State real-time monitoring method, system and equipment for blood dialysis

Abstract

The invention discloses a state real-time monitoring method, system and equipment for blood dialysis, and aims to collect physiological parameters such as urea concentration, creatinine concentration, potassium concentration, weight and body fluid quantity of a dialysate in real time through multi-sensor equipment, and perform data analysis and prediction in combination with a machine learning model. Continuous monitoring and abnormal early warning of the state of a dialyzer in the dialysis process are achieved. The system comprises an online blood analyzer, an electronic weighing scale, a biological impedance analyzer and a central intelligent device, and efficient data transmission and processing are realized through a data transmission interface. If it is detected that the physiological parameters exceed the safety threshold value, the system automatically triggers early warning and informs medical staff, so that the safety and effectiveness of treatment are improved.

Description

Technical Field

[0001] The present invention relates to the field of medical monitoring, and in particular to a method, system and equipment for real-time monitoring of the state of dialyzed blood. Background Art

[0002] Dialysis treatment has become an important means for hemodialysis patients with renal failure to maintain their lives. Dialysis treatment is divided into two categories: hemodialysis and peritoneal dialysis. Hemodialysis has become the most common form of dialysis due to its significant therapeutic effect. The progress of modern hemodialysis technology is not only reflected in the continuous optimization of equipment, but also in the improvement of the means of monitoring the state of dialysis blood. The state monitoring of dialysis blood is not only related to the effect of treatment, but also directly affects the quality of life of the hemodialysis patients. Real-time and accurate acquisition and analysis of the physiological parameters of the hemodialysis patients can prevent and reduce the occurrence of dialysis complications to the greatest extent and improve the safety and effectiveness of treatment. In the prior art, the state monitoring of dialysis blood mainly relies on traditional periodic physical examinations and manual monitoring. However, the traditional manual monitoring method relies on the experience and reaction speed of medical staff, and it is difficult to achieve 24-hour uninterrupted real-time monitoring. Traditional monitoring systems usually only monitor single or limited physiological parameters and lack the ability to comprehensively analyze multiple physiological indicators. Due to the complex interrelationships between different physiological parameters, relying only on a single data may lead to deviations in monitoring and judgment results, reducing the reliability of monitoring. Manual monitoring often has a delayed response to emergencies and is unable to promptly identify and handle drastic changes in the status of hemodialysis patients. Although existing monitoring equipment can collect and display physiological parameters, it often lacks intelligent analysis and early warning functions. Medical staff need to manually analyze data and cannot promptly identify potential risks, which greatly reduces the efficiency and accuracy of monitoring.

[0003] For the real-time monitoring system of the dialyzer status in the emergency room based on the Internet of Things, the invention patent publication number is CN202410026974.0, the vital signs data are obtained and analyzed through the Internet of Things technology to realize real-time monitoring of emergency dialyzers. The system improves the real-time and accuracy of dialyzer monitoring through means such as facial images and blood oxygen analysis. However, the technology is still limited to the use scenarios of emergency wards, and lacks long-term monitoring and data prediction functions for chronic disease dialyzers, such as dialyzed blood. There are also inventors who use image crowd density recognition detection methods to implement monitoring. Although it can detect and monitor the group status in a specific environment, its technology does not involve individual physiological status monitoring of dialyzed blood, and lacks personalized health management solutions for chronic disease dialyzers. For an intelligent management system for a hemodialysis center with the invention patent publication number CN112509676A, it covers the information management, safety management and on-site management units of the hemodialysis center, aiming to improve the overall management efficiency of the hemodialysis center and the safety of dialyzers. Although the system is helpful for the safe management of dialyzed blood, its monitoring function is mainly aimed at the operation status of the entire hemodialysis center, rather than the real-time monitoring and prediction of the physiological status of a single dialyzer. Some inventors have also disclosed a peritoneal dialysis system and monitoring method that can be remotely monitored, which can remotely monitor the dialysis status of multiple dialyzers and update treatment prescriptions. However, the system is mainly aimed at peritoneal dialysis blood, the monitored physiological indicators are relatively limited, and there is a lack of in-depth analysis of the data and real-time early warning functions, which makes it difficult to meet the needs of hemodialysis blood for real-time monitoring of multiple parameters.

[0004] Although the above-mentioned related inventions have proposed solutions for monitoring the status of dialyzers in different application scenarios, they have the following major defects:

[0005] Limitations of monitoring objects: Most existing technologies are aimed at monitoring emergency hemodialysis patients or in specific environments, and lack long-term and continuous status monitoring of chronic diseases such as hemodialysis blood.

[0006] Imperfect early warning mechanism: Although most current systems can monitor physiological parameters, they lack an effective early warning mechanism and are unable to promptly notify medical staff for intervention when the dialyzer's condition becomes abnormal.

[0007] Insufficient data processing capabilities: Existing technologies can often only process a limited number of physiological parameters and lack the ability to comprehensively analyze and predict multi-dimensional data, making it difficult to accurately determine the health trends and potential risks of dialysis patients. Summary of the invention

[0008] In view of the above-mentioned deficiencies, the present invention proposes a method, system and device for real-time monitoring of the status of dialyzed blood. The system integrates a variety of sensor devices and can continuously collect multiple physiological parameters of dialyzers, including urea concentration, creatinine concentration, potassium concentration, body weight and body fluid volume. Combined with machine learning models, the system can perform real-time analysis and prediction of these physiological parameters, and issue early warning notifications in time when any abnormalities are detected. Through full-process monitoring, the system can significantly improve the safety of dialyzed blood, reduce the occurrence of complications, and improve the quality of life of dialyzers. The method specifically includes the following steps:

[0009] S1: a sensor device integrated on the dialysis device, which continuously collects physiological parameters of the dialyzed blood, including urea concentration, creatinine concentration, potassium concentration, body weight and body fluid volume;

