Dynamic optimization method for proportion of hemodialysis fluid

Abstract

The invention relates to a dynamic optimization method for the proportion of hemodialysis fluid, which comprises the following steps: forming a migration trend domain reflecting the active behavior trend of ions according to the migration directivity, strength and time variation trend of target ions in transmembrane exchange and in combination with the ion exchange stability at the initial dialysis stage, the buffer response hysteresis characteristic in the body of a patient and exchange inertia; the result is used as a core criterion for dynamic adjustment of the dialysate ratio; according to the physiological action inertia and membrane permeation rate of different ion components in the dialysis process, an adjustment lag response time window of each ion is established, and time-sensitive differential blending is carried out on each component; aiming at indirect cross interference caused in the ion component adjustment process, presetting an interaction mapping rule between ions, preferentially isolating a sensitive ion path, and then implementing adjustment; the frequency and the amplitude of the real-time ratio adjustment process of each ion component are dynamically monitored, and the adjustment threshold value and the adjustment frequency of the component are limited, so that blood microenvironment oscillation caused by excessive adjustment is inhibited.

Description

Technical Field

[0001] This invention relates to a method for optimizing the ratio of hemodialysis fluid, specifically a method for dynamically optimizing the ratio of hemodialysis fluid. Background Technology

[0002] Currently, a dialysate monitoring method and device based on a hemodialysis device, as disclosed in Chinese patent CN113289098A, relies on the core technology of inputting parameters such as conductivity, temperature, and transmembrane pressure into a dialysis risk assessment model, and then using patient vital signs data for incremental learning to achieve dynamic identification and early warning of dialysis risks. However, from the perspective of dynamic optimization of actual dialysate ratios, this approach has significant shortcomings and structural flaws, mainly in the following aspects: First, this technology is essentially a risk assessment path driven by static physical parameters. It constructs an initial risk model using transmembrane pressure changes and temperature parameters derived from conductivity and electrolyte concentration, and then uses patient vital signs data for model learning and updating. While this approach can improve the fitting of risk judgments, it does not achieve dynamic generation and real-time adjustment of the dialysate composition ratio itself. Its control logic is unidirectional monitoring-driven rather than closed-loop regulation, lacking the ability to fundamentally intervene in the ion ratio composition mechanism. Therefore, it remains essentially a passive risk warning system, unable to support proactive optimization of dialysate ratio formulations.

[0003] Secondly, this scheme lacks the ability to model the deep-level time-varying characteristics of ion behavior. The dialysate conductivity it relies on is merely a macroscopic representation of ion concentration, failing to distinguish the individual behavioral differences of specific ions during transmembrane migration, such as sodium, potassium, calcium, magnesium, and HCO3-. - Different physiological inertia, membrane permeability rates, and metabolic feedback pathways exist during dialysis, making conductivity parameters unable to reflect their microscopic dynamic responses or assess cross-interference effects during regulation. Furthermore, while the proposed scheme mentions using user vital sign information for incremental learning in its risk assessment model, most of these inputs are general indicators (such as blood pressure and heart rate), failing to clearly establish the mapping relationship between individual vital sign parameters and specific dialysate components. For example, when HCO3-... - When changes are slow, they do not trigger changes in the risk warning threshold in the model, but in actual clinical practice, HCO3... -Slow imbalances can easily lead to acid-base abnormalities and subsequent loss of electrolyte compensation control. Without the ability to identify specific time windows for slowly changing indicators, risk trends are easily misjudged. Furthermore, the models used in this technology mainly focus on the numerical calculation of risk values ​​and the assessment of early warning levels, lacking structural management of the dialysate regulation behavior itself. For example, it fails to identify whether the regulation behavior is a high-frequency fine-tuning or rhythmic perturbation behavior, whether it triggers cross-indicator oscillations, or whether it is necessary to construct sequence control boundaries between regulation pathways. The role of such key intervention strategies in clinical practice is to prevent multiple regulation behaviors from clustering into high-stress points on the time axis.

[0004] More importantly, the prior art still treats regulation and feedback as a linear, unidirectional causal process, ignoring the nonlinear propagation and regulatory lag phenomena in dialysis. For example, delayed feedback of electrolyte regulation often takes several minutes or even tens of minutes to manifest. If this physiological inertia is not modeled, it will easily lead to mistimed regulation or frequent over-regulation, resulting in physiological oscillations. Finally, from the perspective of system objectives, this invention is still mainly focused on risk detection and alarm, and its function is limited to reminders and data correction. It lacks closed-loop feedback control and path reconstruction capabilities and cannot cope with the path control problems caused by the high complexity of the regulatory behavior itself. Summary of the Invention

[0005] The purpose of this invention is to provide a dynamic optimization method for hemodialysis fluid ratio, thereby addressing some of the drawbacks and shortcomings pointed out in the background art.

[0006] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: a dynamic optimization method for hemodialysis fluid ratio, comprising: based on the migration directionality, intensity and time change trend of target ions in transmembrane exchange, combined with the ion exchange stability at the beginning of dialysis, the buffer response hysteresis characteristics in the patient's body and the exchange inertia between blood and dialysate, forming a migration trend domain that reflects the active behavior trend of ions, and using the migration trend domain as the core criterion for dynamic adjustment of dialysate ratio;

[0007] Based on the physiological inertia and membrane permeability of different ionic components during dialysis, and considering the migration trend domain, a regulatory hysteresis response time window for each ion is established. The timing, intensity, and duration of intervention of dialysate components are dynamically set to enable time-sensitive differentiated formulation of each component.

[0008] To address the indirect cross-interference caused during the adjustment of ion components, a pre-defined mapping rule for inter-ion interactions is established. When dynamically adjusting the dialysate ratio, sensitive ion pathways are preferentially isolated before adjustments are implemented through priority sorting, time-sequential separation, and the introduction of buffer components.

[0009] The frequency and amplitude of the real-time ratio adjustment process of each ionic component are dynamically monitored. When a cumulative fluctuation trend caused by high-frequency small-amplitude adjustments is detected, a rhythm gating restriction mechanism is triggered to limit the adjustment threshold and adjustment frequency of the components in order to suppress the blood microenvironment oscillation caused by excessive adjustment.

[0010] Furthermore, the migration trend domain establishes a contradiction judgment mechanism based on the consistency between the rate of change of ion concentration difference and the direction of the osmotic gradient of dialysate components, which is used to identify the deviation between surface migration and substantial metabolic traction. After the migration trend domain is formed, the ion behavior inertia back judgment logic is introduced. If an ion frequently reverses its direction in the historical trend, the current trend needs to be delayed for confirmation before it can be used for regulation.

