A system for preventing arteriovenous fistula thrombosis in hemodialysis patients based on the knowledge-belief-practice health education model.

Abstract

This invention discloses a system for preventing arteriovenous fistula thrombosis in hemodialysis patients based on a knowledge-belief-practice (KBPP) health education model. The system includes a KBPP health education module, a combined traditional Chinese and Western medicine monitoring module, a thrombosis early warning module, a personalized intervention module, a data storage module, and an interactive module, forming a closed-loop prevention system. This invention creatively integrates the KBPP model with combined traditional Chinese and Western medicine techniques to achieve "proactive prevention – dynamic monitoring – precise early warning – personalized intervention" for arteriovenous fistula thrombosis in hemodialysis patients. It significantly improves patients' health awareness and adherence to proactive prevention behaviors, reduces the incidence of arteriovenous fistula thrombosis, and prolongs the lifespan of the fistula. Simultaneously, it takes into account the personalized advantages of TCM syndrome differentiation and nursing care and the accuracy of Western medicine monitoring, adapting to the individual differences of hemodialysis patients with TCM-related kidney disease. It provides reliable technical support for arteriovenous fistula protection in hemodialysis patients undergoing combined TCM and Western medicine treatment, and has significant clinical application value and promising prospects for promotion.

Description

Technical Field

[0001] This invention belongs to the technical fields of TCM kidney disease diagnosis and treatment, hemodialysis nursing and health education, and in particular relates to a system for preventing arteriovenous fistula thrombosis in hemodialysis patients based on the knowledge-belief-practice health education model. Background Technology

[0002] Arteriovenous fistulas (AVFs) are the "lifeline" for hemodialysis patients. Their patency directly determines the effectiveness of hemodialysis treatment and the patient's quality of life. Thrombosis is the leading cause of AVF failure, with an incidence rate as high as 20%-30%. This is especially true in patients with TCM-prescribed kidney diseases (such as edema, general weakness, and obstruction syndromes), where the incidence of AVF thrombosis is even higher and the prognosis is worse due to the pathogenesis of qi and blood deficiency, blood stasis, and damp-heat accumulation. Currently, the prevention of AVF thrombosis in hemodialysis patients mainly relies on Western clinical nursing and routine health education, combined with TCM syndrome differentiation and treatment. However, this approach has many technical limitations and cannot meet the personalized prevention needs of hemodialysis patients undergoing integrated TCM and Western medicine treatment. Specifically: Current health education mainly consists of lectures and brochure distribution, without combining the knowledge-belief-practice model to build a closed-loop system of "knowledge transfer - belief establishment - behavior formation". As a result, patients do not have a solid grasp of the pathogenesis of kidney disease in traditional Chinese medicine and the protection of arteriovenous fistula, have weak health beliefs, poor compliance with proactive preventive behaviors (such as arteriovenous fistula self-monitoring, dietary regulation, and emotional regulation), and poor thrombosis prevention effect.

[0003] Existing prevention programs mainly rely on Western medicine anticoagulation and arteriovenous fistula compression care. Traditional Chinese medicine interventions are limited to routine syndrome differentiation and medication. They do not deeply integrate traditional Chinese medicine syndrome differentiation (such as Qi deficiency and blood stasis syndrome, damp-heat accumulation syndrome) with patients' knowledge, attitudes and practices and arteriovenous fistula monitoring data. Therefore, they cannot develop personalized prevention programs based on individual patient differences (condition, constitution, cognitive level) and have poor adaptability.

[0004] Therefore, it is necessary to provide a new arteriovenous fistula thrombosis prevention system for hemodialysis patients based on a knowledge-belief-practice health education model to solve the above-mentioned technical problems. Summary of the Invention

[0005] The technical problem solved by this invention is to provide a creative integration of the knowledge-attitude-practice (KAP) model and traditional Chinese and Western medicine techniques to achieve "proactive prevention, dynamic monitoring, precise early warning, and personalized intervention" for arteriovenous fistula thrombosis in hemodialysis patients. This significantly improves patients' health awareness and adherence to proactive prevention behaviors, reduces the incidence of arteriovenous fistula thrombosis, and prolongs the lifespan of the fistula. At the same time, it takes into account the personalized advantages of TCM syndrome differentiation and nursing care and the accuracy of Western medicine monitoring, adapts to the individual differences of hemodialysis patients with TCM kidney disease, and provides reliable technical support for the protection of arteriovenous fistulas in hemodialysis patients undergoing integrated TCM and Western medicine treatment. This system, based on the KAP health education model, is a system for the prevention of arteriovenous fistula thrombosis in hemodialysis patients and has significant clinical application value and promotion prospects.

[0006] To address the aforementioned technical problems, this invention provides a blood fistula thrombosis prevention system for hemodialysis patients based on the knowledge-belief-practice health education model. This system is applied in the fields of traditional Chinese medicine nephrology and integrated traditional Chinese and Western medicine hemodialysis. It includes a knowledge-belief-practice health education module, an integrated traditional Chinese and Western medicine monitoring module, a thrombosis early warning module, and a personalized intervention module, which are sequentially connected by signals. It also includes a data storage module that is connected by signals to all of the above modules, and an interaction module that is connected by signals to the knowledge-belief-practice health education module, the personalized intervention module, and the thrombosis early warning module. The Knowledge, Belief, and Action (KBE) health education module is based on the KBE theory and is constructed according to three levels: cognition, belief, and behavior. It is used to accurately convey knowledge related to TCM kidney disease and arteriovenous fistula thrombosis prevention, guide patients' health beliefs, and standardize and monitor proactive preventive behaviors. It has a built-in TCM kidney disease syndrome differentiation knowledge base and arteriovenous fistula protection knowledge system, and can customize personalized health education content according to patients' TCM syndrome differentiation, cognitive level, and cultural background. The integrated traditional Chinese and Western medicine monitoring module includes a TCM syndrome differentiation monitoring unit, an arteriovenous fistula physiological index monitoring unit, and a behavioral data monitoring unit. The TCM syndrome differentiation and monitoring unit is used to collect patients' TCM symptoms, tongue and pulse information, and complete the automatic differentiation of common TCM kidney disease syndromes such as Qi deficiency and blood stasis syndrome, damp-heat accumulation syndrome, and spleen and kidney deficiency syndrome. The arteriovenous fistula physiological index monitoring unit is used to collect physiological indicators such as blood flow velocity, pulsation intensity, skin temperature, and circumference of the patient's arteriovenous fistula in real time; The behavioral data monitoring unit is used to collect data on patients' self-monitoring of arteriovenous fistulas, dietary management, emotional regulation, and medication adherence. The thrombosis early warning module constructs a thrombosis risk assessment model based on an improved BP neural network algorithm. It inputs TCM syndrome differentiation scores, arteriovenous fistula physiological index data and behavioral data collected by the integrated TCM and Western medicine monitoring module. It outputs the thrombosis risk level through multi-dimensional weighted calculation and issues an early warning signal when the risk level reaches medium or above. The personalized intervention module combines the patient's knowledge, attitudes and behaviors level, TCM syndrome differentiation and thrombosis risk level to output a personalized intervention plan that combines TCM and Western medicine, including TCM syndrome differentiation nursing plan, Western medicine nursing plan and knowledge, attitudes and behaviors reinforcement plan; The data storage module adopts a dual storage approach of local and cloud storage to achieve encrypted storage and traceability of patient basic information, TCM syndrome differentiation information, arteriovenous fistula monitoring data, health education records, intervention plans, and thrombosis risk assessment results throughout the entire process.

[0007] As a further aspect of the present invention, the three levels of the knowledge-belief-practice health education module are specifically as follows: (1) Cognitive level: It contains content such as the pathogenesis of kidney disease in traditional Chinese medicine, the thrombosis mechanism of arteriovenous fistula, self-monitoring methods of arteriovenous fistula, and prevention knowledge in traditional Chinese and Western medicine. It is pushed in various forms such as text, video, voice, and online Q&A, and supports adjusting the difficulty and presentation of the content according to the patient's cognitive level. (2) Belief level: Through personalized psychological intervention, sharing of successful cases, doctor-patient interaction and communication, etc., correct patients' misconceptions, strengthen patients' understanding of "the importance of fistula protection" and "the effectiveness of proactive prevention", and establish correct health beliefs; (3) Behavioral level: Set personalized behavioral goals and regulate patients’ preventive behaviors through mechanisms such as check-in, reminders, and points rewards. At the same time, collect the completion status of behaviors in real time through the behavioral data monitoring unit to form a behavioral assessment report.