[0010] S2: Collect the parameters every 5 seconds and transmit them to the central intelligent device in real time;

[0011] S3: On the central intelligent device, use the machine learning model to predict the values ​​of various physiological parameters of blood in 5 seconds;

[0012] S4: Make a judgment by comparing the differences between the parameter values ​​with the preset safety threshold;

[0013] S5: The central large screen receives the physiological parameter values ​​stored and predicted by the central intelligent device, and displays real-time data and warning information;

[0014] S6: If a parameter in step S4 is judged to exceed the threshold, the warning system is automatically triggered, and a warning notification is sent to the responsible doctor and nursing staff via SMS or application push;

[0015] S7: During the entire dialysis process of the dialyzer, steps S1 to S6 are continuously repeated to maintain real-time monitoring and early warning of the dialyzed blood status.

[0016] As a technical preferred solution of the present invention, the sensor device in step S1 is specifically:

[0017] T1: Online blood analyzer, used to collect urea concentration, creatinine concentration and potassium concentration; the measurement unit of urea concentration and creatinine concentration is mg / dL, mg / dL represents milligram per deciliter; the measurement unit of potassium concentration is mmol / L, mmol / L represents millimole per liter;

[0018] T2: electronic weighing scale, used to collect body weight; the unit of measurement for body weight is kg, which stands for kilogram;

[0019] T3: Bioimpedance analyzer, used to collect body fluid volume; the unit of measurement for body fluid volume is L, where L stands for liter.

[0020] As a technical preferred solution of the present invention, the central intelligent device in step S2 includes the following steps to predict the values ​​of various physiological parameters of the dialyzer:

[0021] S2-1: predicting five machine learning models on the central intelligent device, wherein the five machine learning models are specifically a urea concentration prediction model, a creatinine concentration prediction model, a potassium concentration prediction model, a body weight prediction model, and a body fluid volume prediction model;

[0022] S2-2: Each machine learning model collects continuous and real-time data collected by corresponding sensor devices. The data collected by the urea concentration prediction model, creatinine concentration prediction model, and potassium concentration prediction model come from online blood analyzers, the data collected by the weight prediction model comes from electronic weight scales, and the data collected by the body fluid volume prediction model comes from bioimpedance analyzers;

[0023] S2-3: Store the collected real-time data in the buffer of the central intelligent device, and continuously record at least the previous 10 minutes of data to form a data time series;

[0024] S2-4: Use each machine learning model to analyze the corresponding data time series stored in the buffer, the model input is the data collected every 5 seconds in the past 10 minutes, and the output is the predicted value of each physiological parameter of the blood at the next moment;

[0025] Among them, the urea concentration prediction model is responsible for the prediction of urea concentration, the creatinine concentration prediction model is responsible for the prediction of creatinine concentration, the potassium concentration prediction model is responsible for the prediction of potassium concentration, the weight prediction model is responsible for the prediction of weight, and the body fluid volume prediction model is responsible for the prediction of body fluid volume. Each prediction model is equipped with the same physiological indicator prediction algorithm.

[0026] As a technical preferred solution of the present invention, the physiological index prediction algorithm specifically includes the following steps:

[0027] P1: Model initialization:

[0028] Initialize state estimate Among them, y 0 is the initial parameter estimate, v 0 is an estimate of the rate of change of a parameter, wherein the state specifically includes urea concentration, creatinine concentration, potassium concentration, body weight and body fluid volume;

[0029] Initialize the state transfer matrix and the measurement matrix H k =

[10] , where Δt is the sampling time interval, specifically 5 seconds;

[0030] P2: Process and measurement noise settings:

[0031] Set process noise covariance and the measurement noise covariance Among them, q yy Reflects the predicted change rate of urea concentration, specifically set to 0.1; q yv and q vy Reflects the relationship between urea concentration and change rate, specifically set to 0; q vv Reflects the prediction uncertainty of the urea concentration change rate itself, which is specifically set to 0.1; z k is the actual measured value of the k measurement point; y k is the reference value; N is the total duration;

[0032] P3: For each measurement point k, predict the parameter estimate at the next time point. The specific calculation formula is:

[0033]

[0034] in, is the forecast error covariance;

[0035] P4: Repeat P2 and P3 to update the measurement, including:

[0036] Calculate the Kalman gain:

[0037] Update the state estimate:

[0038] Update the error covariance:

[0039] Among them, z k is the actual measurement at time k;

[0040] P5: The model prediction value at the last moment Output as the predicted value of physiological indicators at the next time point.

[0041] The present invention proposes a real-time monitoring system for the status of dialysis blood, which specifically includes:

[0042] Central intelligent device: equipped with a processor and a memory, wherein the processor is used to execute a software program including data reception, analysis and prediction, and the memory is used to store real-time and historical data;

[0043] Central large screen: electrically connected to the central intelligent device, used to display physiological parameters of blood and early warning information;

[0044] Online blood analyzer: comprising a plurality of sensors, including sensors for measuring urea, creatinine and potassium concentrations, and the online blood analyzer is provided with a data transmission interface, through which the collected data is sent to the central intelligent device;

[0045] Electronic weight scale: used to measure the dialyzer's weight, including a load sensor and a display, wherein the load sensor is used to detect the dialyzer's weight and send the weight data to the central intelligent device through a built-in transmission interface;

[0046] Bioimpedance analyzer: comprising electrodes and a signal processing unit, wherein the electrodes are attached to the body surface of the dialyzer to measure the amount of body fluid, and the signal processing unit is used to process the data collected from the electrodes and send the data to the central intelligent device through a connecting line;

[0047] Medical bed: a bed body used to support the body of the dialyzer, the bed body being provided with an interface for connecting with the electronic weight scale and the bioimpedance analyzer.