[0011] Furthermore, a metabolic compliance score is introduced into the trend domain to measure whether the current migration behavior is consistent with the patient's basal metabolic pattern. If not, the activation of the regulation window is delayed. During the establishment of the response time window, if there is no significant in vivo feedback response to the regulation of an ion within a preset time, the regulation authority of that component will be temporarily frozen. The response time window will adaptively scale based on fluctuations in parameters reflecting metabolic activity, such as patient body temperature, pH, or red blood cell distribution width, to synchronize with the metabolic rate.

[0012] Furthermore, in the mapping rule for multi-ion interference, induced associations are dynamically identified based on the dialysis process. The second-order reaction triggered by the adjustment of one ion indirectly pulls on another ion, and non-dominant linkages are identified first. For ion combinations with high interference levels, an interlocking regulation authorization mechanism is adopted, requiring that the minimum response feedback time between the adjustments of two ions must be met before they can be rotated for regulation.

[0013] Furthermore, the rhythm gating mechanism is preferentially applicable to regulatory behaviors with fluctuation periods shorter than the average period of a patient's single heart rate; after the regulation frequency self-suppression mechanism is triggered, the regulation is not blocked, but enters a delayed feedback observation period to determine whether it is a false fluctuation caused by physiological oscillation interference; and the rhythm gating threshold is dynamically set by the system's cumulative regulation energy density, which is a composite load index composed of three factors: regulation amplitude, regulation interval, and regulation duration.

[0014] Furthermore, the regulatory behavior with a fluctuation period shorter than the average heart rate period must satisfy the requirement that the fluctuation spectrum energy is concentrated in the heart rate fundamental frequency and high-frequency harmonic range, so as to determine the interference with the heart rate rhythm and to preferentially activate the rhythm gating mechanism; before the rhythm gating mechanism is activated, it is necessary to verify whether there is a time distribution overlap in the regulatory behavior of the target ion and the regulatory path of other high-risk components, and adjust the gating delay or advance window.

[0015] To address high-frequency fine-tuning behaviors that may interfere with patients' spontaneous heart rhythm, a custom spectral coupling integral function is introduced to identify and intervene in potential overlapping regulatory paths, dynamically activate gating mechanisms, and adjust their activation time windows. This method defines and calculates the energy density focusing of regulatory behaviors in the frequency domain and combines the synchronous behavior of regulatory behaviors with other high-risk components in the time domain to achieve adaptive optimization of the rationality of gating trigger determination and activation timing.

[0016] The core technology includes two aspects:

[0017] ①Spectral energy focusing judgment:

[0018] Define the spectral interference exponential function Λ(t) of regulatory behavior to quantify the potential interference risk of regulatory behavior on heart rate rhythm, as follows:

[0019]

[0020] in:

[0021] Λ(t) is the spectral interference exponent of the regulatory behavior, serving as the trigger threshold parameter for rhythm gating; t0, t1 are the start and end points of the calculation time window, and t is the current system processing time; ω is the frequency variable; ψ(ω,t): the instantaneous amplitude weighting function of the regulatory behavior spectral function at frequency ω and time t; E(ω) is the spectral energy function of the regulatory behavior, representing the energy density of the regulation at frequency ω; Ω(ω) is the frequency window function of the reference heart rate rhythm spectrum, used to represent the energy distribution curve of the baseline heart rate band; κ(t) is the gating exponent function, a dynamic time-varying parameter, representing the sensitivity amplification coefficient of the current system to high-frequency interference, whose value can be updated in real time by the system's metabolic stress level;

[0022] When Λ(t)>θ Λ If the threshold is reached, the system determines that the regulatory behavior is significantly coupled with the heart rhythm in the frequency domain and has potential interference. It immediately marks the behavior as a candidate event for rhythm impact and enters the rhythm gating pre-triggered state.

[0023] ② Adjusting path overlap judgment and window adjustment mechanism:

[0024] Before the rhythm gating mechanism is officially triggered, the system needs to verify whether there is a calculable overlap between the current regulatory behavior and the regulatory paths of other highly sensitive components on the time axis; if the overlap is higher than the preset correlation threshold, the behavior is considered to be a multi-point simultaneous intervention scenario, triggering coordinated oscillation;

[0025] To address this, the system introduces an adjustment path crossover coefficient Ξ. ij :

[0026]

[0027] in:

[0028] τ i ,τ j τ represents the duration of the adjustment window for two regulating components (e.g., sodium and potassium). i ∩τ j Indicates the overlapping interval of two moderating events on the time axis; Ξ ij ∈[0,1]Ξ ij >ξ th (Cross threshold), the system automatically delays the rhythm gating window to prevent false triggering; conversely, if Ξ ij <ξ min And Λ(t) >> θ Λ If so, the gating start window will be activated in advance, and the system will enter the fast gating response state.

[0029] Furthermore, the delayed feedback observation period monitors the regulation results and its cross-index transduction effects, including osmotic pressure fluctuations, vascular tone parameters, and metabolic reflex delays, to comprehensively determine whether the regulation is triggered by a false signal. If synchronous fluctuations of indicators that have no direct physiological link with the regulation target occur during the observation period, the system classifies this as a systemic oscillation and adds a verification delay to that segment of the regulation path.

[0030] Furthermore, the delayed feedback observation period is divided into segmented response windows according to different physiological feedback types, which are used to quickly measure response indicators including blood pressure and those including HCO3. - The slow-changing indicators are detected synchronously with a delay; when the regulatory behavior is determined to be a false fluctuation and the gating restriction is about to be released, a secondary safety confirmation mechanism must be entered first, which includes trend stability backtesting and post-wave recovery analysis of patient vital signs.

[0031] Furthermore, the factor of the regulation amplitude and the frequency factor are combined in the assessment to introduce the time and amplitude coupling tension curve, which is used to assess whether the regulation path constitutes a high-stress regulation zone within a certain time window; before judging that the regulation energy density threshold exceeds the limit, a predictive resistance analysis window is run first to assess whether the regulation response zone that the patient is about to enter has physiological buffering potential. If not, the gating is triggered in advance.

[0032] Furthermore, if the energy density trend continues to rise rapidly after the gating is triggered, the regulation path reconstruction mechanism is executed to re-plan the regulation timing, frequency, and amplitude distribution of each ion component; the rhythm gating mechanism generates a regulation behavior entropy value after the regulation path reconstruction, which is used to evaluate the stability of the current ratio mode and the complexity of the regulation behavior, so as to determine whether further dimensionality reduction processing is needed.