[0008] As a further aspect of the present invention, the specific parameters and functions of the integrated traditional Chinese and Western medicine monitoring module are as follows: (1) Traditional Chinese Medicine syndrome differentiation monitoring unit: The traditional Chinese medicine syndrome differentiation scale is built in. The patient's symptoms, tongue and pulse information are collected through the interactive module. The fuzzy comprehensive evaluation method is used to complete the syndrome differentiation and output the syndrome differentiation score (0-10 points). The higher the score, the more typical the corresponding syndrome. (2) Arteriovenous fistula physiological index monitoring unit: A portable blood flow velocity monitor, temperature sensor and circumference measuring instrument are used to collect blood flow velocity, skin temperature and circumference data of the arteriovenous fistula in real time. The sampling frequency is once every 30 minutes. Abnormal data will trigger an alert in real time. (3) Behavioral data monitoring unit: Patient behavioral data are collected through the check-in function of the interactive module and the smart wearable device, including the completion rate of arteriovenous fistula self-monitoring, dietary compliance, emotional regulation, and medication compliance.

[0009] As a further aspect of the present invention, the improved BP neural network algorithm of the thrombosis early warning module includes the following steps and formulas: S1. Data Preprocessing: The TCM syndrome differentiation score (X1), arteriovenous fistula blood flow velocity (X2), skin temperature (X3), arteriovenous fistula circumference (X4), and behavioral completion rate (X5) collected by the integrated TCM and Western medicine monitoring module were normalized. The normalization formula is as follows: ; in, For the first Original data, For the normalized data, This is the minimum value of the indicator. This is the maximum value of the indicator. =1,2,3,4,5; S2. Construct an improved BP neural network: the input layer has 5 neurons (corresponding to 5 normalized indicators), the hidden layer has 2 layers (10 neurons in the first layer and 8 neurons in the second layer), and the output layer has 1 neuron (corresponding to the thrombosis risk score, 0-100 points); introduce a momentum factor. ( =0.9) Optimize the gradient descent algorithm to avoid model overfitting; S3. Weight Calculation and Forward Propagation: The formula for the output of hidden layer neurons is: ; The output formula for the output layer neuron is: ; in, The weights from the input layer to the hidden layer. The weights from the hidden layer to the output layer. , For bias terms, The Sigmoid activation function ( (x)=1 / (1+e^(-x))); S4. Error Calculation and Backpropagation: The mean squared error (MSE) is used to calculate the error between the predicted and actual values. The error formula is as follows: ; in, For the sample size, For actual thrombosis risk score, The model is assigned a score; the weights and biases are updated via backpropagation, using the following formula: ; ; in, Learning rate ( =0.001), This represents the current iteration number. Momentum factor; S5. Risk Level Determination: Based on the output thrombosis risk score Y, set thresholds: low risk (Y<30 points), intermediate risk (30≤Y<60 points), and high risk (Y≥60 points). Issue warning signals for intermediate risk and above.

[0010] As a further aspect of the present invention, the integrated traditional Chinese and Western medicine intervention program of the personalized intervention module is as follows: (1) TCM syndrome differentiation and nursing plan: Based on the TCM syndrome differentiation, the following treatments are given: Qi deficiency and blood stasis syndrome is given TCM decoction to replenish Qi and promote blood circulation and acupoint massage; damp-heat syndrome is given TCM decoction to clear heat and promote diuresis and dietary advice; spleen and kidney deficiency syndrome is given TCM decoction to strengthen the spleen and benefit the kidney and moxibustion. (2) Western medicine nursing plan: For low-risk patients, provide guidance on fistula protection; for medium-risk patients, provide guidance on low-dose anticoagulants and enhance fistula blood flow monitoring; for high-risk patients, immediately send an emergency reminder and recommend that medical staff perform thrombolysis and thrombectomy. (3) Knowledge, Attitude and Behavior Reinforcement Program: Based on the patient's behavior assessment report, targeted reminders and guidance are sent to patients who have not completed the behavior goals; and points are awarded to patients who have completed the goals well to strengthen proactive preventive behaviors.

[0011] As a further aspect of the present invention, the data storage module adopts a dual-mode of 1TB local SSD storage + cloud server storage. The local storage stores nearly 12 months of patient monitoring data, intervention plans, and health education records, while the cloud storage enables long-term data storage. The data uses the AES-256 encryption algorithm and supports three-level access control for administrators, medical staff, and patients. Administrators are responsible for system maintenance, medical staff can view the full-process data of the corresponding patients and adjust the intervention plan, and patients can view their own monitoring results, health education content, and intervention suggestions.

[0012] As a further embodiment of the present invention, the interactive module includes a 12-inch touch screen display, a voice interaction unit, a printing unit, and multiple data interfaces; the touch screen display enables patient information entry, TCM symptom collection, health education content viewing, behavior check-in, and monitoring data display; the voice interaction unit supports voice Q&A and command control; the printing unit can print the patient's diagnosis report, thrombosis risk assessment report, and personalized intervention plan; the data interfaces include USB, Ethernet, and Bluetooth interfaces, supporting connection with hospital HIS systems, LIS systems, and smart wearable devices to achieve data interconnection and interoperability.

[0013] As a further aspect of the present invention, the adaptive optimization function of the system comprises the following steps: S1. Collect patient behavioral feedback data, intervention effect data, and thrombosis occurrence data monthly; S2. Analyze the collected data. If the patient's behavior completion rate is less than 60%, optimize the presentation format and reminder mechanism of the knowledge, attitude and behavior health education content; if several prognostic thrombosis risk scores do not decrease significantly, adjust the parameters of the TCM syndrome differentiation and nursing plan and the Western medicine nursing plan. S3. Incremental training is adopted to input new data into the thrombosis risk assessment model, fine-tune the model weights and bias terms, and optimize the warning threshold; S4. Store the optimized health education content, intervention plans, and model parameters to the data storage module, and update all modules of the system synchronously to achieve continuous optimization.

[0014] As a further aspect of the present invention, the knowledge-belief-practice health education module also includes a built-in doctor-patient interaction unit, which supports patients to consult medical staff online. Medical staff can adjust the health education content and intervention suggestions in real time based on the patient's questions and monitoring data, thereby achieving personalized doctor-patient communication.

[0015] As a further aspect of the present invention, the integrated traditional Chinese and Western medicine monitoring module also includes an abnormal data filtering unit, employing 3 The criteria for filtering outlier data are as follows: ; in, This is the average value of the indicator. The standard deviation is used to determine the data that meets the formula. Data that meets the standard deviation is considered outlier, automatically removed, and re-collected to ensure the accuracy of the monitoring data.

[0016] The purpose of this invention is to provide a blood fistula thrombosis prevention system for hemodialysis patients based on a knowledge-belief-practice (KBPP) health education model. By deeply integrating the KBPP health education model with traditional Chinese medicine (TCM) syndrome differentiation and nursing care and Western medicine clinical monitoring techniques, a closed-loop prevention system is constructed. This system aims to improve patients' health awareness, cultivate proactive preventative behaviors, provide accurate early warning of blood fistula thrombosis, and enable personalized TCM and Western medicine interventions. It addresses the pain points of existing technologies, extends the lifespan of blood fistulas in hemodialysis patients, improves the treatment safety and quality of life for TCM-treated hemodialysis patients with kidney disease, and promotes the development of integrated TCM and Western medicine hemodialysis nursing techniques.

