous pressure deviation value abnormally increased, the blood flow deviation value often simultaneously decreased abnormally, indicating a strong negative correlation between venous pressure and blood flow. Regression analysis is used to predict the future trend of parameter changes. Based on a large amount of historical data and current deviation values, the system can predict that during the next dialysis session, if other conditions remain unchanged, Mr. Li's blood flow may fluctuate within a certain range; for example, the predicted range for blood flow in the next 10 minutes is 230-250 ml / min. Similarly, similar analyses are performed on other parameters such as dialysate flow rate and body temperature to obtain their correlation and predicted ranges.

[0048] The correlation between various dialysis parameters and the predicted range of each parameter are processed to generate a hemodialysis prediction and monitoring feature vector. This feature vector is a multi-dimensional data structure that integrates the dynamic changes of various parameters during Mr. Li's dialysis process and their interrelationships. For example, the hemodialysis prediction and monitoring feature vector might be represented as [(correlation between blood flow and venous pressure: -0.8), (predicted blood flow range: 230-250 ml / min), (correlation between dialysate flow and body temperature: 0.3), (predicted dialysate flow range: 480-520 ml / min)...]. This feature vector comprehensively reflects Mr. Li's current dialysis status and potential future trends, providing crucial information for assessing infection risk based on a target hemodialysis event early warning model. Medical staff can use this feature vector to identify potential problems in advance and adjust the treatment plan promptly to reduce Mr. Li's risk of infection and other adverse events.

[0049] S105, based on the target hemodialysis event early warning model, processes the influencing factors of hemodialysis events, the hemodialysis monitoring parameter information of the target patient, the hemodialysis prediction monitoring feature vector, and the infection test results of the target patient to generate early warning information for hemodialysis infection events.

[0050] In one implementation, a target hemodialysis event early warning model is used to process hemodialysis event influencing factors, hemodialysis monitoring parameters of the target patient, hemodialysis predictive monitoring feature vectors, and infection test results of the target patient to generate an infection risk score for the target patient. The hospital's monitoring system collects Mr. Li's hemodialysis event influencing factors, hemodialysis monitoring parameters, hemodialysis predictive monitoring feature vectors, and infection test results. Assuming Mr. Li's hemodialysis event influencing factor is 1.1 (this factor integrates factors such as changes in catheter type, catheter assessment, and vascular access location), the hemodialysis monitoring parameters show that his blood flow occasionally falls below the normal threshold during recent dialysis, and venous pressure shows an abnormal negative correlation with blood flow; the hemodialysis predictive monitoring feature vector indicates a correlation of -0.8 between blood flow and venous pressure, and predicts future blood flow fluctuations within the range of 230-250 ml / min; the infection test results show that Mr. Li's most recent postoperative immune checkup results were all negative.

[0051] This information is input into the target hemodialysis event early warning model. The model assigns different weights to each piece of information based on its correlation with the infection risk. For example, the weight of the hemodialysis event impact factor is 0.4, the weight of the hemodialysis monitoring parameter information is 0.3, the weight of the hemodialysis prediction monitoring feature vector is 0.2, and the weight of the infection test result information is 0.1. After calculation, Mr. Li's infection risk score is found to be 0.6 (the specific calculation method is as follows).

[0052] S = F × v1 + P1 × v2 + P2 × v3 + P3 × v4. Where S is the infection risk score, F is the hemodialysis event influencing factor, v1 is the weight of the hemodialysis event influencing factor, P1 is the score of abnormal blood flow and related factors, v2 is the weight of hemodialysis monitoring parameter information, P2 is the prediction feature vector score, v3 is the weight of the hemodialysis prediction monitoring feature vector, P3 is the infection detection result score, and v4 is the weight of the infection examination result information.

[0053] The infection risk score of the target patient is processed to generate the patient's current infection risk level. The infection risk score is divided into different intervals to correspond to different infection risk levels. An infection risk score of 0-0.4 is defined as low risk, 0.4-0.7 as medium risk, and 0.7-1 as high risk. Mr. Li's infection risk score is 0.6, falling within the 0.4-0.7 interval, so Mr. Li's current infection risk level is determined to be medium risk. This means that Mr. Li has a certain possibility of infection and requires close monitoring by medical staff.