[0033] The beneficial effects of this invention are as follows: By introducing mechanisms such as migration trend domains, metabolic compliance scores, and response lag windows, it can dynamically identify the ion metabolism behavior of each patient at different stages, avoiding the use of fixed ratios or single target value adjustment modes, and truly realizing an individualized dialysate ratio strategy based on dynamic feedback of physiological processes. By introducing spectral interference identification, rhythm gating mechanisms, and regulatory path overlap analysis, it can effectively identify the synergistic oscillation effects of high-frequency fine-tuning behavior on the patient's heart rhythm, osmolarity, or cross-ion-induced oscillations, thereby achieving systemic oscillation suppression and rhythmic protective regulation. By setting a delayed feedback observation period and a secondary safety confirmation mechanism, combined with cross-link response analysis of multi-channel physiological parameters, it can intelligently identify false signals caused by surface migration, charge drift, and initial dialysis inertia, avoiding premature or excessive regulation due to misjudgment of trends, and improving the reliability of judgment.

[0034] By introducing time-amplitude coupling tension curves and regulatory behavior entropy analysis, the system can identify the stress level and complexity of regulatory behavior within a specific time window and perform behavioral dimensionality reduction, effectively controlling the complexity of the regulatory path and improving system stability and resource allocation efficiency. The reconstruction mechanism allows the regulatory path to be dynamically adjusted based on regulatory feedback, buffering capacity, and physiological state, enabling the system to possess self-optimization, self-convergence, and complexity self-inhibition capabilities, adapting to the complex physiological intervention needs under different stages and dialysis conditions. Utilizing ion mapping rules and interlocking authorization mechanisms, a safe boundary framework for coordinating the timing and influencing pathways of multi-component regulation is provided, making multi-ion synergistic regulation no longer dependent on empirical judgment, thus improving the reliability and safety of coordinated regulation. Attached Figure Description

[0035] Figure 1 This invention provides a dynamic optimization process for the preparation of hemodialysis fluid ratios.

[0036] Figure 2 This is a diagram showing the multi-level control function relationship of dynamic ratio of hemodialysis fluid in this invention.

[0037] Figure 3 This is the high-frequency micro-adjustment law gating process of the present invention.

[0038] Figure 4 This is a flowchart of the intelligent dynamic ratio control embodiment of hemodialysis fluid in Embodiment 1 of the present invention.

[0039] Figure 5 This is the flowchart of Embodiment 2 of the present invention, which describes a multi-ion high-frequency regulation intelligent gating system. Detailed Implementation

[0040] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0041] Combined with appendix Figure 1The present invention provides a dynamic optimization method for the ratio of hemodialysis fluid. It constructs and applies a migration trend domain as the core criterion for the dynamic adjustment of the ratio of dialysis fluid, so as to achieve high-precision, individualized, and real-time control of the ion environment in the patient's body. First, based on real-time monitoring data during dialysis, the migration directionality, migration intensity, and temporal trend of target ions (such as potassium, sodium, calcium, and bicarbonate ions) during transmembrane transport are captured and analyzed. Directionality is used to identify the current net migration state of the ions (e.g., whether they are moving from blood to dialysate or vice versa), intensity is used to measure the rate of concentration change per unit time, and temporal trend is used to identify whether the behavior is in the rising, stable, or reversing phase. Based on this, ion exchange stability is collected during the initial stage of dialysis (generally 5 to 15 minutes after start). Stability is defined as the coefficient of variation of the target ion concentration change per unit time. If this coefficient tends to stabilize, the system is considered to be in the initial steady-state stage, and the ion change trend during this stage can be used as one of the trend domain initialization parameters. Simultaneously, the system records the response hysteresis characteristics of the patient's internal buffer systems (including but not limited to bicarbonate buffer pairs, protein buffer systems, and phosphate buffer systems). This characteristic is obtained by monitoring blood pH and HCO3-. - The response time differences of concentration and metabolism-related electrolytes are quantified and a response hysteresis matrix is ​​formed. Furthermore, the system acquires the exchange inertia between blood and dialysate in the current dialysis device. This exchange inertia includes not only membrane permeability, fluid flow rate ratio, and transmembrane pressure changes, but also patient-specific transmembrane exchange behavior patterns reflected in historical data. This information is used to identify the delay and inertia amplitude of ion exchange under the current device settings. These multiple factors together constitute a migration trend domain, which is a dynamically updated multivariate behavioral field capable of depicting the active migration intention of target ions in real time. The system identifies potential reversal risks. During system operation, when the trend domain shows that a certain ion is continuously migrating positively and its intensity is higher than the predicted baseline and exceeds the system's preset intervention threshold, the dialysate concentration of that component is reduced. Conversely, if it shows negative migration or a significant buffer hysteresis response, the system postpones the adjustment window or implements a slow increase strategy to prevent electrolyte over-adjustment or reverse adjustment caused by adjustment lag. Ultimately, the entire dialysate ratio adjustment behavior uses a trend domain-driven, stability confirmation-hysteresis offsetting-inertial feedback closed-loop execution logic to achieve dynamic dialysate ratio optimization based on the actual ion behavior trend.

[0042] To address the differences in physiological inertia and membrane permeability rates among various ionic components during dialysis, and combining real-time behavioral criteria from the migration trend domain, a regulatory hysteresis response time window is established to achieve time-sensitive differentiated formulation of each dialysate component. The system first extracts the fundamental characteristics of ion migration behavior based on the migration direction, rate, and trend stability information of each target ion (e.g., potassium, sodium, calcium, magnesium, bicarbonate, etc.) in the previously constructed migration trend domain. This characteristic is then combined with a physiological response inertia factor to form a preliminary framework for assessing the feasibility of regulatory intervention. Physiological inertia refers to the lag in the effect of a particular ion on changes in physiological indicators (e.g., blood pressure, heart rate, acid-base balance, electrophysiological activity, etc.) in vivo. This inertia factor can be calculated from historical dialysis data or the input-output response delay monitored during the first dialysis session. Simultaneously, the membrane permeability rate describes the speed characteristics of the ion crossing the dialysis membrane, and its value depends on parameters such as membrane pore size, charge state, blood-to-dialysate flow rate ratio, and ion electrochemical gradient. Based on this, the system establishes a regulatory hysteresis response time window for each target ion. The definition is: the shortest predicted time period from the start of adjusting the dialysate component concentration to the target ion reaching a steady-state feedback change in the patient's blood, and this window is dynamically updated to adapt to different dialysis stages and individual status changes. The system then sets the corresponding ion adjustment trigger conditions and execution rhythm based on this time window, meaning only one effective adjustment action is allowed within this time window, and the next adjustment must wait until the physiological response triggered by the previous intervention is completed or enters a stable period before being triggered, to avoid ion oscillations or buffer overshoot caused by frequent fine-tuning. Furthermore, combining the trend strength (e.g., slope coefficient) and fluctuation amplitude (e.g., peak difference) of the ion in the migration trend domain, the system calculates the corresponding adjustment intensity. The intensity setting adopts a graded logic: for ions with clear trends and low inertia (e.g., sodium ions), medium to high adjustment intensity is allowed; while for ions with high inertia and delayed feedback (e.g., HCO3-),... - The system employs a fine-tuning strategy and sets a minimum step size to prevent hysteresis imbalance. The adjustment intensity is also linked to the intervention duration setting. The system dynamically sets the execution length of the current adjustment based on the window duration. If the trend is strong and the response feedback is stable, the intervention duration is extended; otherwise, a short, pulsed intervention is performed. Throughout the process, the system uses the migration trend domain as the main control input and the adjustment hysteresis response window as the constraint boundary. It links the three dimensions of adjustment timing, adjustment amplitude, and intervention time to construct a multi-component dynamic ratio adjustment strategy that takes into account real-time performance, hysteresis, and individual differences. This ensures that the intervention of each ion component during dialysis is more precise, its effects are more controllable, and the physiological feedback is more stable, thereby significantly improving the safety and physiological matching of dialysate regulation.