[0017] The core technical solution of this invention is to construct an integrated system for preventing arteriovenous fistula thrombosis in hemodialysis patients, which combines "knowledge, belief, and practice health education + integrated traditional Chinese and Western medicine monitoring + precise early warning + personalized intervention + closed-loop optimization". Through modular structural design and core algorithm innovation, combined with the characteristics of TCM syndrome differentiation and nursing care for kidney diseases, it overcomes the shortcomings of existing technologies. The specific technical solution is as follows: The system is equipped with six modules: health education, integrated traditional Chinese and Western medicine monitoring, thrombosis early warning, personalized intervention, data storage, and interaction. Each module works independently but is interconnected, enabling full automation of the entire process from patient information entry, health education, dynamic monitoring to early warning intervention and feedback optimization, reducing manual intervention and improving prevention efficiency.

[0018] The innovative Knowledge-Belief-Practice (KBPP) health education model breaks down the KBPP theory into three levels: cognition, belief, and behavior. It combines the characteristics of TCM kidney disease pathogenesis with knowledge of arteriovenous fistula thrombosis prevention to customize personalized health education content. Through multi-form delivery, psychological intervention, and behavioral reinforcement mechanisms, it constructs a closed loop of "knowledge transfer - belief establishment - behavior formation" to fundamentally improve patients' proactive prevention capabilities.

[0019] The integration of traditional Chinese and Western medicine monitoring technologies: The design includes three units: TCM syndrome differentiation monitoring, arteriovenous fistula physiological index monitoring, and behavioral data monitoring. It simultaneously collects patients' TCM syndrome differentiation information (symptoms, tongue and pulse), arteriovenous fistula physiological indexes (blood flow velocity, temperature, etc.), and behavioral data, achieving comprehensive collection of multi-dimensional data from both TCM and Western medicine, and providing data support for thrombosis early warning and personalized intervention.

[0020] Improved thrombosis early warning algorithm: Based on the BP neural network algorithm, a momentum factor optimization model is introduced, and a thrombosis risk assessment model is constructed by integrating multi-dimensional indicators from traditional Chinese and Western medicine. The risk level is calculated and output through a refined formula, so as to achieve accurate early warning of thrombosis and solve the problem of delayed warning.

[0021] Personalized TCM and Western medicine intervention plan: Based on the patient's knowledge, attitudes and behaviors, TCM syndrome differentiation and thrombosis risk level, a customized intervention plan combining TCM syndrome differentiation and nursing (Chinese medicine, acupoint massage, diet therapy) and Western medicine nursing (anticoagulation, arteriovenous fistula care) is developed to adapt to the individual differences of TCM kidney disease patients and improve the intervention effect.

[0022] Closed-loop feedback optimization mechanism: The system is equipped with an adaptive optimization function. By collecting patient behavioral feedback, intervention effects and thrombosis occurrence, the health education content, early warning model parameters and intervention plan are dynamically adjusted to form a closed-loop logic of "health education-monitoring-early warning-intervention-feedback-optimization" to ensure the system's generalization ability and long-term effectiveness.

[0023] Compared with related technologies, the arteriovenous fistula thrombosis prevention system for hemodialysis patients based on the knowledge-belief-practice health education model provided by this invention has the following beneficial effects: This invention overcomes the limitation of existing technologies that "disconnect health education from clinical intervention" by integrating the knowledge-belief-practice (KBP) health education model into the entire process of integrated traditional Chinese and Western medicine for thrombosis prevention. Through knowledge transfer at the cognitive level, psychological guidance at the belief level, and reinforcement and standardization at the behavioral level, combined with the characteristics of TCM nephrology diagnosis, it transforms patients from "passively receiving care" to "actively preventing" thrombosis. This significantly improves patients' understanding of TCM nephrology and arteriovenous fistula protection knowledge and their adherence to proactive preventive behaviors. At the same time, it integrates Western medicine monitoring and TCM intervention to achieve multi-dimensional thrombosis prevention, which can reduce the incidence of arteriovenous fistula thrombosis in hemodialysis patients from the current 20%-30% to below 5%, extend the lifespan of arteriovenous fistulas, and solve the core pain point of poor preventive effect of existing technologies.

[0024] This invention targets the pathogenesis characteristics of TCM patients with kidney disease, setting up a TCM syndrome differentiation and monitoring unit to automatically differentiate common syndromes such as Qi deficiency and blood stasis syndrome and damp-heat accumulation syndrome. It creatively combines the TCM syndrome differentiation results with the patient's knowledge, attitude, and behavior level and thrombosis risk level to customize personalized TCM and Western medicine intervention plans, avoiding the "one-size-fits-all" intervention model of existing technologies. The TCM syndrome differentiation and nursing plan (Chinese medicine, acupoint massage, and dietary therapy) and the Western medicine nursing plan (anticoagulation and arteriovenous fistula care) work synergistically, taking into account the advantages of TCM "syndrome differentiation and treatment" while ensuring the accuracy of Western medicine monitoring, significantly improving the intervention effect of thrombosis prevention and arteriovenous fistula protection, and is suitable for TCM patients with kidney disease and hemodialysis with different constitutions and conditions.

[0025] This invention employs an improved BP neural network algorithm, introducing a momentum factor to optimize the gradient descent process and avoid model overfitting. It integrates multi-dimensional indicators such as TCM syndrome differentiation scores, arteriovenous fistula physiological indicators, and behavioral data to construct a thrombosis risk assessment model. Through refined formula calculations and threshold settings, it achieves accurate thrombosis risk classification and early warning. Compared with existing conventional monitoring (palpation, auscultation), the warning time is 24-48 hours earlier, which can promptly detect early thrombosis signals and push intervention suggestions, effectively preventing thrombosis formation or progression, reducing the risk of arteriovenous fistula failure, and solving the problem of delayed warning in existing technologies.

[0026] This invention constructs a closed-loop logic of "health education-monitoring-early warning-intervention-feedback-optimization". Through the system's adaptive optimization function, it regularly collects patients' behavioral feedback, intervention effects, and thrombosis occurrence, dynamically adjusting health education content, early warning model parameters, and intervention plans. This allows the system to adapt to changes in the condition and behavioral habits of TCM kidney disease patients, continuously improving the accuracy and effectiveness of prevention. At the same time, it adopts a dual storage method of local + cloud to achieve full-process data traceability, facilitating medical staff to review and optimize, providing data support for clinical research and protocol improvement, ensuring long-term stable operation of the system, and adapting to multiple clinical applications. Attached Figure Description

[0027] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0028] Figure 1 The overall system structure block diagram provided by the present invention; Figure 2 A schematic diagram of the system provided by this invention; Figure 3 This is the workflow of the Knowledge, Integrity, and Practice Health Education module in the system provided by this invention; Figure 4 The structure of the integrated traditional Chinese and Western medicine monitoring module in the system provided by this invention; Figure 5 A diagram of the improved BP neural network model in the system provided by this invention; Figure 6 This is a schematic diagram of the system workflow and closed-loop feedback logic provided by the present invention. Figure 7 A logic diagram is generated for the personalized intervention scheme in the system provided by this invention. Detailed Implementation