[0048] As a technical preferred solution of the present invention, the online blood analyzer further includes the following parts:

[0049] Automatic sampling mechanism: used to automatically extract blood samples from the blood circulation of the dialysis machine at regular intervals;

[0050] Analysis chamber: The analysis chamber is equipped with a chemical reactor for reacting with the blood sample to generate a detectable signal;

[0051] Detector: used to detect the signal generated by the chemical reaction and convert the detection result into digital data, which is then sent through the data transmission interface.

[0052] The present invention provides a monitoring device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] Strong real-time performance and high accuracy: The present invention integrates multiple sensor devices, such as online blood analyzers, electronic weight scales, and bioimpedance analyzers, to collect key physiological parameters of dialyzed blood in real time and continuously. Compared with traditional manual monitoring methods, the present invention achieves high-frequency data collection (such as once every 5 seconds), and combines machine learning models to perform real-time analysis and prediction of data, greatly improving the accuracy and real-time performance of monitoring, and can capture subtle changes in the status of dialyzers in a timely manner.

[0055] Comprehensive analysis and prediction of multiple parameters: The present invention not only monitors a single physiological parameter, but also comprehensively monitors and predicts multiple important physiological indicators such as urea concentration, creatinine concentration, potassium concentration, body weight and body fluid volume. By using multiple machine learning models to perform continuous data time series analysis on these parameters, the present invention can accurately predict multiple physiological indicators at the same time and compare the predicted values ​​with the safety threshold, thereby comprehensively evaluating the health status of the dialyzer and avoiding the risks caused by misjudgment of a single data.

[0056] Intelligent early warning system: Another important advantage of the present invention is its intelligent early warning mechanism. When a physiological parameter exceeds the preset safety threshold, the system can automatically trigger an early warning and immediately notify the responsible doctor and nursing staff through SMS or application push. This early warning mechanism greatly improves the timeliness of medical response and helps medical staff take appropriate intervention measures before the dialysis patient is in danger, thereby effectively preventing the occurrence of complications.

[0057] Systematic and integrated design: The present invention realizes the systematic management of dialysis blood monitoring by integrating various sensor devices, data processing center and display terminal. The central intelligent device is responsible for data collection, analysis and prediction, and the central large screen is used to display real-time data and early warning information, ensuring efficient data processing and timely communication of information. The overall system design is reasonable and comprehensive, so that every link in the dialysis process can be closely connected, improving the safety and reliability of the overall treatment.

[0058] Strong adaptability, suitable for a variety of scenarios: The present invention is not only suitable for dialysis centers in hospitals, but also for other scenarios such as home dialysis. The system is flexible in design and can be adjusted and configured according to specific needs, such as increasing or decreasing monitoring parameters, changing data collection frequency, etc., to adapt to different usage environments and dialyzer needs, so that it can be more widely used in different dialysis treatment scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A method flow chart of the method for real-time monitoring of the state of dialyzed blood provided by the present invention;

[0060] Figure 2 A module structure diagram of a real-time monitoring system for the state of dialyzed blood according to an embodiment of the present invention;

[0061] Figure 3 A central intelligent device prediction flow chart of an embodiment of the present invention is provided;

[0062] Figure 4 A simulation diagram of the change of urea concentration in dialyzed blood according to an embodiment of the present invention is provided;

[0063] Figure 5A simulation diagram of the change of creatinine concentration in dialyzed blood according to an embodiment of the present invention is provided;

[0064] Figure 6 A simulation diagram of the change of potassium concentration in dialysis blood provided by the present invention;

[0065] Figure 7 Provides a simulation diagram of the change in dialysis blood weight of an embodiment of the present invention;

[0066] Figure 8 A simulation diagram of the change in the volume of dialysis blood fluid is provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0067] The present invention is further described below in conjunction with the accompanying drawings and examples. However, the present invention can be implemented in many different ways and should not be construed as being limited to the embodiments shown; on the contrary, these embodiments provide those skilled in the art with implementation methods that meet applicable legal requirements.

[0068] Example 1: Figure 1 As shown, in this embodiment, the workflow of the real-time monitoring system for the state of dialyzed blood based on the present invention is integrated with a variety of sensor devices, central intelligent devices and early warning systems to achieve continuous monitoring, data analysis and automatic early warning of the physiological state of dialyzed blood. The specific implementation process of this embodiment will be described in detail below:

[0069] The various modules of the real-time monitoring system for dialysis blood status work together through reasonable layout and electrical connection to ensure efficient operation of data collection, transmission, analysis and early warning during dialysis. The main modules of the system include online blood analyzer, electronic weight scale, bioimpedance analyzer, central intelligent equipment and central large screen. These devices are installed around the dialysis bed and connected to the central intelligent equipment by wire or wireless means to form a complete monitoring network:

[0070] T1: Online blood analyzer: This device is usually installed next to the dialysis machine and directly connected to the blood circulation system of the dialysis machine. The online blood analyzer is connected to the dialysis machine through a sampling line, automatically extracts samples from the blood circulation and analyzes them. The analyzed data is sent to the central intelligent device through the data transmission interface.

[0071] T2: Electronic weight scale: The electronic weight scale is installed under the dialysis bed as part of the bed. When the dialyzer lies on the bed, the weight scale can measure the dialyzer's weight in real time. The weight data is transmitted to the central intelligent device through the sensor interface to ensure the continuity and accuracy of the data.

[0072] T3: Bioimpedance Analyzer: The electrodes of this device are usually fixed on both sides of the dialysis bed. The electrodes are connected to the body surface of the dialyzer through wires to measure the amount of body fluid. The signals collected by the electrodes are processed by the signal processing unit and transmitted to the central intelligent device through the connecting line for further analysis.

