Intelligent hemodialysis system based on sodium clearance ratio and control method thereof

Abstract

The invention discloses an intelligent hemodialysis system based on a sodium clearance ratio (RNA) and a control method of the intelligent hemodialysis system. The system integrates five layers of functions to form a complete closed loop: (1) a sensing layer: based on a dialysate conductivity difference value, applying a corrected Gibbs-Donnan formula (Ceq = [Na] d / sigma, sigma = 0.97) and Kalman filtering to reversely deduce plasma sodium concentration in real time; (2) a calculation layer is used for calculating RNa = Naactual / Narget in real time, and RNa = Naactual / Narget; (3) a monitoring layer, namely an RNa + Kt / V double-index instrument panel + three-level alarm; (4) a [Na] d regulation and control curve, RNa closed-loop regulation and three safety rules are calculated through a control layer-double-chamber model; and (5) a feedback layer, cross-course trend analysis and automatic model optimization. The system is compatible with dialysis machines of various brands and can be integrated as an additional module.

Description

Technical Field

[0001] This invention relates to the field of blood purification equipment and intelligent medical monitoring technology, and in particular to an intelligent hemodialysis system and its control method based on sodium removal ratio (R_Na), which integrates three major functions: real-time estimation of plasma sodium concentration, online calculation and dual-indicator monitoring of R_Na, and dynamic closed-loop regulation of dialysate sodium concentration. Background Technology

[0002] Hemodialysis is the primary renal replacement therapy for patients with end-stage renal disease (ESRD). The assessment of dialysis adequacy has long relied on the urea clearance index Kt / V as the core standard. While Kt / V reflects the clearance efficiency of small molecule solutes (urea), it fails to reflect another crucial therapeutic dimension during dialysis—sodium balance.

[0003] Sodium is the main cation in extracellular fluid, determining its osmotic pressure and volume. Abnormal sodium balance in dialysis patients (insufficient sodium excretion leading to volume overload, or excessive sodium excretion leading to hypotension and muscle cramps) is a key factor affecting long-term prognosis. However, current dialysis equipment only uses Kt / V as an indicator of dialysis adequacy, lacking the ability to quantitatively monitor sodium clearance effectiveness.

[0004] The inventors proposed a novel sodium clearance ratio, R_Na = Na_actual / Na_target, as an indicator for assessing dialysis adequacy. Here, Na_target = (W_pre - W_dry) × [Na]s / 0.93 represents the target sodium clearance, and Na_actual represents the actual sodium clearance calculated based on a sodium kinetic model. R_Na = 1.0 indicates that sodium clearance is just within the target range, R_Na < 1.0 indicates insufficient sodium removal (hypersodium dialysis), and R_Na > 1.0 indicates excessive sodium removal (hyposodium dialysis).

[0005] However, to achieve real-time monitoring and closed-loop control of R_Na in dialysis equipment, three major technical challenges are faced: (1) Sensing layer - how to obtain plasma sodium concentration in real time without relying on blood sampling? (2) Monitoring layer - how to integrate R_Na with the existing Kt / V assessment system? (3) Control layer - how to dynamically adjust the sodium concentration of dialysate based on real-time feedback of R_Na? The existing dialysis machine's "sodium curve" function adopts a fixed mode, is not based on a sodium kinetic model, and cannot be individualized.

[0006] Existing dialysis machines (such as the Gambro / Baxter Diascan system and Fresenius OCM system) are equipped with dialysate inlet and outlet conductivity sensors, primarily used to calculate the ion dialysis volume to estimate Kt / V. Since sodium ions account for over 95% of the total cations in the dialysate, conductivity is highly correlated with sodium concentration. However, a key drawback of existing methods is that the traditional Gibbs-Donnan formula uses C_eq = σ × [Na]d (σ = 0.97), while the correct formula is C_eq = [Na]d / σ, resulting in approximately 6% systematic error.

[0007] Therefore, there is an urgent need for an intelligent dialysis system that integrates real-time plasma sodium estimation, online monitoring of R_Na, and dynamic closed-loop regulation of [Na]d. Summary of the Invention

[0008] The technical problem to be solved by this invention is to provide an intelligent hemodialysis system and its control method based on sodium clearance ratio R_Na, which integrates three major functions: real-time estimation of plasma sodium concentration (sensing layer), real-time calculation of R_Na and dual-index monitoring (monitoring layer), and dynamic closed-loop regulation of dialysate sodium concentration (control layer), forming a complete closed loop of "sensing → calculation → monitoring → control → feedback".

[0009] To achieve the above objectives, the present invention adopts the following technical solution: I. Sensing Layer – Real-time Estimation of Plasma Sodium Concentration Step S1: Real-time conductivity acquisition. Measure the dialysate inlet conductivity κ_in and outlet conductivity κ_out in real time at a frequency of at least once per minute. A built-in conductivity sensor of the dialysis machine (such as Diascan or OCM systems) can be used, or an external high-precision conductivity sensor can be connected.