[0029] Please refer to the following: Figure 1 and Figure 7 ,in, Figure 1 The overall system structure block diagram provided by the present invention; Figure 2 A schematic diagram of the system provided by this invention; Figure 3 This is the workflow of the Knowledge, Integrity, and Practice Health Education module in the system provided by this invention; Figure 4 The structure of the integrated traditional Chinese and Western medicine monitoring module in the system provided by this invention; Figure 5 A diagram of the improved BP neural network model in the system provided by this invention; Figure 6 This is a schematic diagram of the system workflow and closed-loop feedback logic provided by the present invention. Figure 7 This invention provides a logic diagram for generating personalized intervention schemes in the system. The system, based on the Knowledge-Instruction-Practice (KIIP) health education model, is applied in the fields of traditional Chinese medicine nephrology and integrated traditional Chinese and Western medicine hemodialysis. It includes a KIIP health education module, an integrated traditional Chinese and Western medicine monitoring module, a thrombosis early warning module, and a personalized intervention module, all sequentially connected by signals. It also includes a data storage module connected to all of the above modules, and an interaction module connected to the KIIP health education module, the personalized intervention module, and the thrombosis early warning module. The Knowledge-Belief-Practice (KBP) health education module is based on the KBP theory and is constructed at three levels: cognition, belief, and behavior. It is used to accurately convey knowledge related to TCM kidney disease and arteriovenous fistula thrombosis prevention, guide patients' health beliefs, and standardize and monitor proactive preventive behaviors. It has a built-in TCM kidney disease syndrome differentiation knowledge base and arteriovenous fistula protection knowledge system, and can customize personalized health education content according to the patient's TCM syndrome differentiation, cognitive level, and cultural background. The integrated traditional Chinese and Western medicine monitoring module includes a traditional Chinese medicine syndrome differentiation monitoring unit, an arteriovenous fistula physiological index monitoring unit, and a behavioral data monitoring unit. The TCM syndrome differentiation and monitoring unit is used to collect patients' TCM symptoms, tongue and pulse information, and complete the automatic differentiation of common TCM kidney disease syndromes such as Qi deficiency and blood stasis syndrome, damp-heat accumulation syndrome, and spleen and kidney deficiency syndrome. The arteriovenous fistula physiological index monitoring unit is used to collect physiological indicators such as blood flow velocity, pulsation intensity, skin temperature, and circumference of the patient's arteriovenous fistula in real time; The behavioral data monitoring unit is used to collect data on patients' self-monitoring of arteriovenous fistulas, dietary management, emotional regulation, and medication adherence. The thrombosis early warning module constructs a thrombosis risk assessment model based on an improved BP neural network algorithm. It inputs TCM syndrome differentiation scores, arteriovenous fistula physiological index data and behavioral data collected by the integrated TCM and Western medicine monitoring module. It outputs the thrombosis risk level through multi-dimensional weighted calculation, which is specifically divided into: low risk, medium risk and high risk. It issues an early warning signal when the risk level reaches medium risk or above. The personalized intervention module combines the patient's knowledge, attitudes, and behaviors (KAP) level, TCM syndrome differentiation, and thrombosis risk level to output a personalized intervention plan that integrates TCM and Western medicine. This includes a TCM syndrome differentiation and care plan, which specifically includes TCM conditioning, acupoint massage, and dietary therapy suggestions, and a Western medicine care plan, which specifically includes guidance on anticoagulant medication, arteriovenous fistula compression care, and a KAP reinforcement plan. The data storage module adopts a dual storage method of local + cloud to realize the encrypted storage and traceability of the entire process of patient basic information, TCM syndrome differentiation information, arteriovenous fistula monitoring data, health education records, intervention plans, and thrombosis risk assessment results; The system has an adaptive optimization function. By collecting patients' behavioral feedback data, intervention effect data, and thrombosis occurrence data, it dynamically adjusts the knowledge, belief, and practice health education content, thrombosis risk assessment model parameters, and personalized intervention plans, forming a closed-loop logic of "health education-monitoring-early warning-intervention-feedback-optimization".

[0030] The three levels of the Knowledge, Belief, and Practice Health Education module are as follows: (1) Cognitive level: The TCM pathogenesis of kidney disease is built in, including qi and blood deficiency and blood stasis, the mechanism of arteriovenous fistula thrombosis, arteriovenous fistula self-monitoring methods, TCM and Western medicine prevention knowledge, etc. It is pushed in various forms such as pictures, videos, voice, online Q&A, and supports adjusting the difficulty and presentation of the content according to the patient's cognitive level, specifically primary school, middle school, university and above. (2) Belief level: Through personalized psychological intervention, sharing of successful cases, doctor-patient interaction and communication, etc., correct patients' misconceptions, strengthen patients' understanding of "the importance of fistula protection" and "the effectiveness of proactive prevention", and establish correct health beliefs; (3) Behavioral level: Set personalized behavioral goals, such as self-monitoring of the arteriovenous fistula twice a day and acupoint massage three times a week. Through mechanisms such as check-in, reminders, and point rewards, standardize the patient's preventive behavior. At the same time, collect the behavior completion status in real time through the behavior data monitoring unit to form a behavior assessment report.

[0031] The specific parameters and functions of the integrated traditional Chinese and Western medicine monitoring module are as follows: (1) Traditional Chinese Medicine syndrome differentiation monitoring unit: The traditional Chinese medicine syndrome differentiation scale is built in. The patient's symptoms, such as fatigue, edema, numbness of limbs, and tongue and pulse information, such as tongue body, tongue coating and pulse, are collected through the interactive module. The syndrome differentiation is completed by fuzzy comprehensive evaluation method and the syndrome differentiation score (0-10 points) is output. The higher the score, the more typical the corresponding syndrome. (2) Arteriovenous fistula physiological index monitoring unit: A portable blood flow velocity monitor is used, with a measurement range of 0.1-10cm / s and an accuracy of ±0.05cm / s; a temperature sensor is used, with a measurement range of 32-40℃ and an accuracy of ±0.1℃; and a circumference measuring instrument is used, with a measurement range of 5-20cm and an accuracy of ±0.1cm. The system collects arteriovenous fistula blood flow velocity, skin temperature, and circumference data in real time. The sampling frequency is once every 30 minutes, and abnormal data triggers an alert in real time. (3) Behavioral data monitoring unit: Through the check-in function of the interactive module and smart wearable devices, such as the linkage bracelet, the patient's behavioral data is collected, including the completion rate of self-monitoring of arteriovenous fistula, dietary compliance, whether the taboos of traditional Chinese medicine dietary therapy are followed, emotional regulation, daily emotional score, medication compliance, and the use of traditional Chinese medicine and Western medicine.

[0032] The improved BP neural network algorithm of the thrombosis early warning module includes the following steps and formulas: S1. Data Preprocessing: The TCM syndrome differentiation score (X1), arteriovenous fistula blood flow velocity (X2), skin temperature (X3), arteriovenous fistula circumference (X4), and behavioral completion rate (X5) collected by the integrated TCM and Western medicine monitoring module were normalized. The normalization formula is as follows: ; in, For the first Original data, For the normalized data, This is the minimum value of the indicator. This is the maximum value of the indicator. =1,2,3,4,5; S2. Construct an improved BP neural network: the input layer has 5 neurons (corresponding to 5 normalized indicators), the hidden layer has 2 layers (10 neurons in the first layer and 8 neurons in the second layer), and the output layer has 1 neuron (corresponding to the thrombosis risk score, 0-100 points); introduce a momentum factor. ( =0.9) Optimize the gradient descent algorithm to avoid model overfitting; S3. Weight Calculation and Forward Propagation: The formula for the output of hidden layer neurons is: ; The output formula for the output layer neuron is: ; in, The weights from the input layer to the hidden layer. The weights from the hidden layer to the output layer. , For bias terms, The Sigmoid activation function ( (x)=1 / (1+e^(-x))); S4. Error Calculation and Backpropagation: The mean squared error (MSE) is used to calculate the error between the predicted and actual values. The error formula is as follows: ; in, For the sample size, For actual thrombosis risk score, The model is assigned a score; the weights and biases are updated via backpropagation, using the following formula: ; ; in, Learning rate ( =0.001), This represents the current iteration number. Momentum factor; S5. Risk Level Determination: Based on the output thrombosis risk score Y, set thresholds: low risk (Y<30 points), intermediate risk (30≤Y<60 points), and high risk (Y≥60 points). Issue warning signals for intermediate risk and above.