[0043] To address the indirect cross-interference problem caused by the regulation of multiple ion components during dialysis, an inter-ion interaction mapping rule based on system intervention link analysis is constructed, and an interference decoupling mechanism is designed to achieve a fine-grained synergistic regulation strategy for multiple ions. The system first establishes a set of ion interaction mapping rules based on basic physiological electrolyte knowledge, biochemical reaction chains, and previous dialysis data during the initialization phase. These rules describe whether two or more ion components exhibit direct charge coupling, synergistic changes in osmotic pressure, membrane potential linkage effects, or indirect metabolic cross-influences during regulation. For example, sodium and potassium ion regulation may cause a linkage in extracellular fluid osmotic pressure changes, thereby affecting potassium transmembrane migration through the water-sodium exchange mechanism. Similarly, bicarbonate regulation may affect blood pH, indirectly altering the bound state and free concentration of calcium, forming implicit cross-interference. These mapping relationships are quantified as interference level indices and labeled as highly sensitive coupling pairs, neutral weak coupling pairs, or safe independent groups for subsequent decision-making. During the dynamic ratio regulation execution phase, the system first uses the mapping rule base to screen ion combinations with potential cross-interference risks based on the migration status and regulatory needs of each ion in the current trend domain. It then initiates a priority ranking algorithm, prioritizing the regulation of components with the lowest coupling strength, shortest interference paths, or smallest feedback delays. Simultaneously, components with highly sensitive coupling relationships are temporarily frozen or their regulation is delayed, ensuring that regulatory actions occur simultaneously. The system minimizes overlap within the window; subsequently, it enters a time-sequential separation strategy scheduling process, dynamically planning the adjustment rhythm according to the ion regulation response window and feedback completion time. For example, it first adjusts potassium, then waits for its effect to converge before adjusting sodium, ensuring that cross-perturbations do not superimpose in the state before feedback is completed. In some highly coupled scenarios, the system can also call preset neutral buffer components, such as isotonic glucose, mannitol, or small doses of lactate, which are neutral solutes that do not participate in the main electrolyte channels, to play a role in osmotic pressure isolation, potential buffering, or indirect pH stabilization, preventing the two types of coupled ions from forming short-term synchronous driving channels on both sides of the membrane, thereby softly isolating sensitive pathways. The entire cross-interference identification and isolation adjustment behavior operates modularly at the system level in the order of interference decoupling → path optimization → feedback separation → buffer assistance, achieving sequential independence in regulation logic and minimal overlap in physiological feedback. This effectively improves the system stability, response predictability, and intervention safety in the process of multi-ion ratio regulation, ensuring that dynamic ratio regulation not only responds to trend changes but also adheres to the cross-control principle of ion regulation, realizing the dialysis system's ability to perceive interference and optimize pathways in complex in vivo ion networks.

[0044] The frequency and amplitude of the real-time ratio adjustment process of each ion component are dynamically monitored, and a rhythm gating restriction mechanism is triggered based on the blood microenvironment response trend caused by the adjustment behavior to suppress the systemic micro-oscillations caused by the accumulation of high-frequency small-amplitude adjustments. The system establishes a real-time recording unit for the regulatory behavior of each target ion component. This unit continuously records the timestamp, concentration change amplitude, response delay time, and in vivo feedback indicators for each regulatory operation based on a sliding time window. The system calculates the number of regulation times per unit time (i.e., regulation frequency) and the absolute value of each concentration change (i.e., regulation amplitude), and further constructs a regulation frequency amplitude integral trend function to describe whether the current regulatory behavior exhibits a high-frequency and small-amplitude repetitive adjustment pattern. Simultaneously, the system introduces a cumulative fluctuation judgment model, comprehensively considering the amplitude amplification effect and non-physiological rhythm synchronization of key indicators such as short-term ion concentration oscillation trends, slight fluctuations in blood pH, minor changes in osmotic pressure, and extracellular fluid tension. If the regulatory behavior does not cause substantial improvement in the target concentration in vivo within a unit time, but instead triggers rhythmic perturbations in systemic indicators (such as frequent but slight arrhythmias, microvascular contraction responses, and transient instability of transmembrane pressure), the system determines that the current regulation... The path has entered a high-frequency, low-efficiency perturbation state and immediately triggers a rhythm gating restriction mechanism. This mechanism does not directly block the regulatory authority of the target ion component, but dynamically increases the regulatory trigger threshold of the component (i.e., it can only be allowed to be regulated again when the migration trend exceeds a higher amplitude or a higher signal intensity) and sets a minimum regulation interval (e.g., not earlier than 5 minutes after the previous regulation). At the same time, the maximum single regulation amplitude is set to a low value to limit the impact intensity. In addition, the system also introduces a regulatory behavior rhythm energy consumption weight index as the basis for dynamic regulation authorization. This index is calculated by comprehensively considering the number of regulation times, regulation amplitude, and in vivo feedback fluctuation amplitude. When this weight exceeds the physiological tolerance range set by the system for several consecutive cycles, it is determined that the system microenvironment has shown a critical oscillation signal. The system temporarily marks the component as a low dynamic activity state, and its ratio will remain at the current value without adjustment until in vivo feedback indicators such as pH, blood pressure, or transmembrane pressure return to within the baseline fluctuation range, and then the regulation channel is reopened.