[0010] Step S2: Conductivity-Sodium Concentration Conversion. The conductivity value is converted to a sodium concentration value using a pre-calibrated conversion function f(κ). Optionally, the conductivity contribution of non-sodium ions is corrected.

[0011] Step S3: Calculation of sodium mass transfer. ΔNa = Q_d × ([Na]d_out - [Na]d_in).

[0012] Step S4: Correct Donnan equilibrium and reverse the plasma sodium calculation. C_eq = [Na]d_out / σ (σ = 0.97), reverse the calculation of [Na]pw using the mass balance equation, and then convert it to [Na]s = 0.93 × [Na]pw.

[0013] Step S5: Kalman filtering for noise reduction. The process model is based on the sodium kinetic equations of the two chambers, and the observation model is based on the difference in conductivity.

[0014] Step S6: Online verification and calibration. Automatic correction is performed when the deviation exceeds 2 mmol / L.

[0015] Computational Layer – R_Na Real-Time Computation Step S7: Na_target = (W_pre - W_dry) × [Na]s_0 / 0.93.

[0016] Step S8: R_Na = Na_actual / Na_target, updated every 5-15 minutes.

[0017] III. Monitoring Layer – Dual Indicator Display and Alarms Step S9: Dual indicator display. R_Na + Kt / V are displayed side-by-side on the dashboard, color-coded.

[0018] Step S10: Level 3 Alarm. Yellow → Orange → Red.

[0019] IV. Control Layer – Dynamic Regulation of Sodium Concentration in Dialysate Step S11: Calculate the initial control curve. "High at the beginning and low at the end" phased strategy.

[0020] Step S12: Closed-loop regulation based on R_Na. R_Na < 0.9 decreases [Na]d; R_Na > 1.1 increases [Na]d.

[0021] Step S13: Three safety correction rules: blood pressure protection, sodium clearance progress, and serum sodium protection.

[0022] Step S14: Safety protection. [Na]d is locked at [130, 150] mmol / L.

[0023] V. Feedback Layer – Post-Analysis Assessment and Learning Step S15: Post-dialysis deviation analysis and parameter optimization.

[0024] Step S16: Cross-treatment trend management and prescription recommendations.

[0025] Step S17: Standardized data output (HL7 FHIR) and centralized monitoring of multiple patients. Attached Figure Description

[0026] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1This is a diagram of the overall architecture of the intelligent hemodialysis system based on sodium clearance ratio of the present invention, showing the integrated structure of five layers: sensing layer, computing layer, monitoring layer, control layer, and feedback layer, as well as the data flow between each layer.

[0028] Figure 2 This is a flowchart of the real-time plasma sodium concentration estimation process of the present invention, which describes in detail the complete steps from conductivity acquisition, conversion, sodium mass transfer calculation, Donnan correction to back-calculate plasma water and sodium concentration, Kalman filtering to the final output of plasma sodium concentration.

[0029] Figure 3 This is a schematic diagram of the dual-indicator dashboard for R_Na and Kt / V and the three-level alarm system of the present invention. It shows the interface layout in which the two indicators of dialysis adequacy are displayed side by side, as well as the triggering conditions and presentation methods of the three-level alarms: yellow, orange, and red.

[0030] Figure 4 The flowchart for the dynamic closed-loop control of dialysate sodium concentration in this invention illustrates the decision-making logic of initial curve generation, R_Na feedback regulation, three safety correction rules (blood pressure protection, sodium clearance progress, and serum sodium protection), and the cyclical process of model prediction and curve update.

[0031] Figure 5 This is a schematic diagram of the hardware structure of the intelligent dialysis system of the present invention, showing the connection relationship of modules such as sodium concentration sensor, controller, sodium concentration adjustment mechanism, and communication interface, as well as the interaction method with the dialysis machine main control system and hospital information system.

[0032] Figure 6 This is a schematic diagram of the interface of the multi-patient R_Na centralized monitoring workstation of the present invention, showing the management interface layout that simultaneously displays information such as the dynamic R_Na, Kt / V value, and alarm status of multiple patients.

[0033] The technical solution and its beneficial effects of the present invention can be more fully understood through the above figures and specific embodiments. Detailed Implementation

[0034] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the specific embodiments. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this patent, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this patent.

[0035] Example 1: Intelligent Hemodialysis System Based on Sodium Clearance Ratio Reference Figure 1This invention provides an intelligent hemodialysis system based on sodium clearance ratio, comprising a sensing module 100, a calculation module 200, a monitoring module 300, a control module 400, and a feedback module 500, which are connected sequentially to form a closed loop. This system can be integrated as an add-on module with existing dialysis machines, acquiring real-time data from the dialysis machine and outputting control commands via a communication interface.