[0033] The integrated traditional Chinese and Western medicine intervention plan of the personalized intervention module is as follows: (1) TCM syndrome differentiation and nursing plan: Based on the TCM syndrome differentiation, the following TCM nursing plan is formulated: Qi deficiency and blood stasis syndrome is given TCM decoction to replenish qi and activate blood, such as Astragalus membranaceus, Angelica sinensis and Salvia miltiorrhiza, and acupoint massage, specifically Neiguan, Hegu and Zusanli; Damp-heat syndrome is given TCM decoction to clear heat and promote diuresis, specifically Phellodendron chinense, Atractylodes lancea and Plantago asiatica, and dietary advice: avoid spicy and greasy foods and eat more foods that clear heat and promote diuresis; Spleen and kidney deficiency syndrome is given TCM decoction to strengthen the spleen and benefit the kidney, specifically Codonopsis pilosula, Atractylodes macrocephala and Lycium barbarum, and moxibustion, specifically Shenshu and Pishu. (2) Western medicine nursing plan: For low-risk patients, provide guidance on fistula protection, avoid compression, and engage in moderate activity; for medium-risk patients, provide guidance on low-dose anticoagulants such as aspirin and enhance fistula blood flow monitoring; for high-risk patients, immediately send an emergency reminder and recommend that medical staff perform thrombolysis and thrombectomy. (3) Knowledge, belief and behavior reinforcement program: Based on the patient behavior assessment report, targeted reminders and guidance are sent to patients who have not completed the behavior goals; patients who have completed the goals well are given points rewards, which can be redeemed for health consultations, Chinese herbal teas, etc., to strengthen proactive preventive behaviors.

[0034] The data storage module adopts a dual mode of 1TB local SSD storage + cloud server storage. The local storage stores nearly 12 months of patient monitoring data, intervention plans and health education records, while the cloud storage enables long-term data storage. The data uses the AES-256 encryption algorithm and supports three-level permission management for administrators, medical staff and patients. Administrators are responsible for system maintenance, medical staff can view the full process data of the corresponding patients and adjust the intervention plan, and patients can view their own monitoring results, health education content and intervention suggestions.

[0035] The interactive module includes a 12-inch touchscreen display, a voice interaction unit, a printing unit, and multiple data interfaces. The touchscreen display enables patient information entry, TCM symptom collection, viewing of health education content, behavior check-in, and monitoring data display. The voice interaction unit supports voice Q&A and command control, such as "querying arteriovenous fistula monitoring data." The printing unit can print the patient's diagnosis report, thrombosis risk assessment report, and personalized intervention plan. The data interfaces include USB, Ethernet, and Bluetooth interfaces, supporting connection with hospital HIS systems, LIS systems, and smart wearable devices to achieve data interconnection.

[0036] The adaptive optimization function of the system comprises the following steps: S1. Collect patient behavioral feedback data monthly, including behavioral completion rate, health education satisfaction, intervention effect data, changes in thrombosis risk score, symptom improvement, and thrombosis occurrence. S2. Analyze the collected data. If the patient's behavior completion rate is less than 60%, optimize the presentation format and reminder mechanism of the knowledge, attitude and behavior health education content; if several prognostic thrombosis risk scores do not decrease significantly, adjust the parameters of the TCM syndrome differentiation and nursing plan and the Western medicine nursing plan. S3. Incremental training is adopted to input new data into the thrombosis risk assessment model, fine-tune the model weights and bias terms, and optimize the warning threshold; S4. Store the optimized health education content, intervention plans, and model parameters to the data storage module, and update all modules of the system synchronously to achieve continuous optimization.

[0037] The Knowledge, Belief, and Practice (KBPA) health education module also includes a built-in doctor-patient interaction unit, supporting patients to consult medical staff (traditional Chinese medicine and Western medicine) online. Medical staff can adjust the health education content and intervention suggestions in real time based on the patient's questions and monitoring data, thus achieving personalized doctor-patient communication.

[0038] The integrated traditional Chinese and Western medicine monitoring module also includes an abnormal data filtering unit, employing 3D... The criteria filter out abnormal data collected (such as abnormal blood flow velocity or temperature data deviation caused by operational errors). The filtering formula is as follows: ; in, This is the average value of the indicator. The standard deviation is used to determine the data that meets the formula. Data that meets the standard deviation is considered outlier, automatically removed, and re-collected to ensure the accuracy of the monitoring data.

[0039] This embodiment provides a fistula thrombosis prevention system for hemodialysis patients based on the knowledge-belief-practice health education model. It is specifically designed for the prevention of fistula thrombosis in patients with traditional Chinese medicine nephropathy and patients undergoing hemodialysis with integrated traditional Chinese and Western medicine. The system hardware is modularly designed, and the software is developed based on the Windows 10 system and Python programming language. It integrates the knowledge base of traditional Chinese medicine syndrome differentiation with Western medicine monitoring technology and can be directly applied to the nephrology department and hemodialysis center of a hospital. The specific parameters, structure, and functions of the ZhiXinXing health education module are as follows: It includes a built-in TCM (Traditional Chinese Medicine) kidney disease syndrome differentiation knowledge base (containing symptoms, pathogenesis, and nursing points of 8 common syndrome types such as Qi deficiency and blood stasis syndrome, damp-heat accumulation syndrome, and spleen and kidney deficiency syndrome), and an arteriovenous fistula thrombosis prevention knowledge system (TCM arteriovenous fistula protection knowledge, Western medicine arteriovenous fistula nursing points, and thrombosis emergency treatment methods). It supports adjusting the content difficulty according to the patient's cognitive level (primary school, middle school, university and above), and the presentation formats include illustrated manuals, short videos (1-3 minutes), voice explanations, and online Q&A. The belief guidance unit includes 200+ successful cases (TCM kidney disease hemodialysis patients' arteriovenous fistula protection cases), supports doctor-patient interaction (online consultation with TCM and Western medicine medical staff); the behavior reinforcement unit sets up a check-in function and a points reward mechanism (points can be redeemed for Chinese herbal teas and health consultation services), and pushes daily behavior reminders (arteriovenous fistula self-monitoring, acupoint massage).

[0040] The specific parameters, structure, and functions of the integrated traditional Chinese and Western medicine monitoring module are as follows: (1) Traditional Chinese Medicine syndrome differentiation monitoring unit: The traditional Chinese medicine syndrome differentiation scale is built in. The patient's symptoms (fatigue, edema, numbness of limbs, dry mouth, etc.) and tongue and pulse information (red / pale tongue, white / yellow greasy tongue coating, wiry / thin pulse) are collected through the touch screen of the interactive module. The syndrome differentiation is completed by using the fuzzy comprehensive evaluation method and the syndrome differentiation score (0-10 points) is output. The score ≥6 points is judged as a typical syndrome. (2) Arteriovenous fistula physiological index monitoring unit: portable blood flow velocity monitor (measurement range 0.1-10cm / s, accuracy ±0.05cm / s), contact temperature sensor (32-40℃, ±0.1℃), electronic circumference measuring instrument (5-20cm, ±0.1cm), sampling frequency 30 minutes / time, abnormal data (such as blood flow velocity <1cm / s, temperature fluctuation >1℃) trigger audible and visual reminders in real time; (3) Behavioral data monitoring unit: Through the touch screen check-in function and Bluetooth linkage smart bracelet, the system collects the patient's self-monitoring completion rate of arteriovenous fistula, dietary compliance (whether or not to follow the taboos of traditional Chinese medicine diet therapy), emotional regulation (daily emotional score of 0-10 points), and medication compliance (check-in of taking Chinese medicine and Western medicine). The data is then uploaded to the data storage module.

[0041] The specific parameters, structure, and functions of the thrombosis early warning module are as follows: A thrombosis risk assessment model was constructed based on an improved BP neural network algorithm. The input layer has 5 neurons (TCM syndrome differentiation score X1, arteriovenous fistula blood flow velocity X2, skin temperature X3, arteriovenous fistula circumference X4, and behavior completion rate X5), two hidden layers (10 neurons in the first layer and 8 neurons in the second layer), and one output layer (thrombosis risk score 0-100). Model parameters: momentum factor α=0.9, learning rate η=0.001, 500 iterations, and batch size 64. The training dataset consisted of clinical data from 1000 TCM-treated hemodialysis patients (500 normal patients and 500 thrombosis patients). After training, the model achieved an accuracy ≥98% and a recall ≥97%. Risk level thresholds were: low risk (Y<30), intermediate risk (30≤Y<60), and high risk (Y≥60). Warning signals for intermediate and higher risks were pushed via an interactive module and a mobile app.