[0073] T4: Central smart device: The central smart device is usually placed on the central console of the dialysis room and maintains a stable connection with other devices. The device is connected to the online blood analyzer, electronic weight scale and bioimpedance analyzer by wire or wireless means, and receives and processes the data transmitted from these devices.

[0074] T5: Central large screen: The central large screen is installed in a prominent position in the dialysis room, usually in the operation area of ​​medical staff, for real-time monitoring. The large screen is connected to the central intelligent device via a data cable or wireless connection, and displays the physiological parameters of the dialyzer's blood and system analysis results in real time.

[0075] T6: Medical bed: The medical bed is equipped with a bed body for supporting the body of the dialyzer, and the bed body is provided with an interface for connecting with the electronic weight scale and the bioimpedance analyzer.

[0076] like Figure 2 As shown in the figure, the various devices of the dialysis blood status real-time monitoring system work together through precise connection and data transmission methods to ensure the real-time collection, stable transmission and accurate analysis of blood physiological parameters. The system includes main components such as online blood analyzer, electronic weight scale, bioimpedance analyzer, central intelligent equipment and central large screen. The following details the information collection and transmission methods of each device and its connection with other parts of the system:

[0077] Online blood analyzers are mainly used to monitor the blood components of dialyzed blood in real time, including key physiological parameters such as urea concentration, creatinine concentration and potassium concentration. Its working process is as follows:

[0078] Sample collection: The online blood analyzer is connected to the blood circulation system of the dialysis machine through a sterile sampling line. During the dialysis process, the analyzer automatically draws blood samples from the blood circulation at regular intervals and generates measurable signals through chemical reactions.

[0079] Data transmission: After the blood analysis results are converted into digital signals, they are transmitted to the central intelligent device through the data transmission interface. Specifically, the data transmission interface is connected by a shielded cable, and the interface type is a standard Ethernet interface or serial port (such as RS-232 or RS-485). To ensure that there is no electromagnetic interference during the transmission process, the data transmission is carried out using the TCP / IP protocol to ensure efficient and reliable data transmission.

[0080] Connection method: The online blood analyzer is fixed to the mounting bracket near the dialysis machine to ensure stable connection with the blood pipeline of the dialysis machine. The transmission cable is guided to the interface module of the central intelligent device through the cable trough. The entire data transmission process is automated and does not require manual intervention.

[0081] Electronic weighing scales are used to measure the weight of dialyzers and provide real-time weight data during the dialysis process. The process of data collection and transmission is as follows:

[0082] Data collection: The electronic weight scale is built into the bottom structure of the dialysis bed and is equipped with a high-precision load sensor, which can monitor and record the weight changes of the dialysis patient in real time while the dialysis patient is lying on the bed. The weight data is collected every 5 seconds and converted into a digital signal.

[0083] Data transmission: Digital weight data is transmitted to the central intelligent device through the built-in sensor interface. The data transmission cable uses a USB interface, and the transmission path is designed to ensure that the cable is hidden and does not interfere with other devices. The data transmission protocol uses the Modbus protocol to ensure the real-time and reliability of data transmission.

[0084] Connection method: The weight scale is fixedly connected to the dialysis bed, and the sensor interface is connected to the data transmission interface under the bed through the reserved wire groove. The data transmission cable is connected to the electrical interface box next to the bed, and then the data is transmitted to the interface module of the central intelligent device through the interface box.

[0085] The bioimpedance analyzer is used to measure the body fluid volume of the dialyzer and provides important health indicators by analyzing the changes in the water content in the body. The data collection and transmission process is as follows:

[0086] Data collection: The electrodes of the bioimpedance analyzer are connected to the surface of the dialyzer's body through wires. The electrodes emit a weak alternating current and measure the impedance change when the current passes through the body, thereby calculating the dialyzer's body fluid volume. The measured data is converted into a digital signal and transmitted to the signal processing unit for preliminary processing.

[0087] Data transmission: The signal processed data is transmitted to the central intelligent device through a high-speed data bus (such as CAN bus). The data transmission cable uses a highly sensitive shielded cable or coaxial cable to minimize external noise interference. The data transmission protocol uses the CAN protocol to ensure the real-time and high-precision data transmission.

[0088] Connection method: The electrodes of the bioimpedance analyzer are fixed on the brackets on both sides of the dialysis bed and connected to the signal processing unit through flexible wires. The signal processing unit and the central intelligent device are connected through the signal transmission interface installed under the bed to ensure the stability and continuity of signal transmission.

[0089] The central intelligent device is the data processing core of the entire system, responsible for receiving data from the online blood analyzer, electronic weight scale and bioimpedance analyzer, and performing analysis and prediction. The data receiving and processing methods are as follows:

[0090] Data reception: The central intelligent device is equipped with a variety of communication interface modules, including Ethernet interface, USB interface, RS-485 interface and CAN bus interface, to meet the data transmission requirements of different devices. All received data is first stored in the built-in memory of the device to ensure the integrity and security of the data.

[0091] Data processing: The stored data is analyzed in real time by the processor. Based on the preset machine learning model, the processor predicts and analyzes the data in real time and stores or sends the results to the central large screen for display.

[0092] Connection method: The central intelligent device is placed on the central console of the dialysis room and connected to each sensor device through a shielded cable. To ensure the stable operation of the equipment, all connection interfaces are insulated, and repeaters or signal amplifiers are used when necessary to ensure signal quality.

[0093] The central large screen is used to display the physiological parameter data collected and analyzed by the system in real time, as well as early warning information. Its connection and display methods are as follows:

[0094] Data display: The central large screen is connected to the central intelligent device via HDMI or VGA data cable to display the data from various sensor devices in real time. The display content includes real-time monitoring of urea concentration, creatinine concentration, potassium concentration, body weight, body fluid volume and other data, and automatically issues an alarm when the parameters exceed the threshold.