[0036] The sensing module 100 is used to estimate plasma sodium concentration in real time; its specific structure and workflow are described in [reference needed]. Figure 2 The sensing module 100 includes: - Conductivity sensors are installed at the inlet and outlet of the dialysate line to collect the inlet and outlet conductivity of the dialysate at a frequency of not less than once per minute; - The conversion unit uses the conductivity-sodium concentration relationship curve established in advance through multi-point calibration to convert conductivity into sodium concentration, thereby obtaining the dialysate inlet sodium concentration [Na]d_in and outlet sodium concentration [Na]d_out; - Sodium mass transfer calculation unit calculates the amount of sodium mass transferred per unit time according to the formula ΔNa = Q_d × ([Na]d_out - [Na]d_in), where Q_d is the dialysate flow rate (provided by the dialysis machine or preset); - The plasma sodium concentration back-calculation unit, based on the corrected Donnan equilibrium formula C_eq = [Na]d / σ (σ=0.97 is the Gibbs-Donnan factor) and the mass balance equation, back-calculates the plasma sodium concentration [Na]pw at the dialyzer outlet, and then obtains the plasma sodium concentration [Na]s according to [Na]s = 0.93 × [Na]pw; the specific back-calculation process is as follows: inside the dialyzer, plasma water and dialysate reach Donnan equilibrium at the outlet, that is, [Na]pw_out = [Na]d_out / σ. Then, combined with the dialyzer inlet plasma sodium concentration [Na]pw_in and ultrafiltration rate, the real-time [Na]pw is obtained by iteratively solving the mass balance equation. - Kalman filter, using a two-chamber sodium kinetic model as the process model and conductivity difference (or the observed value derived from it) as the observation model, performs noise reduction on the estimated plasma sodium concentration; the two-chamber sodium kinetic model divides human body fluid into two chambers, extracellular fluid and intracellular fluid, describing the transfer of sodium between the chambers and the dialysis clearance process. The model parameters include the volume of each chamber, transfer coefficient, etc., and can be initialized according to the individual characteristics of the patient; - Online validation and automatic calibration unit automatically corrects key parameters (such as sodium transfer coefficient K_d) in the dual-chamber model when the estimated plasma sodium concentration deviates from the occasional measured value of blood sampling by more than 2 mmol / L, in order to improve the accuracy of subsequent estimations.

[0037] The calculation module 200 is used to calculate the sodium removal ratio R_Na in real time, and includes: - The target sodium clearance calculation unit calculates the target sodium clearance according to the formula Na_target = (W_pre - W_dry) × [Na]s_0 / 0.93, where W_pre is the predialysis body weight (which can be obtained from the dialysis machine or electronic scale), W_dry is the dry body weight (preset value), and [Na]s_0 is the predialysis plasma sodium concentration (provided by the sensor module or manually input). - The actual sodium removal calculation unit integrates the sodium mass transfer amount ΔNa per unit time over time to obtain the cumulative actual sodium removal amount Na_actual; - The R_Na calculation unit updates R_Na = Na_actual / Na_target every 5 to 15 minutes and outputs the results to the monitoring module and the control module.

[0038] The monitoring module 300 is used to display both R_Na and Kt / V indicators and has a three-level alarm mechanism. For example... Figure 3 As shown, the monitoring interface uses a dashboard format to display R_Na and Kt / V side-by-side, with color coding indicating compliance status: green indicates compliance (R_Na between 0.95 and 1.05, Kt / V ≥ 1.2), yellow indicates slight deviation, orange indicates moderate deviation, and red indicates severe deviation. The three-level alarm is as follows: - Yellow alert: R_Na deviates from 1.0 but remains between 0.9 and 1.1 for more than 10 minutes, triggering a local notification to alert medical staff; - Orange Alarm: R_Na exceeds the range of 0.8 to 1.2, or continues to deviate for more than 30 minutes, triggering a notification to the nurse station and flashing on the monitoring interface; - Red Alert: When R_Na exceeds the range of 0.7 to 1.3, or is accompanied by dangerous situations such as a sharp drop in blood pressure, the entire system will be triggered by an audible and visual alarm and safety protection will be automatically activated (such as pausing regulation and restoring the default sodium concentration).

[0039] The control module 400 is used for dynamic closed-loop regulation of the sodium concentration [Na]d in the dialysate, referring to... Figure 4 It includes: - The initial curve generation unit, based on the dual-chamber sodium kinetic model and individual patient parameters (dry weight, predialysis weight, predialysis serum sodium, etc.), generates an initial [Na]d regulation curve in stages of "high at the beginning and low at the end" through model simulation optimization; for example, with a total dialysis time of 4 hours, the [Na]d level is maintained at 142 mmol / L for the first 100 minutes and then linearly decreases to 136 mmol / L for the next 140 minutes. - The R_Na feedback adjustment unit compares R_Na with the target value of 1.0 in real time during dialysis: if R_Na < 0.9, the [Na]d setpoint is reduced (e.g., reduced by 1.5 mmol / L) to increase sodium excretion; if R_Na > 1.1, the [Na]d setpoint is increased (e.g., increased by 1.5 mmol / L) to reduce sodium excretion; the adjustment step size and frequency can be set according to clinical needs. - Safety correction unit, integrating three safety correction rules, with higher priority than R_Na feedback adjustment: 1) Blood pressure protection rule: If the systolic blood pressure drops by more than 20 mmHg from the baseline value, the current [Na]d will be increased by 3 mmol / L immediately, the ultrafiltration rate will be reduced by 20%, and the remaining time curve will be recalculated by the model; 2) Sodium clearance progress rule: If the cumulative sodium clearance is less than 80% of the expected progress (the expected progress is calculated based on the initial control curve), and blood pressure is stable and serum sodium is not lower than 135 mmol / L, then [Na]d will be reduced by 1.5 mmol / L to accelerate sodium excretion. 3) Serum sodium protection rule: If the model predicts that serum sodium will be lower than 135 mmol / L before the end of dialysis, [Na]d will be automatically increased by 2 mmol / L, and the ultrafiltration rate will be increased if necessary to compensate for sodium removal through convection. - The safety lockout unit keeps [Na]d within a safe range of 130–150 mmol / L. If any control command exceeds this range, it will automatically lock at the boundary value and issue an alarm. At the same time, it monitors the rate of change of plasma sodium. If |d[Na]s / dt| > 3 mmol / L / h, it will issue a rapid change alarm and suspend control.