[0054] The system processes the current infection risk level of the target patient based on a preset warning threshold to generate a hemodialysis infection event warning. The hospital has pre-set warning thresholds; if the infection risk level reaches or exceeds a certain threshold, the system will trigger a warning. Assume the preset warning threshold is a medium-risk level (i.e., an infection risk score of 0.4). Since Mr. Li's infection risk level is medium-risk, the warning threshold has been reached, and the system generates a hemodialysis infection event warning. This warning will be displayed on the hospital's monitoring system interface, indicating that Mr. Li may have an infection risk, along with relevant risk factors, such as a high hemodialysis event influencing factor (suggesting that changes in catheter type may increase the infection risk), abnormal blood flow in hemodialysis monitoring parameters and abnormal correlations between parameters, and potential risks reflected in the hemodialysis predictive monitoring feature vector. After seeing the warning, medical staff will further consider Mr. Li's specific situation, such as reviewing his recent dialysis records and physical symptoms, to determine whether appropriate measures need to be taken, such as increasing the monitoring frequency during dialysis, conducting further examinations on Mr. Li, or adjusting the treatment plan to reduce the infection risk.

[0055] Furthermore, another aspect of this application involves processing hemodialysis event influencing factors, hemodialysis monitoring parameters of the target patient, hemodialysis prediction monitoring feature vectors, and infection test results of the target patient based on a target hemodialysis event early warning model to generate early warning information for hemodialysis infection events. This process further includes:

[0056] The system processes early warning information on hemodialysis infection events to generate information on the source and transmission routes of the infectious pathogens.

[0057] Based on the source information, transmission route information, and infection test results of the target patient, the system processes the data to generate target triage information and response priority information.

[0058] The target triage information and response priority information are processed to generate early warning measures for hemodialysis infection.

[0059] In one implementation, assuming Mr. Li's hemodialysis infection event warning information indicates a high risk of infection, the system first analyzes the possible sources and transmission routes of the infectious pathogen from various data. Mr. Li's dialysis history reveals that he recently had his temporary catheter (ECC) replaced, a procedure that may increase the risk of infection. Combined with catheter assessment results, if redness was observed around the puncture site, and although blood cultures during that period did not identify the pathogen, inflammatory markers were elevated, it is speculated that the infectious pathogen may have originated from contact infection during the catheter insertion procedure, with bacteria entering through the puncture site. Simultaneously, reviewing hospital environmental monitoring data reveals abnormalities in the dialysis equipment cleaning and disinfection records, and other patients during the same period also exhibited similar infection warnings, raising the possibility of incomplete disinfection of the dialysis equipment leading to pathogen transmission. Based on this information, the initial assessment is that the infectious pathogen is likely Staphylococcus aureus, originating from contamination during the catheter insertion procedure or incomplete disinfection of the dialysis equipment, with contact transmission as the transmission route.

[0060] Based on the information regarding the source and transmission route of the infectious pathogens, as well as Mr. Li's infection test results (such as inflammatory markers and bloodborne pathogen screening results), target triage information and response priority information are generated. Given Mr. Li's high infection risk and potential bloodborne infection risk, the first consideration is to triage him to the infection isolation area for further examination and treatment to avoid cross-infection. Regarding response priority, catheter-related infection risk is ranked first, as the catheter placement procedure is closely related to the current infection symptoms, and catheter-related infections, if not treated promptly, may lead to serious complications. Secondly, the issue of dialysis equipment disinfection is considered, as this involves the infection control safety of the entire dialysis center and requires timely investigation and rectification. Therefore, the target triage information is to transfer Mr. Li to the infection isolation area, and the response priority information is: first, address the risk of catheter-related infection; second, resolve the issue of dialysis equipment disinfection.

[0061] In response to the risk of infection associated with catheter placement, medical staff immediately reassessed Mr. Li's catheter, including replacing it or enhancing catheter care. Samples were collected from the catheter tip for bacterial culture and drug sensitivity testing to more accurately identify the pathogen and select appropriate antibiotics for treatment. Simultaneously, Mr. Li's vital signs and infection indicators, such as temperature and white blood cell count, were closely monitored. Regarding the disinfection of dialysis equipment, the logistics department was instructed to conduct a thorough disinfection of the dialysis equipment and related accessories used by Mr. Li, and to strengthen the supervision and inspection of daily disinfection management procedures for dialysis equipment to ensure that disinfection operations comply with regulations. Infection risk screening was conducted for other patients who used the same batch of dialysis equipment during the same period, and their health status was monitored. Furthermore, infection control training for medical staff was strengthened, emphasizing standardized procedures for catheter placement and equipment disinfection to prevent similar infection incidents from recurring.

[0062] The system acquires multifaceted data from target patients, such as dialysis physiological parameters and vascular access information. It generates hemodialysis event influencing factors by processing vascular access and outpatient hemodialysis catheterization records; processes dialysis physiological parameters to obtain hemodialysis monitoring parameters; and analyzes historical infection data to generate hemodialysis prediction and monitoring feature vectors. These data processing steps provide multi-dimensional evidence for subsequent infection risk assessment. A training sample set and a pre-set model are acquired, and the model is trained through statistical analysis, sampling, and data grouping to select an effective early warning model for target hemodialysis events. This model is then used to generate an infection risk score by integrating various data sources, further determining the infection risk level, and issuing early warning information for hemodialysis infection events based on pre-set thresholds.