[0095] Connection method: The central large screen is installed on the wall or special bracket in the dialysis room and connected to the central intelligent device through a long-distance data cable. The data cable has been specially processed to reduce electromagnetic interference and ensure high-quality image and data display.

[0096] Through the design of the above connection and transmission methods, this embodiment ensures efficient collaboration between the various modules of the dialysis blood status real-time monitoring system, and realizes real-time collection, stable transmission, accurate analysis and timely warning of the dialyzer's physiological parameters.

[0097] Example 2: Figure 1 As shown, this embodiment shows in detail a method flow for real-time monitoring of dialysis blood status, covering the entire process from data collection, transmission, processing, prediction, warning to feedback. The method combines multiple sensor devices, advanced machine learning algorithms and efficient communication means to monitor the key physiological parameters of the dialyzer in real time and provide accurate health assessment and warning services, specifically including the following steps:

[0098] S1: Real-time acquisition of multi-parameter physiological data.

[0099] S1-1: Collection parameters and corresponding equipment: Urea concentration, creatinine concentration and potassium concentration are collected in real time by an online blood analyzer, body weight is measured in real time by an electronic weight scale, and body fluid volume is monitored in real time by a bioimpedance analyzer.

[0100] S1-2: Collection frequency and accuracy: All physiological parameters are collected every 5 seconds to ensure the real-time and continuity of the data. The collection equipment is calibrated to meet medical-grade accuracy requirements, including:

[0101] Urea concentration: ±0.1mg / dL

[0102] Creatinine concentration: ±0.05 mg / dL

[0103] Potassium concentration: ±0.02mmol / L

[0104] Weight: ±0.1kg

[0105] Body fluid volume: ±0.1L

[0106] S1-3: Equipment workflow:

[0107] Online blood analyzer: Automatic sampling: Through the sampling line connected to the blood circuit of the dialysis machine, a trace blood sample (about 0.5mL) is automatically drawn every 5 seconds to avoid additional burden on the dialyzer. Chemical analysis: The sample enters the analysis chamber and reacts with specific reagents to generate optical or electrochemical signals. Signal conversion: The signal is converted into digital data through the built-in sensor, and preliminary filtering and denoising are performed.

[0108] Electronic weight scale: Real-time measurement: Built-in high-precision load sensor to continuously monitor the weight changes of the dialyzer. Data stabilization: Adopt average filtering algorithm to eliminate measurement fluctuations caused by the dialyzer's slight movements.

[0109] Bioimpedance Analyzer: Electrode measurement: Through the four-electrode method, a weak alternating current (frequency 50kHz, current 1mA) is applied to both ends of the body of the dialyzer to measure the body impedance value. Parameter calculation: According to the impedance value, the body fluid volume is calculated using the formula:

[0110]

[0111] The correction factor is adjusted according to the age, gender and body size of the dialysis patient.

[0112] S1-4: Data formatting: All collected data are packaged in a unified data format, including the following fields: Timestamp: The exact time of data collection (format: YYYY-MM-DD HH:MM). Device ID: The unique identifier of the collection device. Parameter type: Such as urea concentration, body weight, etc. Value: The actual measured value. Data integrity check code: The integrity of the data during transmission is ensured by CRC check.

[0113] S2: Stable data transmission to the central intelligent device.

[0114] S2-1: Transmission network architecture: Wired transmission: Protocol: Adopt **Ethernet** protocol to support high-speed and stable data transmission. Connection: Each sensor device is connected to the switch interface of the central intelligent device through Cat6 shielded twisted pair cable, with a transmission bandwidth of up to 1Gbps. Wireless transmission (backup): Protocol: Adopt **Wi-Fi 6 (802.11ax)** standard to provide high-speed and low-latency wireless connection. Security: Enable WPA3 encryption protocol to ensure the security and privacy of data transmission. Application scenario: When the wired connection is limited or fails, it automatically switches to wireless transmission mode to ensure the continuity of data transmission.

[0115] S2-2: Data transmission process: Data packaging: Each device encapsulates the formatted data into a TCP / IP data packet, adding network information such as the source address and destination address. Transmission control: Confirmation mechanism: After receiving the data packet, the central intelligent device returns an ACK confirmation signal to the sending device to ensure that each data packet is successfully transmitted. Retransmission mechanism: If the ACK is not received within the set time, the sending device will automatically retransmit the data packet, and the maximum number of retransmissions is set to 3 times.

[0116] S2-3: Data reception and caching: Receiving module: The central intelligent device is equipped with a high-performance network interface card to monitor and receive data from various sensor devices in real time. Data caching: Short-term caching: The received data is first stored in the cache (RAM) for use by the real-time processing module; long-term storage: The processed data is regularly written to the persistent storage (such as SSD hard disk) to facilitate historical data query and analysis.

[0117] S3: Figure 3 As shown, the physiological parameter prediction based on the machine learning model specifically includes an algorithm, and the specific calculation formula of the algorithm is as follows:

[0118] P1: Model initialization:

[0119] Initialize state estimate Among them, y 0 is the initial parameter estimate, v 0is an estimate of the rate of change of a parameter, wherein the state specifically includes urea concentration, creatinine concentration, potassium concentration, body weight and body fluid volume;

[0120] Initialize the state transfer matrix and the measurement matrix H k =

[10] , where Δt is the sampling time interval, specifically 5 seconds;

[0121] P2: Process and measurement noise settings:

[0122] Set process noise covariance and the measurement noise covariance Among them, q yy Reflects the predicted change rate of urea concentration, specifically set to 0.1; q yv and q vy Reflects the relationship between urea concentration and change rate, specifically set to 0; q vv Reflects the prediction uncertainty of the urea concentration change rate itself, which is specifically set to 0.1; z k is the actual measured value of the k measurement point; y k is the reference value; N is the total duration;

[0123] P3: For each measurement point k, predict the parameter estimate at the next time point. The specific calculation formula is:

[0124]

[0125] Among them, P k - is the forecast error covariance;

[0126] P4: Repeat P2 and P3 to update the measurement, including:

[0127] Calculate the Kalman gain:

[0128] Update the state estimate:

[0129] Update the error covariance:

[0130] Among them, z k is the actual measurement at time k;

[0131] P5: The model prediction value at the last moment Output as the predicted value of physiological indicators at the next time point.