[0040] The feedback module 500 is used for post-dialysis evaluation and parameter optimization, and includes: - Deviation analysis unit: After dialysis, the deviation between the actual total sodium clearance and the target value is calculated, and the deviation information is stored in the patient's file; - Cross-treatment trend management unit: When the same patient’s R_Na deviates from 1.0 multiple times in a row (e.g., R_Na < 0.9 for 3 consecutive times), the system provides prescription adjustment suggestions (such as adjusting dry weight, modifying initial curve parameters, etc.). - Data output unit, according to HL7 FHIR standard, outputs data such as R_Na, Kt / V, control curve, alarm events, etc. of each dialysis to the hospital information system (HIS / EMR / DIMS). - Multi-patient centralized monitoring workstation, such as Figure 6 As shown, it is used to simultaneously display information such as R_Na dynamics, Kt / V values, and alarm status of multiple patients, which facilitates centralized management by medical staff.

[0041] Example 2: Real-time estimation method for plasma sodium concentration This embodiment details the specific steps by which the sensing module 100 achieves real-time estimation of plasma sodium concentration. Refer to... Figure 2 .

[0042] Step 2.1: Collect the inlet conductivity Cd_in and outlet conductivity Cd_out of the dialysate at a fixed frequency of 1 time / minute.

[0043] Step 2.2: Using a pre-calibrated conductivity-sodium concentration conversion curve, convert Cd_in and Cd_out to sodium concentrations [Na]d_in and [Na]d_out. The calibration method is as follows: prepare a series of dialysate standard solutions with different sodium concentrations (covering 130–150 mmol / L), measure their conductivity, fit a linear or quadratic polynomial relationship, and calibrate periodically.

[0044] Step 2.3: Calculate the sodium mass transfer per unit time ΔNa = Q_d × ([Na]d_out - [Na]d_in), where Q_d is the dialysate flow rate (usually 500 mL / min or set according to the prescription).

[0045] Step 2.4: According to the corrected Donnan equilibrium formula, the plasma sodium concentration at the dialyzer outlet [Na]pw_out = [Na]d_out / σ, where σ is taken as 0.97. Simultaneously, according to mass balance, the plasma sodium concentration at the dialyzer inlet [Na]pw_in, the plasma sodium concentration at the outlet [Na]pw_out, the ultrafiltration rate QUF, and the blood flow rate QB satisfy the following relationship: QB × [Na]pw_in = QB × [Na]pw_out + QUF × [Na]pw_out (assuming the ultrafiltrate sodium concentration equals the plasma sodium concentration; this is a simplified mass balance equation for convective sodium transfer, while diffuse sodium transfer has been calculated separately using the conductivity difference method). From this, [Na]pw_in, i.e., the plasma sodium concentration at the current moment, can be deduced. In actual calculations, dynamic updates of the dual-chamber model need to be considered.

[0046] Step 2.5: Convert plasma water and sodium concentration to serum sodium concentration: [Na]s = 0.93 × [Na]pw, where 0.93 is the volume fraction of plasma water in serum (because about 7% of serum is solid components such as protein).

[0047] Step 2.6: Optimal estimation of [Na]s is performed using a Kalman filter. The state equation is a two-chamber sodium kinetic model, and the observation equation is [Na]s obtained in Step 2.5 (which can be considered as noisy observations). The Kalman filter, through a prediction-update step, outputs a smoothed estimate of plasma sodium concentration, effectively suppressing measurement noise and model errors.

[0048] Step 2.7: Online Validation and Automatic Calibration. When there are actual blood sampling values ​​(e.g., blood is occasionally drawn before or during dialysis), the estimated value is compared with the measured value. If the deviation exceeds 2 mmol / L, the parameters in the dual-chamber model (such as extracellular fluid volume, transfer coefficient, etc.) are automatically adjusted to make the model more closely match the individual characteristics of the patient.

[0049] Example 3: Real-time R_Na Calculation Method This embodiment details the workflow of the calculation module 200.