[0063] In-depth analysis of early warning information identifies the source and transmission route of the infectious pathogen, such as determining whether the infection was caused by catheter placement procedures or dialysis equipment. Based on this information, target triage information and response priority information are generated to guide subsequent targeted measures. Finally, based on the above information, early warning measures for hemodialysis infection are formulated, such as adjusting treatment plans and strengthening equipment disinfection, to reduce the risk of infection.

[0064] Through comprehensive data processing and multi-step analysis and evaluation, effective monitoring and early warning of infection risks can be achieved for outpatient hemodialysis patients. This helps to identify potential infection risks in a timely manner, provides strong support for medical staff to take corresponding measures, and is of great significance for ensuring the safety of patients' dialysis and reducing the incidence of infection.

[0065] In one implementation, such as Figure 2 As shown, this application also provides a monitoring device for infection events during outpatient hemodialysis, comprising:

[0066] The acquisition module 201 is used to acquire the dialysis physiological parameters of the target patient undergoing maintenance hemodialysis within a preset time, the vascular access information of the target patient, the outpatient hemodialysis catheterization record of the target patient, the infection test results of the target patient, and the historical infection data of the target hemodialysis patient.

[0067] The processing module 202 is used to process the vascular access information and outpatient hemodialysis catheterization records of the target patient to generate hemodialysis event influencing factors; process the dialysis physiological parameter information of the target patient within a preset time to generate hemodialysis monitoring parameter information of the target patient; process the historical infection data of the target hemodialysis patient to generate hemodialysis prediction monitoring feature vector; and process the hemodialysis event influencing factors, the target patient's hemodialysis monitoring parameter information, the hemodialysis prediction monitoring feature vector, and the target patient's infection test results based on the target hemodialysis event early warning model to generate hemodialysis infection event early warning information.

[0068] The various embodiments in this application are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for the monitoring method, electronic device, electronic device, and readable storage medium for assessing infection events in outpatient hemodialysis are basically similar to the above-described embodiment for monitoring infection events in outpatient hemodialysis, and therefore are described simply. Relevant parts can be referred to in the description of the above-described embodiment for monitoring infection events in outpatient hemodialysis.

Claims

1. A method for monitoring infection events in outpatient hemodialysis, characterized in that, include: Acquire dialysis physiological parameters, vascular access information, outpatient hemodialysis catheterization records, infection test results, and historical infection data of target patients undergoing maintenance hemodialysis within a preset time period; The vascular access information and outpatient hemodialysis catheterization records of target patients are processed to generate hemodialysis event influencing factors. This includes processing the outpatient hemodialysis catheterization records to obtain information on changes in catheter type and catheter assessment results; processing the catheter type change information to generate catheter types for different time periods; processing the catheter assessment results to generate hematoma and bleeding information; and processing the catheter type, hematoma, and bleeding information for different time periods to generate influence degree information, where imaging quality information is used to characterize the influence weights corresponding to different catheter type changes and catheter assessment results. The vascular access information and imaging quality information of target patients are processed to generate hemodialysis event influencing factors. When calculating these factors, changes in catheter type, hematoma condition, bleeding condition, and vascular access location information must be considered, and each factor... Assign corresponding weights to the impact factors. Let the impact factor of the hemodialysis event be I. The impact weight of the change in catheter type is ω1, set to 0.8; the impact weight of hematoma status is ω2, set to 0.1; the impact weight of bleeding status is ω3, set to 0.1; the impact weight of vascular access location information is ω4, with an impact weight of ω4 of 0.3 when the vascular access location is the right internal jugular vein. For the change in catheter type, the target patient changes from long-term catheter to temporary catheter, which has a significant impact on the risk of infection. Let x1 represent the degree of impact of the change in catheter type, with x1 = 1 when the change occurs. Regarding the hematoma status, the target patient's catheter assessment result is no hematoma, so x2 represents the degree of impact of the hematoma status, with x2 = 0 when there is no hematoma. The target user has no bleeding, so the degree of impact of bleeding status is x3 = 0. The target patient's catheter location is the right internal jugular vein, so the degree of impact of the vascular access location is set to x4 = 1. The dialysis physiological parameter information of the target patient within a preset time is processed to generate the hemodialysis monitoring parameter information of the target patient; The system processes historical infection data of target hemodialysis patients to generate a hemodialysis prediction and monitoring feature vector. This includes obtaining the physiological parameters and target historical time period for each dialysis session; processing the physiological parameters and target historical time period for each dialysis session to generate the deviation values ​​between the parameters for each dialysis session and the historical data; obtaining the fluctuation range and abnormality of the parameter deviation values ​​based on the deviation values; processing the fluctuation range and abnormality of the parameter deviation values ​​based on correlation analysis and regression analysis to generate the correlation degree between each dialysis parameter and the predicted value range of each parameter; and processing the correlation degree between each dialysis parameter and the predicted value range of each parameter to generate the hemodialysis prediction and monitoring feature vector. Based on the target hemodialysis event early warning model, the influencing factors of hemodialysis events, the hemodialysis monitoring parameters of target patients, the hemodialysis prediction monitoring feature vector, and the infection test results of target patients are processed to generate early warning information for hemodialysis infection events. The system processes early warning information on hemodialysis infection events to generate information on the source and transmission routes of the infectious pathogens; and processes information on the source and transmission routes of the infectious pathogens and the infection test results of the target patients to generate target triage information and response priority information.