[0132] All physiological parameters are combined into a joint state vector, and the correlation between parameters is considered to improve the overall prediction accuracy.

[0133]

[0134] Covariance matrix expansion: Construct a joint error covariance matrix that contains the covariance terms between parameters and captures the mutual influence between parameters.

[0135] Parallel computing optimization: Utilize a multi-threaded parallel computing framework to update the predicted values ​​of each parameter simultaneously, reduce computing latency, and meet real-time requirements.

[0136] S4: Comparison and judgment of prediction results and safety thresholds.

[0137] S4-1: Safety threshold setting: Urea concentration: Normal range is 7-20mg / dL, and more than 25mg / dL is considered dangerous. Creatinine concentration: Normal range is 0.5-1.2mg / dL, and more than 1.5mg / dL is considered dangerous. Potassium concentration: Normal range is 3.5-5.0mmol / L, and more than 5.5mmol / L or less than 3.0mmol / L is considered dangerous. Weight: Set according to the baseline weight of the dialyzer before dialysis, more than **±2kg** is considered abnormal. Body fluid volume: Set according to the standard body fluid volume of the dialyzer, more than **±0.5L** is considered abnormal.

[0138] S4-2: Threshold comparison process: Single parameter judgment: Compare each predicted value with the corresponding safety threshold to determine whether there is an abnormality. Multi-parameter comprehensive judgment: Risk level assessment: According to the number and degree of abnormal parameters, the overall risk level is assessed and divided into low risk, medium risk and high risk. Trend analysis: Combined with historical data, it is determined whether the parameter abnormality is a short-term fluctuation or a continuous deterioration, providing a more accurate risk estimate.

[0139] S4-3: Decision rules: Low risk: If a parameter slightly exceeds the threshold and there is no continuous trend, it can be temporarily monitored without immediate intervention. Medium risk: If a parameter significantly exceeds the threshold, or multiple parameters are slightly abnormal, the system recommends that medical staff pay attention and may need to adjust dialysis parameters. High risk: If multiple parameters seriously exceed the threshold, or the abnormal trend is obvious, the system immediately triggers an emergency warning and recommends immediate medical intervention measures.

[0140] S5: Generation and transmission of early warning information.

[0141] S5-1: Generation of early warning information: Content composition: Basic information of the dialyzer: name, age, medical record number, etc. Abnormal parameter details: specific abnormal parameters, current predicted value, degree of excess. Risk assessment results: risk level, recommended measures. Timestamp: the exact time when the early warning was generated. Information format: The early warning information is generated in JSON format, which is convenient for parsing and display on multiple platforms.

[0142] S5-2: Early warning information transmission: Delivery channel: SMS notification: The early warning information is sent to the mobile phones of responsible doctors and nursing staff through the integrated SMS gateway. Mobile application push: The early warning notification is pushed in real time through the dedicated medical monitoring App, supporting detailed information viewing and feedback. Central large screen display: The abnormal situation is prompted by highlighting and sound alarm on the central large screen in the dialysis room. Email notification: The detailed early warning report is sent to the work mailbox of the relevant medical staff for subsequent tracking and recording.

[0143] S6: Continuous monitoring and loop execution.

[0144] Loop execution: The entire monitoring process from step S1 to step S5 is executed in a loop with a period of 5 seconds to ensure continuous monitoring of the dialyzer's status.

[0145] Through the method described in this embodiment, high-frequency, accurate, and real-time monitoring and prediction of dialysis blood physiological parameters are achieved, potential risks can be discovered in a timely manner, and timely and effective warnings and intervention suggestions can be provided. The entire system has the following advantages:

[0146] High real-time performance: A monitoring cycle of once every 5 seconds ensures immediate awareness of the status of the dialyzer.

[0147] High Accuracy: Combining multiple high-precision sensors and advanced machine learning algorithms to provide reliable prediction results.

[0148] High stability: Dual data transmission mode and perfect error handling mechanism are adopted to ensure stable operation of the system in various environments.

[0149] High scalability: The system architecture is flexible, and new monitoring parameters and functional modules can be added as needed to adapt to different medical needs.

[0150] High security: Comprehensive data encryption and backup strategies to protect the privacy and data security of dialyzers.

[0151] The successful implementation of this embodiment will significantly improve the safety and effectiveness of dialysis treatment, reduce the incidence of complications, and enhance the quality of life and medical service level of hemodialysis patients.

[0152] Embodiment 3: In this embodiment, by Figures 4 to 8 The changes of five key physiological parameters of the dialyzer during the dialysis process, namely, urea concentration, creatinine concentration, potassium concentration, body weight and body fluid volume, are shown. These charts not only provide visual support for the monitoring of physiological parameters during the dialysis process, but also provide simulation basis for verifying the effectiveness of the monitoring system and algorithm proposed in the present invention.