[0050] Step 3.1: Before dialysis begins, obtain the patient's dry weight W_dry (set by the doctor), predialysis weight W_pre (input from the electronic scale), and predialysis serum sodium [Na]s_0 (estimated by the sensor module or manually entered). Calculate the target sodium clearance: Na_target = (W_pre - W_dry) × [Na]s_0 / 0.93. For example, if W_pre = 67 kg, W_dry = 65 kg, and [Na]s_0 = 140 mmol / L, then Na_target = 2 × 140 / 0.93 ≈ 301 mmol.

[0051] Step 3.2: During dialysis, the amount of sodium mass transferred per unit time, ΔNa, is integrated every 5 minutes (or according to a set interval) to obtain the cumulative actual sodium removal amount, Na_actual = ∫ΔNa dt. The integration starts at the beginning of dialysis, and the integration step size is synchronized with the sampling cycle.

[0052] Step 3.3: Calculate R_Na = Na_actual / Na_target, update it every 5 to 15 minutes, and send the result to the monitoring module and control module.

[0053] Example 4: Dynamic Closed-Loop Control Method for Sodium Concentration in Dialysis Fluid This embodiment details the specific steps by which the control module 400 implements dynamic regulation. (Refer to...) Figure 4 .

[0054] Step 4.1: Before dialysis, an initial [Na]d regulation curve is generated using a two-chamber model simulation based on patient parameters (dry weight, pre-dialysis weight, pre-dialysis serum sodium, etc.). Design principle: A higher sodium concentration (e.g., 142 mmol / L) is used for the first 40%–50% of the time to maintain stable plasma osmolality; the sodium concentration is gradually reduced (e.g., linearly decreased to 136 mmol / L) for the subsequent 50%–60% of the time to increase diffuse sodium excretion. The model verifies that this curve enables the total sodium clearance to reach the target value.

[0055] Step 4.2: After dialysis begins, the system monitors parameters such as blood pressure, actual ultrafiltration volume, and estimated plasma sodium in real time, and calculates the current R_Na value every 10 minutes.

[0056] Step 4.3: Compare R_Na with the target value of 1.0 and perform feedback adjustment: - If R_Na < 0.9, it indicates insufficient sodium excretion. Reduce the current [Na]d setting by 1.5 mmol / L (but not below 130 mmol / L) to enhance the diffusion gradient; - If R_Na > 1.1, it indicates excessive sodium excretion. Increase the current [Na]d setting by 1.5 mmol / L (but not higher than 150 mmol / L) to reduce the diffusion gradient. - After adjustment, the control curve for the remaining time is recalculated using the sodium kinetic model to ensure that the final total sodium removal still approaches the target.

[0057] Step 4.4: Based on the feedback adjustment, check the three safety correction rules in real time, and execute the corresponding actions first if they are triggered: - Rule 1 (Blood Pressure Protection): If the systolic blood pressure drops by more than 20 mmHg from the baseline value, immediately increase [Na]d by 3 mmol / L and reduce the ultrafiltration rate by 20%, while reprogramming the remaining curve; - Rule 2 (Sodium clearance progress): If the cumulative sodium clearance is less than 80% of the expected progress (the expected progress is allocated by time period according to the initial curve), and blood pressure is stable and serum sodium is ≥135 mmol / L, then [Na]d will be reduced by 1.5 mmol / L; - Rule 3 (Serium sodium protection): If the model predicts that serum sodium will be below 135 mmol / L before the end of dialysis, increase [Na]d by 2 mmol / L and the ultrafiltration rate may be increased appropriately.

[0058] Step 4.5: When executing the control command, always ensure that [Na]d is within the range of 130–150 mmol / L. If the correction command causes the limit to be exceeded, lock at the boundary value and issue an alarm. At the same time, monitor the rate of change of plasma sodium. If |d[Na]s / dt| > 3 mmol / L / h, immediately pause the control and issue a rapid change alarm.

[0059] Step 4.6: After dialysis, save the actual control curve, event records, etc. to the database for analysis by the feedback module.

[0060] Example 5: Specific Structure and Working Process of the Hardware Device Reference Figure 5 The hardware device for implementing the above method includes: - Sodium concentration sensor: Installed on the dialysate line, it can be an ion-selective electrode or a conductivity sensor to monitor the sodium concentration of the dialysate in real time and send the data to the controller; - Controller: Built-in microprocessor, stores the dual-chamber sodium kinetic model and the program for executing dynamic regulation algorithms, receives sensor data and real-time information such as blood pressure and ultrafiltration volume obtained through the communication interface, and outputs regulation commands; - Sodium concentration adjustment mechanism: Driven by a controller, it precisely changes the sodium concentration of the dialysate by adjusting the mixing ratio of concentrated sodium solution (such as high-concentration sodium chloride solution) and pure water; specifically, it can be controlled by a proportional pump or an electric valve; - Communication interface: Connects to the main control system of the dialysis machine to obtain data such as blood pressure, ultrafiltration rate, and blood flow. It can also interface with the hospital information system or AI-assisted decision-making system to achieve data sharing and remote monitoring. - Human-machine interface: Used to display monitoring parameters, control curves, alarm information, and receive parameters and instructions input by doctors.