2. The method as described in claim 1, characterized in that, The acquisition of the target hemodialysis event early warning model includes: Acquire a training sample set and a preset hemodialysis event early warning model. The training sample set includes dialysis physiological parameters, vascular access information, outpatient hemodialysis catheterization records, infection test results, and historical infection data of other patients undergoing maintenance hemodialysis within a preset time period. The number of each data feature in the training sample set is counted, and a sampling ratio is generated based on these counts. The training sample set is sampled based on the sampling ratio to generate a preset number of sampled features. Based on any data feature and each sampling feature, the data is divided into a response group and a poor response group. Each group contains a preset number of data samples, and at least one data sample carries identification information. The pre-set hemodialysis event early warning model is processed based on the response group and the poor response group to generate the trained hemodialysis event early warning model and training results; If the data samples containing labeling information in the training results are factors characterizing the risk of hemodialysis infection, then the trained hemodialysis event early warning model will be used as the target hemodialysis event early warning model.

3. The method as described in claim 1, characterized in that, The dialysis physiological parameter information of the target patient within a preset time period is processed to generate hemodialysis monitoring parameter information for the target patient, including: The dialysis physiological parameter information of the target patient within a preset time period is processed to generate data for each dialysis session and the corresponding timestamp. Noise filtering and normalization are performed on each dialysis data to generate dialysis parameter sequence information for the target patient; Based on the preset threshold ranges of each parameter and the monitoring objectives, dynamic analysis and correlation analysis are performed on the dialysis parameter sequence information of the target patient to generate hemodialysis monitoring parameter information for the target patient.

4. The method as described in claim 1, characterized in that, Based on the target hemodialysis event early warning model, the influencing factors of hemodialysis events, hemodialysis monitoring parameters of target patients, hemodialysis predictive monitoring feature vectors, and infection test results of target patients are processed to generate early warning information for hemodialysis infection events, including: Based on the target hemodialysis event early warning model, the influencing factors of hemodialysis events, the hemodialysis monitoring parameters of target patients, the hemodialysis prediction monitoring feature vector, and the infection test results of target patients are processed to generate the infection risk score of target patients. The infection risk score of the target patient is processed to generate the current infection risk level of the target patient; Based on a preset warning threshold, the current infection risk level of the target patient is processed to generate early warning information for hemodialysis infection events.

5. The method as described in claim 1, characterized in that, Based on the target hemodialysis event early warning model, the influencing factors of hemodialysis events, hemodialysis monitoring parameters of target patients, hemodialysis predictive monitoring feature vectors, and infection test results of target patients are processed to generate early warning information for hemodialysis infection events. This process also includes: The target triage information and response priority information are processed to generate early warning measures for hemodialysis infection.

6. A monitoring device for infection events in outpatient hemodialysis, characterized in that, The apparatus for implementing the method of claim 1 includes: The acquisition module is used to acquire the dialysis physiological parameters, vascular access information, outpatient hemodialysis catheterization records, infection test results, and historical infection data of the target patients undergoing maintenance hemodialysis within a preset time period. The processing module is used to process the vascular access information and outpatient hemodialysis catheterization records of the target patient to generate hemodialysis event influencing factors; process the dialysis physiological parameter information of the target patient within a preset time to generate hemodialysis monitoring parameter information of the target patient; process the historical infection data of the target hemodialysis patient to generate hemodialysis prediction and monitoring feature vectors; and process the hemodialysis event influencing factors, the target patient's hemodialysis monitoring parameter information, the hemodialysis prediction and monitoring feature vectors, and the target patient's infection test results based on the target hemodialysis event early warning model to generate hemodialysis infection event early warning information.

7. An electronic device, characterized in that, include: First processor; and memory for storing executable instructions of the first processor; The first processor is configured to execute the method for monitoring infection events in outpatient hemodialysis as described in any one of claims 1 to 5 by executing the executable instructions.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the second processor, it implements the monitoring method for infection events in outpatient hemodialysis as described in any one of claims 1 to 5.

Images