[0153] During the dialysis process, the change in urea concentration is one of the important indicators for evaluating dialysis efficiency. Figure 4The trend of urea concentration gradually decreasing with the passage of dialysis time is shown, reflecting the removal effect of metabolic waste during dialysis. The simulation shows that at most time points, the predicted value of urea concentration is basically consistent with the actual monitoring value, showing a trend of gradual decline, which is consistent with the expected results of dialysis. However, in the later stage of simulation (45 seconds), the actual measured value of urea concentration suddenly increases, and this abnormal fluctuation may be caused by temporary failure of dialysis equipment, unstable dialysis fluid or sudden change of physiological state of dialyzer. This abnormality is marked in the figure to remind medical staff to pay attention to potential dialysis efficiency problems or sudden physiological changes of dialyzers. The monitoring system of the present invention accurately predicts the change of urea concentration most of the time, indicating that the system can effectively monitor the urea concentration during dialysis. However, for the occurrence of abnormal values, the response of the system needs to be further optimized so that it can be predicted and issued earlier when similar events occur in the future to prevent potential crises.

[0154] Figure 5 The change of creatinine concentration during dialysis is shown. Creatinine is an important indicator of renal function. The decrease in its concentration means that dialysis is effectively removing metabolic waste from the body. The simulation shows that the actual measured value and the predicted value of creatinine concentration decrease synchronously in most cases, which shows that the system of the present invention can accurately predict the change trend of creatinine concentration and provide reliable data support. During the simulation, Figure 6 A slight outlier appeared at the 45th second, and the actual creatinine concentration was slightly higher than the predicted value. This abnormality may be caused by metabolic fluctuations of the dialyzer, temporary inefficiency of the dialyzer, or measurement error. Although the magnitude of the abnormal value is not large, it reminds us that the system needs higher sensitivity and accuracy in predicting and responding to sudden changes. When dealing with similar situations, medical staff should further observe and evaluate in combination with the overall condition of the dialyzer to ensure the continuity and safety of treatment.

[0155] Potassium concentrations must be carefully monitored during dialysis because either too high or too low a potassium level can have serious effects on heart function. Figure 6 The figure shows the trend of potassium concentration gradually decreasing during dialysis, which is a direct reflection of the removal of excess potassium ions from the body during dialysis. In the simulation diagram, the curve of potassium concentration changes showed a slight fluctuation at the 30th second, and the true value increased slightly. This small abnormality may be due to changes in the composition of the dialysate, a temporary increase in potassium load due to the dialyzer's diet, or a temporary decrease in potassium removal efficiency during dialysis. Although such fluctuations are not significant, they may indicate potential risks of larger problems in actual clinical settings. Therefore, the system should immediately alarm when potassium concentration fluctuations are detected, and advise medical staff to further check and adjust dialysis parameters to ensure that the dialyzer's electrolyte balance is maintained within a safe range.

[0156] Changes in body weight during dialysis directly reflect the amount of fluid removed. Figure 7 The simulation shows the gradual reduction in the patient's weight during dialysis, which is the result of the dialysis machine effectively removing excess fluid from the body. Overall, the change in weight is gentle and continuous, which is consistent with the normal dialysis process. At the 25th second, the simulation shows that the actual measured weight is slightly higher than the predicted value, forming a slight anomaly. This may be due to the patient's slight movement during dialysis or a measurement error of the equipment. Although the weight change is relatively small, accurate weight monitoring is crucial to adjusting the speed and total amount of fluid removal during dialysis. If the system can issue an early warning when a measurement error or abnormal weight change occurs, it can help medical staff better manage the patient's fluid balance and prevent the patient from dehydration or edema.

[0157] Fluid volume is another important parameter that is closely related to body weight. The reduction in fluid volume during dialysis reflects the successful removal of excess fluid from the body. Figure 8 It shows a trend of gradual decrease in body fluid volume during dialysis, which basically meets the expected goal of dialysis. However, at the 40th second, the actual value of body fluid volume showed a slight increase. This abnormality may be related to the measurement deviation of the equipment or the dialyzer's own fluid redistribution. Since changes in body fluid volume have a direct impact on the dialyzer's blood pressure, cardiac function, etc., this slight fluctuation needs to be taken seriously. When abnormal changes in body fluid volume are detected, the dialysis system should be able to quickly analyze the cause and provide adjustment suggestions, such as adjusting the composition or speed of the dialysate, to ensure the dialyzer's fluid balance and dialysis effect.

[0158] This embodiment demonstrates the changing trends of five key physiological parameters during dialysis through detailed simulation. The simulation results show that the monitoring system of the present invention can accurately capture most of the normal changes of these parameters and provide prompts when abnormalities occur. These simulation diagrams provide reliable data support for the practical application of the system, and also reveal the room for improvement of the system in predicting abnormalities and handling emergencies.

[0159] During the dialysis process, the stability of physiological parameters is crucial, and the monitoring system provided by the present invention can significantly improve the safety and effectiveness of the dialysis process by real-time prediction and monitoring of these key parameters. Although several minor anomalies occurred in the simulation, these anomalies remind us that we need to continuously optimize the sensitivity and response mechanism of the system to ensure timely and accurate warnings in actual clinical applications to protect the health and life safety of the dialyzer.

[0160] In summary, the simulation content of this embodiment shows the changes of key physiological parameters during the dialysis process through a series of detailed simulation diagrams, providing a valuable reference for system optimization and practical application.

[0161] The above embodiments only express the implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention.

Claims

1. A method for real-time monitoring of the state of dialysis blood, characterized in that: The specific steps include: S1: a sensor device integrated on the dialysis device, which continuously collects physiological parameters of the dialyzed blood, including urea concentration, creatinine concentration, potassium concentration, body weight and body fluid volume; S2: Collect the parameters every 5 seconds and transmit them to the central intelligent device in real time; S3: On the central smart device, use the machine learning model to predict the physiological parameter values ​​of the dialyzer in 5 seconds; S4: Make a judgment by comparing the differences between the parameter values ​​with the preset safety threshold; S5: The central large screen receives the physiological parameter values ​​stored and predicted by the central intelligent device, and displays real-time data and warning information; S6: If a parameter in step S4 is judged to exceed the threshold, the warning system is automatically triggered, and a warning notification is sent to the responsible doctor and nursing staff via SMS or application push; S7: During the entire dialysis process of the dialyzer, steps S1 to S6 are continuously repeated to maintain real-time monitoring and early warning of the dialyzed blood status.