[0061] Operating Process: Before dialysis, the doctor inputs parameters such as the patient's dry weight and pre-dialysis weight through the human-machine interface. The controller automatically generates and displays the initial control curve. During dialysis, the controller collects data such as sodium concentration and blood pressure in real time, calculates the new target sodium concentration according to the dynamic control algorithm, and drives the sodium concentration adjustment mechanism to adjust the mixing ratio. All monitoring parameters and control records can be exported through the communication interface and viewed on a multi-patient centralized monitoring workstation.

[0062] Example 6: Centralized Monitoring and Feedback Optimization for Multiple Patients This embodiment illustrates the specific application of the feedback module 500. It describes the deployment of a multi-patient centralized monitoring workstation (e.g., in a dialysis center) Figure 6 As shown), it can display patient information for multiple beds simultaneously, including: - Patient name / ID; - Current R_Na value and trend curve; - Real-time estimated Kt / V value; - The set value and actual value of sodium concentration in the dialysate; - Vital signs such as blood pressure and heart rate; - Alarm status (yellow / orange / red).

[0063] Healthcare staff can promptly detect and intervene in abnormalities through the workstation. After each dialysis session, the system automatically performs deviation analysis, comparing the actual total sodium clearance with the target value and storing the deviation in the patient's electronic record. The cross-treatment trend analysis module tracks the patient's R_Na levels over multiple consecutive sessions. If a persistent deviation occurs (e.g., R_Na < 0.9 for three consecutive sessions), the system prompts the physician to adjust the dry weight or modify the initial curve parameters. Furthermore, all data is output to the hospital information system according to the HL7 FHIR standard, facilitating research and quality management.

[0064] As can be seen from the above embodiments, the present invention achieves the following beneficial effects: 1. For the first time, a five-layer integrated intelligent dialysis system integrating "sensing → calculation → monitoring → control → feedback" has been realized, combining the three major functions of plasma sodium estimation, R_Na monitoring and [Na]d regulation into a closed loop.

[0065] 2. For the first time, the correction Donnan factor (C_eq=[Na]d / σ, σ=0.97) was used to back-calculate serum sodium from conductivity, eliminating approximately 6% systematic error in the traditional formula.

[0066] 3. Kalman filtering combined with a two-chamber sodium kinetic model achieves serum sodium estimation accuracy of ±0.5 mmol / L.

[0067] 4. Establish a dual-index assessment system for dialysis adequacy, including R_Na+Kt / V, to overcome the limitation of existing equipment that only focuses on urea removal.

[0068] 5. R_Na closed-loop control replaces the fixed-mode sodium curve, enabling individualized adaptive sodium balance management.

[0069] 6. Three security correction rules and three levels of alarms ensure system security.

[0070] 7. Online validation and feedback learning continuously improve accuracy and adapt to individual differences.

[0071] 8. Compatible with various brands of dialysis machines, with an add-on module design that eliminates the need for machine replacement and enables low-cost deployment.

[0072] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A control method for an intelligent hemodialysis system based on sodium clearance ratio, characterized in that, Includes the following steps: Step S1: Sensing Layer – Real-time Estimation of Plasma Sodium Concentration. During dialysis, the inlet conductivity κ_in and outlet conductivity κ_out of the dialysate are measured in real time. The conductivity values ​​are converted to sodium concentration values ​​using a pre-calibrated conductivity-sodium concentration conversion relationship. The inlet sodium concentration [Na]d_in and outlet sodium concentration [Na]d_out of the dialysate are calculated. The sodium mass transfer per unit time ΔNa = Q_d × ([Na]d_out - [Na]d_in) is calculated based on the dialysate flow rate Q_d. The diffusion equilibrium concentration is calculated using the corrected Gibbs-Donnan equilibrium formula C_eq = [Na]d_out / σ (σ = 0.97 is the Gibbs-Donnan factor). The current plasma sodium concentration [Na]pw is back-calculated using the sodium mass balance equation ΔNa = K_d × (C_eq - [Na]pw) + Q_uf × [Na]pw × S_Na, and then converted to serum sodium concentration [Na]s = 0.93 × [Na]pw. Step S2: Calculation Layer – Real-time Calculation of Sodium Clearance Ratio R_Na. Obtain the patient's predialysis weight W_pre, dry weight W_dry, and predialysis serum sodium concentration [Na]s_0. Calculate the target sodium clearance Na_target = (W_pre - W_dry) × [Na]s_0 / 0.93; calculate the cumulative actual sodium clearance Na_actual based on the sodium mass transfer time integral from Step S1; calculate the real-time sodium clearance ratio R_Na = Na_actual / Na_target. Step S3: Monitoring Layer – Dual Indicator Display and Alarm. The dialysis equipment interface simultaneously displays two indicators: R_Na and Kt / V. R_Na reflects the adequacy of sodium balance, and Kt / V reflects the adequacy of urea clearance. When R_Na deviates from the target range [0.9, 1.1], a graded alarm signal is generated. Step S4: Control Layer – Dynamic Regulation of Dialysate Sodium Concentration. Based on a dual-chamber sodium kinetic model, the [Na]d regulation curve of dialysate sodium concentration throughout the dialysis process is calculated, employing a phased strategy of high concentration at the beginning and low concentration at the end. During dialysis, the regulation curve is dynamically corrected based on the real-time R_Na value obtained in Step S2: when R_Na < 0.9, [Na]d is decreased by 1-3 mmol / L to increase sodium excretion; when R_Na > 1.1, [Na]d is increased by 1-3 mmol / L to reduce sodium excretion, forming a closed-loop control that drives R_Na to approach 1.