[0042] The specific parameters, structure, and functions of the personalized intervention module are as follows: (1) TCM syndrome differentiation and nursing plan: Qi deficiency and blood stasis syndrome (syndrome differentiation score ≥6 points), Huangqi Danshen Decoction (Astragalus membranaceus 30g, Angelica sinensis 15g, Salvia miltiorrhiza 15g, etc.) is given, 1 dose per day, and acupoint massage (Neiguan, Hegu, Zusanli) for 10 minutes each time, twice a day; Damp-heat syndrome, Huangbai Cangzhu Decoction (Phellodendron chinense 12g, Atractylodes lancea 10g, Plantago asiatica 15g, etc.) is given, and dietary advice: eat more winter melon and coix seed, and avoid spicy and greasy food; Spleen and kidney deficiency syndrome, Dangshen Baizhu Decoction (Dangshen 15g, Atractylodes macrocephala 12g, Lycium barbarum 10g, etc.) is given, and moxibustion is applied to Shenshu and Pishu, twice a week; (2) Western medicine nursing plan: Low risk, send guidance on fistula protection (avoid weight-bearing and compression on the limb on the side of the fistula, and moderate daily activity); Medium risk, give guidance on anticoagulation with aspirin (100mg / day) and strengthen fistula monitoring (once every 15 minutes); High risk, immediately send an emergency reminder and recommend medical staff to perform thrombolytic therapy (intravenous drip of urokinase) or thrombectomy. (3) Knowledge, Belief and Action Reinforcement Program: If the completion rate of behavior is less than 60%, two reminders will be added daily, and targeted guidance (such as a video on the self-monitoring steps of arteriovenous fistula) will be pushed; if the completion rate of behavior is greater than or equal to 80%, a reward of 10 points will be given, and accumulated points can be redeemed for one TCM health consultation.

[0043] The specific parameters, structure, and functions of the data storage module are as follows: Local storage: 1TB SSD, storing nearly 12 months of patient basic information, TCM syndrome differentiation data, arteriovenous fistula monitoring data, health education records, intervention plans, and thrombosis risk assessment results; Cloud storage: Alibaba Cloud server, using AES-256 encryption to achieve long-term data storage; Access control: three-level access control for administrators (system maintenance, parameter settings), medical staff (viewing patient data, adjusting intervention plans), and patients (viewing their own data, health education content), supporting access from multiple terminals (computers, mobile phones).

[0044] The specific parameters, structure, and functions of the interaction module are as follows: A 12-inch touchscreen display (1920×1080 resolution) supports patient information entry, TCM symptom collection, health education content viewing, and behavior check-in; a voice interaction unit (supporting Mandarin and dialects) enables voice Q&A and command control; a small thermal printer prints diagnostic reports, thrombosis risk assessment reports, and personalized intervention plans; data interfaces: USB 3.0, Ethernet, and Bluetooth 5.0, supporting connection with hospital HIS systems, LIS systems, and smart bracelets to achieve data interconnection.

[0045] The working principle of the arteriovenous fistula thrombosis prevention system for hemodialysis patients based on the knowledge-belief-practice health education model provided by this invention is as follows: This invention relates to a blood fistula thrombosis prevention system for hemodialysis patients based on the knowledge-belief-practice (KBPP) health education model. Addressing the needs of hemodialysis patients with traditional Chinese medicine-related kidney disease and those undergoing integrated traditional Chinese and Western medicine treatment for KBPP thrombosis prevention, the system achieves full-process automation and precision "from health education to system optimization." Its core working principle includes three-level guidance based on knowledge, belief, and practice; multi-dimensional monitoring using both traditional Chinese and Western medicine; improved BP neural network early warning; personalized syndrome differentiation intervention; and closed-loop feedback optimization. Specific detailed steps (see attached diagram) are detailed below. Figure 1-6 )as follows: S1: Patient Information Entry and Initial Assessment (Appendix) Figure 1 ) 1.1 Medical staff can input basic patient information (name, age, gender, hemodialysis duration, TCM diagnosis of kidney disease, etc.) through the touch screen of the interactive module. Patients can complete the collection of TCM symptoms and tongue and pulse information through the interactive module. The TCM syndrome differentiation monitoring unit uses the fuzzy comprehensive evaluation method to complete the initial syndrome differentiation, output the syndrome differentiation type and syndrome differentiation score, and store them synchronously in the data storage module. 1.2 The Knowledge, Attitude and Practice (KAP) health education module pushes a cognitive level test questionnaire (10 questions) through the interactive module. Based on the patient's answer, the cognitive level (primary school, middle school, university and above) is determined. At the same time, the patient's health beliefs (such as the awareness of "the importance of fistula protection") and current preventive behavior habits are collected to form an initial KAP assessment report. 1.3 The fistula physiological index monitoring unit starts initial monitoring, collects basic data such as blood flow velocity, skin temperature, and circumference of the patient's fistula, which serve as the benchmark values ​​for subsequent risk assessment and are simultaneously uploaded to the thrombosis early warning module and data storage module.

[0046] S2: Three-Tier Health Education Based on Knowledge, Belief, and Practice (with appendix) Figure 2 ) 2.1 Cognitive Hierarchical Intervention: The Knowledge, Belief, and Practice (KBAP) health education module customizes personalized knowledge content based on the patient's cognitive level and TCM syndrome differentiation. For example, patients with primary school level knowledge are given basic content in the form of pictures and short videos (steps for self-monitoring of arteriovenous fistula, simple knowledge of TCM dietary therapy), while patients with university level or above are given in-depth content (pathogenesis of TCM kidney disease, mechanism of thrombosis). One piece of knowledge content is pushed out every day, and after the push, a short quiz (3 questions) is used to check the patient's understanding. Those who do not meet the standard are pushed out the relevant content again. 2.2 Belief Hierarchy Guidance: Successful cases of arteriovenous fistula protection in TCM hemodialysis patients with kidney disease are pushed through the interactive module. A doctor-patient interactive exchange is organized once a week (online). Medical staff answer patients' questions, correct misconceptions (such as "arteriovenous fistula does not require daily monitoring"), and strengthen patients' belief that "proactive prevention can reduce the risk of thrombosis". For patients with weak belief, one-on-one psychological intervention is arranged by a dedicated person once a month. 2.3 Behavioral Hierarchy Reinforcement: Based on the patient's initial behavioral habits, personalized behavioral goals are set (such as self-monitoring of the arteriovenous fistula twice a day, acupoint massage three times a week, and a daily mood score of ≥7 points). Daily behavioral reminders are pushed through the interactive module. After completing the behavior, the patient checks in through the touch screen. The behavior data monitoring unit collects the check-in status in real time and generates a weekly behavior assessment report. Patients who complete the task well are given points as a reward, and those who do not complete the task are pushed targeted guidance videos.

[0047] S3: Multi-dimensional dynamic monitoring of integrated traditional Chinese and Western medicine (with appendix) Figure 3 ) 3.1 Dynamic monitoring of TCM syndrome differentiation: The patient's TCM symptoms and tongue and pulse information are collected once a week. The TCM syndrome differentiation monitoring unit updates the syndrome differentiation type and syndrome differentiation score. If the syndrome differentiation type changes (such as from Qi deficiency and blood stasis syndrome to spleen and kidney deficiency syndrome), it is synchronized to the personalized intervention module and data storage module in real time. 3.2 Real-time monitoring of arteriovenous fistula physiological indicators: The arteriovenous fistula physiological indicator monitoring unit collects data on blood flow velocity, skin temperature, and circumference of the patient's arteriovenous fistula at a frequency of 30 minutes / time. The abnormal data filtering unit uses the 3σ criterion to filter abnormal data (such as blood flow velocity suddenly dropping below 0.5cm / s). The filtered data is uploaded to the thrombosis early warning module in real time. Abnormal data triggers audible and visual alerts to remind patients and medical staff to pay attention. 3.3 Continuous collection of behavioral data: The behavioral data monitoring unit collects patients' self-monitoring completion rate of arteriovenous fistula, dietary compliance, emotional regulation, and medication compliance in real time through the check-in function and smart bracelet. The behavioral data is summarized daily to generate a daily behavioral report, which is then uploaded to the data storage module and the personalized intervention module.