2. A method for real-time monitoring of the state of dialysis blood according to claim 1, characterized in that: The sensor device in step S1 is specifically: T1: Online blood analyzer, used to collect urea concentration, creatinine concentration and potassium concentration; the measurement unit of urea concentration and creatinine concentration is mg / dL; the measurement unit of potassium concentration is mmol / L; T2: electronic weighing scale, used to collect body weight; the unit of measurement of body weight is kg; T3: Bioimpedance analyzer, used to collect body fluid volume; the unit of measurement of body fluid volume is L.

3. A method for real-time monitoring of the state of dialysis blood according to claim 1, characterized in that: In step S2, the central intelligent device includes the following steps to predict the values ​​of various physiological parameters of the dialyzer: S2-1: predicting five machine learning models on the central intelligent device, wherein the five machine learning models are specifically a urea concentration prediction model, a creatinine concentration prediction model, a potassium concentration prediction model, a body weight prediction model, and a body fluid volume prediction model; S2-2: Each machine learning model collects continuous and real-time data collected by corresponding sensor devices. The data collected by the urea concentration prediction model, creatinine concentration prediction model, and potassium concentration prediction model come from online blood analyzers, the data collected by the weight prediction model comes from electronic weight scales, and the data collected by the body fluid volume prediction model comes from bioimpedance analyzers; S2-3: Store the collected real-time data in the buffer of the central intelligent device, and continuously record at least the previous 10 minutes of data to form a data time series; S2-4: Use each machine learning model to analyze the corresponding data time series stored in the buffer, the model input is the data collected every 5 seconds in the past 10 minutes, and the output is the predicted value of each physiological parameter of the dialyzer at the next moment; Among them, the urea concentration prediction model is responsible for the prediction of urea concentration, the creatinine concentration prediction model is responsible for the prediction of creatinine concentration, the potassium concentration prediction model is responsible for the prediction of potassium concentration, the weight prediction model is responsible for the prediction of weight, and the body fluid volume prediction model is responsible for the prediction of body fluid volume. Each prediction model is equipped with the same physiological indicator prediction algorithm.

4. A method for real-time monitoring of the state of dialysis blood according to claim 3, characterized in that: The physiological index prediction algorithm specifically includes the following: P1: Model initialization: Initialize state estimate Wherein, y0 is the initial parameter estimate, v0 is the estimate of the parameter change rate, wherein the state specifically includes urea concentration, creatinine concentration, potassium concentration, body weight and body fluid volume; Initialize the state transfer matrix and the measurement matrix H k =[10], where Δt is the sampling time interval, specifically 5 seconds; P2: Process and measurement noise settings: Set process noise covariance and the measurement noise covariance Among them, q yy Reflects the predicted change rate of urea concentration, specifically set to 0.1; q yv and q vy Reflects the relationship between urea concentration and change rate, specifically set to 0; q vv Reflects the prediction uncertainty of the urea concentration change rate itself, which is specifically set to 0.1; z k is the actual measured value of the k measurement point; y k is the reference value; N is the total duration; P3: For each measurement point k, predict the parameter estimate at the next time point. The specific calculation formula is: in, is the forecast error covariance; P4: Repeat P2 and P3 to update the measurement, including: Calculate the Kalman gain: Update the state estimate: Update the error covariance: Among them, z k is the actual measurement at time k; P5: The model prediction value at the last moment Output as the predicted value of physiological indicators at the next time point.

5. A real-time monitoring system for the status of dialysis blood, characterized in that: The method for real-time monitoring of the state of dialysis blood used to perform any one of claims 1 to 6 specifically comprises the following: A central intelligent device, wherein the central intelligent device is configured with a processor and a memory, wherein the processor is used to execute a software program including data reception, analysis and prediction, and the memory is used to store real-time and historical data; A central large screen, electrically connected to the central intelligent device, for displaying the physiological parameters and warning information of the dialyzer; An online blood analyzer, comprising a plurality of sensors, wherein the sensors include sensors for measuring urea, creatinine and potassium concentrations, and the online blood analyzer is provided with a data transmission interface, through which the collected data is sent to the central intelligent device; An electronic weight scale, used to measure the dialyzer's weight, comprising a load sensor and a display, wherein the load sensor is used to detect the dialyzer's weight and send the weight data to the central intelligent device through a built-in transmission interface; A bioimpedance analyzer, comprising electrodes and a signal processing unit, wherein the electrodes are attached to the body surface of the dialyzer to measure the amount of body fluid, and the signal processing unit is used to process the data collected from the electrodes and send the data to the central intelligent device through a connecting line; The medical bed is equipped with a bed body for supporting the body of a dialyzer, and the bed body is provided with an interface for connecting with the electronic weight scale and the bioimpedance analyzer.

6. A real-time monitoring system for blood dialysis according to claim 5, characterized in that: The online blood analyzer further comprises: An automatic sampling mechanism for automatically extracting blood samples from the blood circulation of the dialysis machine at regular intervals; An analysis chamber, wherein a chemical reactor is installed in the analysis chamber for reacting with the blood sample to generate a detectable signal; The detector is used to detect the signal generated by the chemical reaction and convert the detection result into digital data, which is then sent through the data transmission interface.

7. A monitoring device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor runs the computer program to implement the method according to any one of claims 1 to 4.

Images