[0045] Combined with appendix Figure 2To construct the migration trend domain, a contradiction judgment mechanism is introduced, comparing the rate of change of ion concentration difference with the direction of the dialysate component osmotic gradient. After the trend domain is generated, an ion behavior inertia back-judgment logic is superimposed to improve the accuracy of trend judgment and the stability of the regulatory response. The system first acquires the rate of change of concentration difference of each ion component between blood and dialysate in real time within each periodic adjustment evaluation window. This rate of change reflects the surface migration intensity of a certain ion during the transmembrane process. However, this data itself cannot be directly used for regulatory decisions because some ions may experience rapid changes in concentration difference due to dialysate concentration adjustments, but this does not necessarily indicate that actual metabolic traction has occurred in vivo. Therefore, a contradiction judgment mechanism is designed to dynamically compare the above-mentioned rate of change of concentration difference with the direction of the dialysate component osmotic gradient within the corresponding time period. The osmotic gradient direction is calculated based on a comprehensive analysis of transmembrane pressure, molecular charge state, and solute permeability index. When the system detects an inconsistency between the direction of change of concentration difference of a certain ion and the direction of the dialysate osmotic gradient (i.e., for example, the blood sodium ion concentration increases while the osmotic gradient remains from the dialysate to the blood), it determines that the migration is surface migration rather than substantial metabolic traction. The directional indicators of ions in the trend domain will not be recorded as valid trend signals to avoid erroneous regulation induced by exogenous transient adjustments. Based on this, after the trend domain is formed, the system further introduces the logic of ion behavior inertia back judgment, that is, to statistically analyze the trend direction of each target ion in multiple historical regulation cycles. If a certain ion has frequent reversals in direction in multiple adjacent cycles (such as first migrating in the positive direction and then in the negative direction, and the number of reversals exceeds the set frequency threshold), the system considers that the ion is currently in an unstable state, and its trend direction lacks coherence and predictive value. Therefore, even if the migration trend calculated in the current cycle is significant, it will be marked as a delayed confirmation trend. This trend will not be used immediately to trigger the adjustment of dialysate concentration, but will enter the system observation period to wait for the trend continuation in the next cycle to be confirmed. If the trend direction is consistent in two consecutive cycles and there is no contradiction in the osmotic gradient judgment, then regulation can be triggered.

[0046] A metabolic compliance scoring mechanism is introduced within the migration trend domain, and metabolic feedback verification, permission freezing, and window adaptive scaling strategies are introduced in the process of establishing and managing the regulatory response time window to achieve dynamic matching of individual patient metabolic status and precise coordination of dialysate regulation rhythm. Based on the constructed migration trend domain, the system further overlays a metabolic compliance scoring module. This scoring is used to measure whether the migration trend of a certain ion component is consistent with the patient's basal metabolic pattern. The scoring mechanism is based on the system's modeling of multiple parameters, including the patient's resting metabolic rate, metabolic rhythm stability, electrolyte adjustment speed before and after meals, body temperature stability, and respiratory and metabolic acid-base compensation capabilities, to obtain a metabolic reference trajectory. The migration trend in the current cycle will be mapped to this trajectory for matching score. If the metabolic compliance score is lower than a preset threshold, the system will consider that the trend may be induced by occasional interference, measurement noise, or in vitro regulation. Therefore, the regulation behavior of this component will be delayed and enter the evaluation and observation state until the trend continues in subsequent cycles and the metabolic score returns to normal before regulation can be carried out. When establishing the regulatory response time window, the system configures a minimum effective response time for each ion component based on historical regulation data and real-time feedback models. That is, the time period from the occurrence of the regulation action to the appearance of a detectable stability change of the component in the blood. If the target index does not show significant changes within the preset time (e.g., potassium ion concentration does not change by more than 0.5%), the system will not be able to regulate the ion component. If the concentration of a component reaches 2 mmol / L or the osmolality remains unchanged, the system determines that the current regulatory behavior has failed and automatically triggers the permission freeze mechanism. The regulatory permission for this component will be temporarily suspended, disallowing continuous adjustment to prevent misjudgment, misadjustment, or system disturbance caused by regulatory lag. The frozen state will remain until the system detects that the target feedback variable has recovered its trend in subsequent cycles or that other parameters have returned to normal, at which point the freeze will automatically be lifted. To improve the system's adaptability, this regulatory response time window has dynamic scaling capabilities. The system monitors patient body temperature, blood pH, and red blood cell distribution width (RDW) to reflect metabolic activity and oxygen delivery. The physiological indicators of efficiency adjust the duration of the response time window in real time. Specifically, when the patient's body temperature rises, the pH becomes more alkaline, or the RDW increases (indicating faster red blood cell turnover or metabolism), the system determines that the current metabolic rate is increased, and the window can be shortened to accelerate the regulatory response cycle. Conversely, when the body temperature drops, the pH becomes more acidic, or the RDW decreases, the system considers metabolism to be slower, and the response window is automatically extended to avoid misadjustment caused by the regulation trying to catch up with the slow metabolic rhythm. This ensures that the dialysate mixing behavior can be synchronously matched with the patient's actual metabolic state in terms of adjustment timing, duration, and magnitude.

[0047] In the process of multi-ion regulation, this system dynamically identifies induced correlations and insubstantive linkages, and introduces an interlocking regulatory authorization mechanism to manage the alternating regulatory behavior of high-interference-risk ion combinations, aiming to improve the system stability and response rationality of complex electrolyte regulation processes. During dialysis, the system calls upon an ion interference mapping rule base in real time. This rule base includes not only direct concentration regulation cross-influences but also indirect regulatory traction relationships formed through physiological response pathways, i.e., induced correlation mechanisms. Specifically, although one ion component may not directly affect the concentration of another component after regulation, it indirectly drives changes in the migration direction or active state of other ions by inducing second-order reactions such as changes in blood pH, osmotic pressure, or membrane potential. For example, pH changes induced by bicarbonate regulation may indirectly affect the free calcium ratio or the extracellular distribution of potassium. Therefore, the system needs to dynamically identify such insubstantive linkage behaviors. To this end, the system constructs an induced correlation mechanism by monitoring the delayed linkage response of multiple indicators after regulation. The system employs a causal link diagram and assigns interference levels based on response amplitude and time lag. When a high induced coupling degree is detected between an ion and another component, the system marks this combination as a high-interference ion pair and activates an interlocking regulation authorization mechanism during ratio adjustments. This mechanism requires that after either ion is authorized for regulation, it must complete its minimum response feedback time window (i.e., the main feedback variable triggered by the regulation stabilizes or enters a plateau period) before the other ion can obtain regulation authorization. This mechanism effectively prevents the two coupled ion components from being repeatedly regulated in the stage before the response is completed, thereby preventing problems such as response aliasing, buffer saturation, or charge imbalance. The interlocking mechanism also sets mutual exclusion logic priorities. If a component is a metabolically dominant ion (such as sodium or HCO3), it will be subject to priority. - Its regulatory priority will be higher than that of relatively dependent ions (such as potassium or calcium), and it allows interruption of the authorization window of secondary channels. In addition, the system is equipped with an emergency release strategy. When a clinical determination is made that a certain component needs to be regulated immediately but the interlocking mechanism restriction has not been released, the system can implement emergency jump authorization under the trigger of multidimensional index consensus, provided that it is predicted that the regulation will not trigger the risk of higher-order linkage.