[0048] S4: Improved BP Neural Network Thrombosis Early Warning (with appendix) Figure 4 ) 4.1 Data Preprocessing: The thrombosis early warning module receives the TCM syndrome differentiation score (X1), arteriovenous fistula blood flow velocity (X2), skin temperature (X3), arteriovenous fistula circumference (X4), and behavior completion rate (X5) collected by the integrated TCM and Western medicine monitoring module, and uses a normalization formula. Normalize the data to eliminate the influence of dimensions; 4.2 Forward Propagation of the Model: The normalized 5 data points are input into the input layer of the improved BP neural network. The model is then processed through the hidden layers (10 neurons in the first layer and 8 neurons in the second layer), using the activation function... The results of each neuron are output, and finally the thrombosis risk score Y (0-100 points) is output through the output layer. 4.3 Error Calculation and Backpropagation: The mean square error formula is used. The error between the predicted and actual values ​​is calculated, and the model weights and bias terms are updated using the backpropagation algorithm, combined with the momentum factor α=0.9 and the learning rate η=0.001, to continuously optimize the model's prediction accuracy. 4.4 Risk Level Determination and Early Warning: Based on the thrombosis risk score Y, the risk level (low, medium, high) is determined. Low-risk patients only have their risk score displayed in the interactive module and no early warning is issued. Medium-risk patients receive audio-visual reminders through the interactive module and push notifications via the mobile app, advising them to strengthen monitoring and follow the intervention plan. High-risk patients receive an immediate emergency alert and their monitoring data is simultaneously sent to the healthcare staff's terminal, advising them to intervene promptly.

[0049] S5: Output and Implementation of Personalized Integrated Traditional Chinese and Western Medicine Intervention Programs (Appendix) Figure 6 ) 5.1 Intervention Plan Generation: The personalized intervention module receives the knowledge, attitude, and behavior assessment report, TCM syndrome differentiation results, and thrombosis risk level. Combined with TCM and Western medicine monitoring data, it generates personalized intervention plans, including TCM syndrome differentiation nursing plans, Western medicine nursing plans, and knowledge, attitude, and behavior reinforcement plans. 5.2 Intervention Plan Delivery and Implementation: Intervention plans are delivered to patients through interactive modules and mobile apps, detailing the methods of taking traditional Chinese medicine, acupoint massage steps, dietary restrictions, and anticoagulant medication guidelines. Medical staff can view the implementation status of the patient's intervention plan through the system and adjust the plan parameters (such as the dosage of traditional Chinese medicine and the frequency of acupoint massage) based on patient feedback. 5.3 Monitoring of intervention effect: Data on the intervention effect of patients (improvement of TCM symptoms, changes in physiological indicators of arteriovenous fistula, and changes in thrombosis risk score) are collected weekly, and an intervention effect evaluation report is generated and uploaded to the data storage module simultaneously to provide a basis for subsequent system optimization.

[0050] S6: Data Storage and System Adaptive Optimization 6.1 Full-process data storage: The patient's basic information, TCM syndrome differentiation data, arteriovenous fistula monitoring data, health education records, behavioral data, thrombosis risk assessment results, intervention plan, and intervention effect data are stored on local SSD (for the past 12 months) and cloud server (long-term), respectively. The data is encrypted with AES-256 and managed according to three levels of access control to ensure data security and traceability. 6.2 Feedback Data Collection: Monthly collection of patient behavioral feedback data (behavior completion rate, health education satisfaction), intervention effect data (changes in thrombosis risk score, improvement in TCM symptoms), and thrombosis occurrence data, followed by data analysis. 6.3 System Optimization and Adjustment: If the patient behavior completion rate is less than 60%, optimize the presentation format of the knowledge, attitude, and practice health education content (e.g., increase the proportion of short videos) and the reminder mechanism (increase the number of reminders); if several prognostic thrombosis risk scores do not show a significant decrease, adjust the TCM syndrome differentiation and care plan (e.g., adjust the dosage of Chinese medicine and change acupoints) and the Western medicine nursing plan (e.g., adjust the dosage of anticoagulants); adopt incremental training method, input new data into the thrombosis risk assessment model, fine-tune the model weights and bias terms, and optimize the early warning threshold; 6.4 Formation of Closed-Loop Logic: Through the process of "patient entry - health education - dynamic monitoring - early warning - intervention - feedback - optimization", a complete closed-loop logic is formed, which enables the system's prevention accuracy and personalization to be continuously improved with clinical application, and to meet the long-term needs of TCM kidney disease hemodialysis patients for fistula thrombosis prevention.

[0051] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions, and variations can be made to these embodiments, or they can be used directly or indirectly, without departing from the principles and spirit of the invention. In other related technical fields, the scope of the invention is defined by the appended claims and their equivalents, and they are similarly included within the scope of patent protection of the invention.

Claims

1. A hemodialysis patient internal fistula thrombosis prevention system based on a health education mode of knowledge, belief and behavior, applied to the fields of traditional Chinese medicine nephropathy and integrated traditional Chinese and western medicine hemodialysis, characterized in that, It includes the ZhiXinXing Health Education Module, the Integrated Traditional Chinese and Western Medicine Monitoring Module, the Thrombosis Early Warning Module, and the Personalized Intervention Module, which are connected by signals in sequence. It also includes a data storage module that is connected by signals to all of the above modules, and an interactive module that is connected by signals to the ZhiXinXing Health Education Module, the Personalized Intervention Module, and the Thrombosis Early Warning Module. The Knowledge, Belief, and Action (KBE) health education module is based on the KBE theory and is constructed according to three levels: cognition, belief, and behavior. It is used to accurately convey knowledge related to TCM kidney disease and arteriovenous fistula thrombosis prevention, guide patients' health beliefs, and standardize and monitor proactive preventive behaviors. It has a built-in TCM kidney disease syndrome differentiation knowledge base and arteriovenous fistula protection knowledge system, and can customize personalized health education content according to patients' TCM syndrome differentiation, cognitive level, and cultural background. The integrated traditional Chinese and Western medicine monitoring module includes a TCM syndrome differentiation monitoring unit, an arteriovenous fistula physiological index monitoring unit, and a behavioral data monitoring unit. The TCM syndrome differentiation and monitoring unit is used to collect patients' TCM symptoms, tongue and pulse information, and complete the automatic differentiation of common TCM kidney disease syndromes such as Qi deficiency and blood stasis syndrome, damp-heat accumulation syndrome, and spleen and kidney deficiency syndrome. The arteriovenous fistula physiological index monitoring unit is used to collect physiological indicators such as blood flow velocity, pulsation intensity, skin temperature, and circumference of the patient's arteriovenous fistula in real time; The behavioral data monitoring unit is used to collect data on patients' self-monitoring of arteriovenous fistulas, dietary management, emotional regulation, and medication adherence. The thrombosis early warning module constructs a thrombosis risk assessment model based on an improved BP neural network algorithm. It inputs TCM syndrome differentiation scores, arteriovenous fistula physiological index data and behavioral data collected by the integrated TCM and Western medicine monitoring module. It outputs the thrombosis risk level through multi-dimensional weighted calculation and issues an early warning signal when the risk level reaches medium or above. The personalized intervention module combines the patient's knowledge, attitudes and behaviors level, TCM syndrome differentiation and thrombosis risk level to output a personalized intervention plan that combines TCM and Western medicine, including TCM syndrome differentiation nursing plan, Western medicine nursing plan and knowledge, attitudes and behaviors reinforcement plan; The data storage module adopts a dual storage approach of local and cloud storage to achieve encrypted storage and traceability of patient basic information, TCM syndrome differentiation information, arteriovenous fistula monitoring data, health education records, intervention plans, and thrombosis risk assessment results throughout the entire process.