0. Step S5: Feedback Layer – Post-dialysis Assessment and Learning. At the end of dialysis, calculate the final R_Na and Kt / V values, analyze the deviation between the actual sodium clearance and the target value, and identify the sources of this deviation. Feedback this deviation information to optimize initial parameters and calibrate the model for the next dialysis session.

2. The method as described in claim 1, characterized in that, Step S1 also includes a non-sodium ion correction step: measuring or pre-setting the concentration of other major ions (potassium, calcium, magnesium, bicarbonate) in the dialysate, calculating the non-sodium ion conductivity contribution value κ_non-Na based on the contribution coefficient of each ion to the total conductivity, subtracting the non-sodium contribution from the total conductivity to obtain the sodium-attributed conductivity κ_Na = κ_total - κ_non-Na, and using κ_Na instead of κ_total for sodium concentration calculation to improve estimation accuracy.

3. The method as described in claim 1, characterized in that, Step S1 also includes a signal filtering step: the original estimate of [Na]pw obtained by back-reasoning is digitally filtered using a Kalman filter. The state variable of the Kalman filter is [Na]pw. The process model is based on the two-chamber sodium dynamics equation (including the perturbation reference ICF exchange model j_ic = K_ic × [(C_e - C_i) - (C_e0 - C_i0)]). The observation model is based on the conductivity-sodium concentration conversion relationship, so that the filtered estimate takes into account both the real-time measurement and the physical constraints of sodium dynamics.

4. The method as described in claim 1, characterized in that, Step S1 also includes an online verification and automatic calibration step: when a blood sample is collected during dialysis and the laboratory test value [Na]s_lab is obtained, the test value is compared with the estimated value [Na]s_est at the same time, and the deviation δ = [Na]s_lab - [Na]s_est is calculated; when |δ| > 2 mmol / L, the conductivity-sodium concentration conversion function parameter and / or the estimated value of diffusion clearance K_d are automatically corrected; the calibration results are stored in the patient's personal parameter file for subsequent dialysis.

5. The method as described in claim 1, characterized in that, The dual-indicator display in step S3 uses a parallel dashboard interface. The left side displays the real-time value of R_Na and its trend curve over time, while the right side displays the real-time value of Kt / V and its trend curve. The two indicators are identified by independent target areas: the target area for R_Na is [0.9, 1.1], and the target area for Kt / V is ≥1.2 (HD) or ≥1.0 (online HDF). When the indicator is within the target area, it is displayed in green, and when it deviates, it is displayed in yellow or red.

6. The method as described in claim 1, characterized in that, The graded alarm in step S3 adopts a three-level mechanism: Yellow alert: R_Na deviates slightly (0.8 ≤ R_Na < 0.9 or 1.1 < R_Na ≤ 1.2), the alert is only displayed on this machine; Orange alarm: R_Na deviates moderately (0.7 ≤ R_Na < 0.8 or 1.2 < R_Na ≤ 1.3), the machine displays the error and notifies the responsible nurse; Red Alert: R_Na deviates significantly (R_Na < 0.7 or R_Na > 1.3), triggering alarms simultaneously on the local machine, central monitoring station, and doctor's mobile terminal, and automatically triggering the safety protection mode to lock [Na]d to the same level as serum sodium.

7. The method as described in claim 1, characterized in that, The dialysate sodium concentration control curve in step S4 is calculated using a dual-chamber sodium kinetic model, rather than being directly determined by a preset empirical model; the phased strategy is as follows: Phase 1 (the first 40%-50% of dialysis time): [Na]d is set to be close to or slightly higher than the patient's serum sodium concentration to reduce the rate of sodium excretion through diffusion, maintain stable plasma osmotic pressure, and avoid early hypotension; The second stage (the last 50%-60% of dialysis time): [Na]d is set 2-5 mmol / L lower than the serum sodium concentration to increase the diffusion sodium excretion gradient and accelerate sodium excretion to ensure that the final R_Na value approaches 1.

0.

8. The method as described in claim 1, characterized in that, The dynamic correction in step S4 also includes the following rules: Blood pressure protection rule: When the patient's systolic blood pressure drops beyond the preset threshold (20 mmHg), the current [Na]d will be increased by 2-4 mmol / L and the ultrafiltration rate will be reduced. The subsequent curve will be recalculated by the sodium kinetic model. Sodium clearance progress rule: When the cumulative sodium clearance is less than 80% of the target progress, reduce [Na]d by 1-2 mmol / L to accelerate sodium excretion, provided that blood pressure is stable and the predicted serum sodium is not lower than the safety limit. Serum sodium protection rule: When the model predicts that serum sodium will be below the safety limit (135 mmol / L), increase [Na]d to reduce diffuse sodium excretion, and if necessary, increase ultrafiltration to compensate for convective sodium excretion.