[0048] When regulatory behaviors are too frequent or too intense, a rhythmic gating mechanism is used to control the regulatory rhythm to prevent oscillations in the blood microenvironment. This mechanism includes three key functions: fluctuation cycle identification, delayed feedback observation, and dynamic threshold management based on regulatory energy density. The system monitors and analyzes the regulatory behavior of each ion component in real time. First, it introduces a precondition for rhythm gating: whether the fluctuation period of the current regulatory behavior per unit time is less than the average period of the patient's single heart rate. The heart rate period is dynamically estimated using synchronously acquired ECG or pulse signals. If the regulatory behavior of a certain component occurs twice or more consecutively within a time interval shorter than this period, the system determines it to be a high-frequency, high-disturbance behavior and enters a rhythm gating pre-judgment state. In the pre-judgment state, if the regulatory behavior triggers a frequency self-suppression mechanism, exhibiting phenomena such as excessive number of regulation times per unit time, excessively short regulation intervals, or incomplete feedback responses, the system does not immediately block the ion regulation function. Instead, it enters a delayed feedback observation period, a dynamically set evaluation window. During this period, the system does not actively execute further regulation but continuously observes the actual response changes of previous regulation to blood-related indicators (such as ion concentration, osmotic pressure, pH, blood pressure, etc.). If the response indicators return to normal fluctuations or show a stable trend within the observation period, the system determines that the original fluctuation was a false fluctuation or physiological oscillation interference, avoiding false triggering of rhythm gating. If the system confirms the regulation after the observation period... If the rhythmic behavior is indeed induced by interference, the rhythm gating mechanism is formally activated. The threshold setting in the rhythm gating mechanism adopts a dynamic algorithm based on regulatory energy density. The system quantifies the regulatory behavior of each component in the most recent several regulatory cycles and constructs a regulatory energy density function. This function comprehensively considers three main factors: first, the regulatory amplitude, i.e., the absolute value of the change in concentration adjustment each time; second, the regulatory interval, i.e., the time difference between two adjacent regulatory operations, reflecting the system recovery time; and third, the regulatory persistence, i.e., the duration or repetitiveness of continuous regulatory behavior. The three factors together constitute a composite load index, representing the total regulatory pressure currently borne by the system for that ion component. When this index exceeds the upper limit threshold set by the system according to the patient's individual tolerance to fluctuations, the rhythm gating mechanism is formally activated, limiting the maximum regulatory amplitude, minimum regulatory interval, and regulatory trigger sensitivity of that component within a certain period of time in the future. This prevents microenvironmental oscillation problems such as increased electrolyte fluctuations, unstable blood pressure, or imbalance of intracellular and extracellular fluid transport caused by excessive, rapid, or excessive regulation. Thus, a ratio optimization method can be achieved that maintains the stability of the regulatory rhythm and the physiological safety boundary even in a highly dynamic dialysis control environment.

[0049] Combined with appendix Figure 3This system is suitable for high-frequency fine-tuning scenarios where regulatory behaviors may interfere with the patient's spontaneous heart rhythm. By identifying the coupling relationship between regulatory behaviors and heart rate rhythm in the frequency domain, and combining the degree of temporal overlap of multiple ion regulatory pathways, it achieves intelligent triggering and adaptive optimization of the rhythm gating mechanism's initiation timing. The system first targets key regulatory ion components (such as sodium, potassium, calcium, and HCO3-) in all dialysate. - The system monitors the adjustment behavior of (etc.) in real time and analyzes its energy accumulation in the frequency domain to determine whether it constitutes potential interference with the heart rate rhythm. The system uses the fluctuation period of the adjustment behavior being less than the patient's current average heart rate period as a preliminary screening condition for interference identification. The heart rate period can be dynamically estimated from the RR interval of the pulse wave or ECG signal. Based on this, the system further calculates the interference exponential function of the adjustment behavior in the spectral space to measure its frequency domain interference potential. This function is defined as follows:

[0050]

[0051] Where Λ(t) is the spectral interference exponent of the regulatory behavior, used for triggering the rhythm gating mechanism; t0 and t1 are the start and end points of the evaluation time window, and t is the current time; ω is the frequency variable (unit: Hz); ψ(ω,t) is the instantaneous weighting function of the regulatory behavior at frequency ω and time t, used to emphasize the contribution of a specific frequency band; E(ω) represents the energy density of the regulatory behavior at frequency ω; Ω(ω) is the reference heart rate spectrum energy curve, describing the energy distribution of the patient's baseline heart rhythm in the frequency domain, used as an interference matching reference; κ(t) is the gating exponent function, representing the system's sensitivity to high-frequency interference at the current time point, and its value is dynamically updated according to the patient's metabolic stress level. When Λ(t) > θ Λ When the interference threshold is reached, the system determines that the regulatory behavior is significantly coupled with the patient's heart rhythm, posing a risk of rhythm interference. It immediately marks this behavior as a candidate event for rhythm impact, and the system then enters a rhythm gating pre-trigger state. To avoid false inhibition due to misjudgment of high-frequency regulation, before formally activating the rhythm gating mechanism, the system further verifies whether the regulatory behavior of the target ion overlaps with the regulatory pathways of other highly sensitive components (such as potassium-sodium coupling, calcium-magnesium interaction, etc.), i.e., whether there is a potential risk of coordinated oscillation. For this purpose, the system defines a regulatory pathway crossover coefficient Ξ. ij :

[0052]

[0053] Where, τ i and τ j τ represents the effective window duration of the two regulating components (e.g., the time period from the start of regulation to the stabilization of the target feedback variable). i ∩τ jThe crossover coefficient Ξ represents the overlapping interval of these two windows on the time axis. ij ∈[0,1] indicates the degree of time synchronization between the two. When Ξ ij >ξ th When the cross threshold is reached, the system assumes significant overlap between regulatory behaviors, potentially leading to superposition of disturbances. Therefore, it automatically delays the rhythm gating initiation time, waiting for feedback from key components before entering the restricted state; if Ξ ij <ξ min And the current interference index Λ(t) >> θ Λ If the system determines that the disturbance has reached a steady-state trend, but there is no interference conflict between the regulatory behaviors, it will activate the rhythm gating mechanism in advance and enter the fast gating response mode. Through the above frequency domain-time domain dual logic criteria, the system can not only identify the rhythm risk of individual regulatory behaviors, but also perceive their collaborative linkage characteristics in the multi-component regulation network, ultimately achieving precise triggering and adaptive window control of the rhythm gating mechanism.