2. The thrombosis prevention system for internal fistula of hemodialysis patients based on health education mode of KAB, according to claim 1, characterized in that: The three levels of the Knowledge, Belief, and Practice Health Education module are as follows: (1) Cognitive level: It contains content such as the pathogenesis of kidney disease in traditional Chinese medicine, the thrombosis mechanism of arteriovenous fistula, self-monitoring methods of arteriovenous fistula, and prevention knowledge in traditional Chinese and Western medicine. It is pushed in various forms such as text, video, voice, and online Q&A, and supports adjusting the difficulty and presentation of the content according to the patient's cognitive level. (2) Belief level: Through personalized psychological intervention, sharing of successful cases, doctor-patient interaction and communication, etc., correct patients' misconceptions, strengthen patients' understanding of "the importance of fistula protection" and "the effectiveness of proactive prevention", and establish correct health beliefs; (3) Behavioral level: Set personalized behavioral goals and regulate patients’ preventive behaviors through mechanisms such as check-in, reminders, and points rewards. At the same time, collect the completion status of behaviors in real time through the behavioral data monitoring unit to form a behavioral assessment report.

3. The thrombosis prevention system for internal fistula of hemodialysis patients based on health education mode of KAB, according to claim 1, characterized in that: The specific parameters and functions of the integrated traditional Chinese and Western medicine monitoring module are as follows: (1) Traditional Chinese Medicine syndrome differentiation monitoring unit: The traditional Chinese medicine syndrome differentiation scale is built in. The patient's symptoms, tongue and pulse information are collected through the interactive module. The fuzzy comprehensive evaluation method is used to complete the syndrome differentiation and output the syndrome differentiation score (0-10 points). The higher the score, the more typical the corresponding syndrome. (2) Arteriovenous fistula physiological index monitoring unit: A portable blood flow velocity monitor, temperature sensor and circumference measuring instrument are used to collect blood flow velocity, skin temperature and circumference data of the arteriovenous fistula in real time. The sampling frequency is once every 30 minutes. Abnormal data will trigger an alert in real time. (3) Behavioral data monitoring unit: Patient behavioral data are collected through the check-in function of the interactive module and the smart wearable device, including the completion rate of arteriovenous fistula self-monitoring, dietary compliance, emotional regulation, and medication compliance.

4. The thrombosis prevention system for internal fistula of hemodialysis patients based on health education mode of KAB, according to claim 1, characterized in that: The improved BP neural network algorithm of the thrombosis early warning module includes the following steps and formulas: S1. Data Preprocessing: The TCM syndrome differentiation score (X1), arteriovenous fistula blood flow velocity (X2), skin temperature (X3), arteriovenous fistula circumference (X4), and behavioral completion rate (X5) collected by the integrated TCM and Western medicine monitoring module were normalized. The normalization formula is as follows: ； wherein, is the first is the original data, is the normalized data, is the minimum value of the index, is the maximum value of the index, = 1, 2, 3, 4, 5; S2. Construct an improved BP neural network: the input layer has 5 neurons (corresponding to 5 normalized indicators), the hidden layer has 2 layers (10 neurons in the first layer and 8 neurons in the second layer), and the output layer has 1 neuron (corresponding to the thrombosis risk score, 0-100 points); introduce a momentum factor. ( =0.9) Optimize the gradient descent algorithm to avoid model overfitting; S3. Weight Calculation and Forward Propagation: The formula for the output of hidden layer neurons is: ； The output formula for the output layer neuron is: ； in, The weights from the input layer to the hidden layer. The weights from the hidden layer to the output layer. , For bias terms, The Sigmoid activation function ( (x)=1 / (1+e^(-x))); S4. Error Calculation and Backpropagation: The mean squared error (MSE) is used to calculate the error between the predicted and actual values. The error formula is as follows: ； in, For the sample size, For actual thrombosis risk score, The model is assigned a score; the weights and biases are updated via backpropagation, using the following formula: ； ； in, Learning rate ( =0.001), This represents the current iteration number. Momentum factor; S5. Risk Level Determination: Based on the output thrombosis risk score Y, set thresholds: low risk (Y<30 points), intermediate risk (30≤Y<60 points), and high risk (Y≥60 points). Issue warning signals for intermediate risk and above.

5. The arteriovenous fistula thrombosis prevention system for hemodialysis patients based on the knowledge-attitude-practice health education model according to claim 1, characterized in that: The integrated traditional Chinese and Western medicine intervention plan of the personalized intervention module is as follows: (1) TCM syndrome differentiation and nursing plan: Based on the TCM syndrome differentiation, the following treatments are given: Qi deficiency and blood stasis syndrome is given TCM decoction to replenish Qi and promote blood circulation and acupoint massage; damp-heat syndrome is given TCM decoction to clear heat and promote diuresis and dietary advice; spleen and kidney deficiency syndrome is given TCM decoction to strengthen the spleen and benefit the kidney and moxibustion. (2) Western medicine nursing plan: For low-risk patients, provide guidance on fistula protection; for medium-risk patients, provide guidance on low-dose anticoagulants and enhance fistula blood flow monitoring; for high-risk patients, immediately send an emergency reminder and recommend that medical staff perform thrombolysis and thrombectomy. (3) Knowledge, Attitude and Behavior Reinforcement Program: Based on the patient's behavior assessment report, targeted reminders and guidance are sent to patients who have not completed the behavior goals; Patients who perform well will be rewarded with points to encourage proactive preventative behaviors.

6. The arteriovenous fistula thrombosis prevention system for hemodialysis patients based on the knowledge-attitude-practice health education model according to claim 1, characterized in that: The data storage module adopts a dual mode of 1TB local SSD storage + cloud server storage. The local storage stores nearly 12 months of patient monitoring data, intervention plans and health education records, while the cloud storage enables long-term data storage. The data uses the AES-256 encryption algorithm and supports three-level permission management for administrators, medical staff and patients. Administrators are responsible for system maintenance, medical staff can view the full process data of the corresponding patients and adjust the intervention plan, and patients can view their own monitoring results, health education content and intervention suggestions.

7. The arteriovenous fistula thrombosis prevention system for hemodialysis patients based on the knowledge-attitude-practice health education model according to claim 1, characterized in that: The interactive module includes a 12-inch touch screen, a voice interaction unit, a printing unit, and multiple data interfaces; the touch screen enables patient information entry, TCM symptom collection, health education content viewing, behavior check-in, and monitoring data display; the voice interaction unit supports voice Q&A and command control; The printing unit can print patients' diagnostic reports, thrombosis risk assessment reports, and personalized intervention plans; the data interfaces include USB, Ethernet, and Bluetooth interfaces, supporting connection with hospital HIS systems, LIS systems, and smart wearable devices to achieve data interconnection.

8. The arteriovenous fistula thrombosis prevention system for hemodialysis patients based on the knowledge-attitude-practice health education model according to claim 1, characterized in that: The adaptive optimization function of the system comprises the following steps: S1. Collect patient behavioral feedback data, intervention effect data, and thrombosis occurrence data monthly; S2. Analyze the collected data. If the patient's behavior completion rate is less than 60%, optimize the presentation format and reminder mechanism of the knowledge, attitude and behavior health education content; if several prognostic thrombosis risk scores do not decrease significantly, adjust the parameters of the TCM syndrome differentiation and nursing plan and the Western medicine nursing plan. S3. Incremental training is adopted to input new data into the thrombosis risk assessment model, fine-tune the model weights and bias terms, and optimize the warning threshold; S4. Store the optimized health education content, intervention plans, and model parameters to the data storage module, and update all modules of the system synchronously to achieve continuous optimization.

9. The arteriovenous fistula thrombosis prevention system for hemodialysis patients based on the knowledge-attitude-practice health education model according to claim 1, characterized in that: The Knowledge, Belief, and Practice (KBPA) health education module also includes a built-in doctor-patient interaction unit, supporting patients to consult with medical staff online. Medical staff can adjust the health education content and intervention suggestions in real time based on the patient's questions and monitoring data, achieving personalized doctor-patient communication.

10. The arteriovenous fistula thrombosis prevention system for hemodialysis patients based on the knowledge-attitude-practice health education model according to claim 1, characterized in that: The integrated traditional Chinese and Western medicine monitoring module also includes an abnormal data filtering unit, employing 3D... The criteria for filtering outlier data are as follows: ； in, This is the average value of the indicator. The standard deviation is used to determine the data that meets the formula. Data that meets the standard deviation is considered outlier, automatically removed, and re-collected to ensure the accuracy of the monitoring data.

Images