9. The method as described in claim 1, characterized in that, In step S4, the sodium concentration of the dialysate is always kept within the safe range of [130, 150] mmol / L. When the dynamically corrected sodium concentration exceeds this range, it is automatically locked at the nearest safe boundary value. The frequency of the dynamic correction is once every 5-15 minutes.

10. The method as described in claim 1, characterized in that, Step S5 also includes cross-treatment trend management: Record the final values ​​of R_Na and Kt / V at the end of each dialysis session; The trend of R_Na in multiple consecutive dialysis sessions is analyzed using a time series approach. When the final R_Na value deviates from the target range for more than three consecutive consecutive times, prescription adjustment suggestions are generated. Generate weekly / monthly sodium balance trajectory reports, including R_Na mean, standard deviation, compliance rate, and comparative analysis with Kt / V compliance rate.

11. The method as described in claim 1, characterized in that, The method also includes standardized data communication steps: The following data are packaged and output using the HL7 FHIR or IHE dialysis extended protocol: dialysis date and time, patient identification, pre- and post-dialysis weight, pre- and post-dialysis serum sodium concentration (including real-time estimated value sequence), Na_target, Na_actual, R_Na final value, Kt / V final value, dialysate sodium concentration execution curve, and alarm event records. Transmitted to HIS / EMR / DIMS via network interface to achieve full-process traceability; It supports centralized monitoring of R_Na for multiple patients, and displays the real-time R_Na status of each dialysis machine in a color-coded manner on the central workstation.

12. The method as described in claim 1, characterized in that, The method is compatible with existing dialysis machine conductivity monitoring systems, including the Gambro / Baxter Diascan system, Fresenius OCM system, and other dialysis equipment equipped with dialysate inlet and outlet conductivity sensors. This compatibility is achieved by reading the signal output of existing conductivity sensors, eliminating the need for additional conductivity sensors. The method is applicable to hemodialysis (HD), hemodiafiltration (HDF), and online hemodiafiltration (online-HDF). For HDF and online-HDF modes, the convective sodium clearance in the sodium kinetic model also includes additional convective clearance caused by the replacement fluid.

13. A smart hemodialysis system based on sodium clearance ratio, characterized in that, include: The real-time plasma sodium concentration estimation module includes a dialysate inlet and outlet conductivity signal acquisition unit (or a signal interface with the conductivity sensor of an existing dialysis machine), a signal filtering unit (Kalman filter), and a Donnan correction calculation unit (built-in C_eq = [Na]d / σ, σ=0.97), which is used to back-calculate the plasma water and sodium concentration from the conductivity difference in real time and convert it into serum sodium concentration; The R_Na calculation module is used to calculate Na_target = (W_pre - W_dry) × [Na]s / 0.93 based on patient parameters, and to calculate Na_actual based on the cumulative value of sodium mass transfer, and then calculate R_Na = Na_actual / Na_target in real time. The dual-indicator display unit is used to simultaneously display the real-time values ​​and trend curves of R_Na and Kt / V, and to use color coding to indicate the compliance status; The dynamic control controller has a built-in dual-chamber sodium kinetic model to calculate the [Na]d control curve throughout the dialysis process and performs closed-loop adjustment based on real-time R_Na feedback to drive R_Na to approach 1.

0. The [Na]d regulation execution unit includes a sodium concentration sensor and a concentrate ratio adjustment mechanism, which is used to precisely adjust the sodium concentration of the dialysate according to controller instructions; The alarm module is used to generate a three-level alarm signal when R_Na deviates from the target range; The data communication module is used to transmit system data to external information systems in a standardized format; The feedback learning module stores the execution records and sodium removal results of each dialysis session, and automatically optimizes the model parameters.

14. The system as claimed in claim 13, characterized in that, The system, as an add-on module, can be integrated with existing dialysis machines via standard communication interfaces (RS-232, USB, Ethernet, or Bluetooth wireless) without requiring the replacement of the entire dialysis device. The standard communication interface supports compatibility with different brands of dialysis machines (Fresenius, B. Braun, Baxter / Gambro, Nipro, etc.) and supports connection with the AI ​​Agent intelligent dialysis prescription optimization system.

15. The system as described in claim 13, characterized in that, The system also includes a dialysis report generation module, used to automatically generate a standardized adequacy assessment report after each dialysis session, the report including: (a) Final values ​​of R_Na and Kt / V in this dialysis session and their criteria for meeting the standards; (b) Time-varying curves of estimated R_Na, Kt / V and serum sodium concentration during dialysis; (c) Dialysis fluid sodium concentration control curve (comparison of planned value and actual value). (d) Comparison of cumulative sodium removal amount with target sodium removal amount and details of diffusion / convection sub-items; (e) R-Na trend graph and sodium balance trajectory analysis of the most recent 10 dialysis sessions; (f) Prescription adjustment recommendations (if any).

Images