[0054] A delayed feedback observation period is introduced during the regulation process. During this period, the in vivo response effect of the regulatory behavior itself and its cross-index transduction effect are monitored simultaneously to determine whether the current regulation is a truly effective response or a false signal triggered by systemic disturbances or physiological oscillations. After the self-inhibition mechanism of the regulation frequency is triggered, the system does not immediately block the regulatory authority of the ion component, but enters a delayed feedback observation period. This observation period is a window with a dynamic start and end range. During this period, the system continuously collects the direct feedback variables triggered by the regulatory behavior of the ion component in vivo, such as changes in ion concentration, pH changes, and adjustments in bicarbonate levels. At the same time, the monitoring range is expanded to a set of indirect cross-index variables, including but not limited to the rate of change of blood osmolality, vascular tension-related parameters (such as pulse pressure and microcirculation perfusion index), extracellular fluid distribution, and reflexive delayed indicators closely related to electrolyte metabolism (such as changes in urine volume, slight changes in respiratory rate, or acid-base compensation behavior). Based on the synchronous fluctuations, lag consistency, and time-domain response correlation of these indicators, the system establishes a regulatory feedback consistency model to comprehensively determine whether the physiological effect produced by the current regulation can be attributed to the regulation itself. If the target feedback variable shows no significant change during the observation period, but other unrelated indicators (such as red blood cell distribution width RDW, lactate level, central venous pressure, etc.) exhibit synchronous fluctuations, and these fluctuations have no direct physiological link with the target regulatory ion, the system classifies this situation as a cross-indicator non-causal response. That is, the regulatory behavior did not lead to the expected result, but the overall system fluctuations intensified. It is inferred that this behavior may be a systemic oscillation caused by hemodynamic instability, fluid transfer stress, or environmental disturbances, and does not represent the true effectiveness of regulation. Therefore, this regulatory pathway will be marked as a potential false positive pathway, and the system will add a verification delay operation to this pathway. That is, before the next permitted regulation, the trend domain direction, metabolic compliance score, and feedback consistency model score must all meet the stability criteria before re-authorization of regulation can be granted.

[0055] During the regulation process, a delayed feedback observation period is set, and a segmented response window is designed for this period based on physiological indicators. Simultaneously, before determining that the regulatory behavior is a false fluctuation and intending to release the rhythm gating constraint, a secondary safety confirmation mechanism, including trend stability backtesting and post-wave recovery analysis, is implemented to ensure the stability of the regulatory path and the integrity of the system's physiological rhythm. When the regulatory behavior of a certain ion component triggers the rhythm gating mechanism constraint due to high-frequency, small-amplitude adjustments, the system first enters the delayed feedback observation period. However, this observation period is not a static time period but is segmented according to the response characteristics of different physiological feedback types. Specifically, a short response window (usually 1 to 3 minutes) is set for rapidly changing indicators such as blood pressure, pulse pressure, and immediate pH response, while a short response window is set for slowly changing indicators such as bicarbonate (HCO3-). -The system sets delayed synchronous detection windows (up to 15 to 30 minutes) for parameters such as concentration, extracellular fluid volume, and electrolyte metabolic buffer curves. Within each window, the system monitors the feedback trajectory of the target regulatory behavior and the associated perturbation trend of the non-target system. If none of the feedback indicators show a change path consistent with the regulatory behavior logic within the corresponding time window, and the trend domain assessment shows that the current behavior lacks a continuous directional drive, the system initially determines that the regulatory behavior may be a false fluctuation signal. However, to prevent premature release of the truly effective regulation due to misjudgment, the system will not immediately degovern, but will enter a secondary safety confirmation mechanism. This mechanism consists of two sub-modules: one is trend stability backtesting, that is, the system re-analyzes the past performance of the ion component. The system employs two main methods: First, it assesses the continuity and consistency of the migration trend within the regulatory cycle. If the direction alternates repeatedly or the trend intensity fluctuates excessively, it is considered an unstable trend. Second, it performs post-gating recovery analysis. During gating, the system continuously monitors the fluctuations of multiple vital signs (such as heart rate, blood pressure, transmembrane pressure, body temperature, and respiratory rate). If these parameters quickly return to baseline or enter a narrow plateau after regulatory blocking, it indicates that the previous regulatory behavior had a potential disturbance to the system, and gating has played a stabilizing role. In this case, the system will continue to maintain the blocked state. Conversely, if the vital signs do not recover significantly after gating is initiated, or if no abnormal fluctuations occurred, it can be presumed that the regulatory behavior was a misjudgment, and the system will release the gating, restoring the regulatory channel for that ion component. This segmented response window combined with a two-level safety confirmation mechanism improves the judgment accuracy and fault tolerance of the rhythm gating mechanism during execution, ensuring that the system can still achieve a stable, efficient, and rhythmically coordinated dialysate ratio control strategy even in the face of complex dynamic feedback and high-noise physiological signals.

[0056] In the assessment of the combination of regulation frequency and amplitude, a time-amplitude coupling tension curve is introduced. When the regulation energy density approaches its limit, a predictive resistance analysis window is run in advance to determine whether the patient possesses sufficient physiological buffering potential, thereby dynamically deciding whether to pre-activate the rhythm gating mechanism to prevent systemic oscillations. After continuously monitoring the frequency factor (number of regulation times per unit time) and amplitude factor (the amount of concentration or ratio change caused by each regulation), the system uses these two dimensions to construct the tension performance of the regulation path within a sliding time window. This tension is modeled as a time-amplitude coupling tension curve, with the horizontal axis representing the timing of regulation actions and the vertical axis representing regulation amplitude. After chronological arrangement, a dynamic curve reflecting the cumulative intensity and rhythm of regulation is formed. The system analyzes the slope (representing regulation density), local peaks (representing regulation impact points), and cumulative area (representing regulation load energy) of the curve. Real-time calculations are performed to assess whether the regulatory pathway constitutes a high-stress regulatory zone in the current or upcoming time period. When the coupling tension curve shows a continuous rise, a short-period high-density distribution, or a multi-peak high-amplitude pattern within a unit of time, the system determines that the current regulatory behavior poses an increased risk of disturbing the patient's blood microenvironment. Further assessment is needed to determine whether the system can maintain stable tolerance. Therefore, before the regulatory energy density exceeds the threshold, the system will prioritize running a predictive resistance analysis window. Within this window, the system models and assesses the patient's physiological buffering capacity, mainly including the current total body fluid volume, plasma osmolality, and HCO3-. - Parameters such as concentration, blood pressure stability, heart rate variability, renal residual metabolic capacity, and extracellular sodium buffer reserve are used to calculate a comprehensive buffer score based on historical data or real-time monitoring. This score represents the patient's current tolerance boundary to regulatory behavior. If the buffer score is lower than the system's preset safety tolerance threshold, it indicates that the patient is in a highly susceptible state. In this case, the system will trigger the rhythm gating mechanism in advance without waiting for the regulatory energy density to actually exceed the limit, thus limiting the amplitude, frequency, and continuity of subsequent regulatory behavior to prevent the cumulative risk of regulation from developing into systemic fluctuations or nonlinear collapse responses. Conversely, if the buffer score is good, the system will temporarily suspend gating activation, allowing regulation to continue until the next rhythm cycle.

[0057] After the rhythm gating mechanism is triggered, the trend of the regulating energy density is continuously tracked. If the system's regulating energy density continues to rise rapidly after gating activation, the regulating path reconstruction mechanism is automatically executed to re-plan the regulating timing, frequency, and amplitude distribution strategies of each ion component to avoid system instability caused by regulating inertia or multi-pathway conflicts. After the path reconstruction is completed, the system further calculates the entropy value of the regulating behavior caused by the current ratio scheme to quantitatively evaluate the system stability and regulating complexity of this round of dynamic regulation strategy during execution, and determines whether further dimensionality reduction processing is needed to simplify the regulation logic. The system performs sliding interval integration on the energy density of the regulating behavior, and comprehensively records the cumulative intervention load obtained by multiplying the sum of the squares of the amplitudes of each ion regulation per unit time by the regulation frequency. If the energy density value does not decrease in several consecutive cycles after the rhythm gating mechanism is triggered, and a sharp rise or plateau inflection characteristic is observed, the system determines that the current ratio strategy has not effectively curbed the regulating fluctuation behavior, and there are problems such as redundancy within the regulating path, regulatory logic conflict, or time window overlap mismatch. Therefore, the regulating path reconstruction mechanism is immediately initiated. The mechanism first freezes the current regulatory behavior sequence and then globally re-evaluates the migration trends, regulatory feedback lag characteristics, physiological importance weights, and interactions with other components of each target ion component. A new regulatory priority list is then established using the system's historical response database. Following the principles of minimum regulatory cross-interference and maximum feedback response efficiency, the temporal arrangement of each ion's regulation, the maximum allowable regulation frequency, and the range of single regulation amplitude are re-planned. Finally, an updated regulatory path blueprint is generated and loaded and executed in real time. After the new path has run for a certain period, the system enters the regulatory behavior entropy calculation stage. This entropy reflects the dispersion and information uncertainty of the current regulatory behavior in a multi-dimensional space, specifically including the dispersion of regulation amplitude distribution, the non-uniformity of regulation temporal intervals, and the non-linearity of regulatory response. The regulatory behavior entropy index is synthesized through a weighted average of multiple factors. If the entropy value is still higher than the system's preset stability reference value after reconstruction, the system will determine that the current ratio strategy is too complex and the rhythm is unstable, requiring further dimensionality reduction processing. This means compressing the number of current regulation variables, temporarily disabling the regulation permissions of low-impact components, or unifying multiple cross-coupled components into a single-channel adjustment mechanism. By reducing the dimension, the convergence of the regulation rhythm and the reduction of system load can be achieved, thereby enabling the dialysate ratio scheme to return to a controllable rhythm and convergence trend, and improving the stability, efficiency, and physiological adaptability of the entire regulation mechanism in dealing with highly dynamic scenarios.

[0058] Example 1:

[0059] Combined with appendix Figure 4In this embodiment, Mr. Liu, a 56-year-old patient with chronic renal insufficiency in stage 3, received routine hemodialysis three times a week. During one routine dialysis session, Mr. Liu's serum sodium ion concentration before dialysis was 137 mmol / L, while the sodium concentration in the prepared dialysate was 140 mmol / L. The system initially judged that a positive osmotic gradient should be formed from the dialysate to the blood, meaning that sodium ions should enter the blood. Within the first 15 minutes, the system's real-time monitoring found that the rate of increase in blood sodium concentration exceeded expectations, reaching +1.5 mmol / L / 15min. However, based on the empirical value of the average sodium migration rate for patients (generally 0.4–0.6 mmol / L / 15min), this rate was marked as abnormal migration, thus triggering the trend domain contradiction judgment mechanism. The system compared the current rate of change of sodium ion concentration (1.5 mmol / L / 15 min) with the direction of the dialysate osmotic gradient and found that although the direction was the same (positive), the increase was much higher than the predicted value. At the same time, the osmotic pressure in the blood did not increase synchronously, indicating that there may be temporary surface accumulation (surface migration) of sodium ions rather than real metabolic pull. Therefore, the system did not immediately include this trend in the trend domain to judge it as an effective regulatory signal, but recorded it as an contradictory trend.

[0060] To verify whether this trend was a genuine metabolic response, the system continued to track the sodium concentration trend at the 30-minute mark. At this point, the blood sodium level had only risen to 139.0 mmol / L, with the increase slowing down. Combined with the urea clearance rate, blood pressure changes, and the absence of typical hypersodium reactions such as thirst recorded within the system, this further supported the conclusion that the rapid increase in the early stage was a false trend driven by surface charge migration. Therefore, the system ultimately removed this trend from the regulatory guidance and did not adjust the sodium ratio concentration due to the temporary rapid increase, thus avoiding the potential pitfalls of hypersodium regulation.

[0061] In another case, during the same dialysis process, the system simultaneously monitored the potassium ion regulation pathway. At the start of dialysis, the blood potassium level was 5.2 mmol / L, the dialysate level was 2.0 mmol / L, and the transmembrane gradient drive direction was negative, indicating a significant removal of potassium from the blood. Over four consecutive dialysis cycles (15 minutes each), the system recorded the potassium ion migration direction as decreasing, slightly increasing, decreasing again, and then increasing again, indicating frequent migration trend reversals. The system activated its ion behavior inertial feedback logic to score the stability of the potassium trend. It found that the trend direction change of this component exceeded the allowable reversal frequency (the system is set to no more than one reversal per cycle), initially determining that its migration trend was directionally unstable. Therefore, even though potassium showed a further decreasing trend in the 5th cycle (with an amplitude of -0.6 mmol / L / 15 min), the system did not immediately adjust the dialysate potassium concentration based on this trend. Instead, it marked this trend as delayed confirmation, waiting for the trend to continue consistently in the next cycle before allowing the triggering of regulatory actions.

[0062] Assuming that potassium continues to decrease to 4.1 mmol / L in cycle 6, forming two consecutive decreasing trends, and the system monitors that the extracellular fluid conductivity is consistent with the potassium decreasing trend, and the patient does not experience adverse reactions to hypokalemia such as muscle tremors or arrhythmias, the system will then confirm the trend as effective and begin to fine-tune the dialysate potassium concentration to 2.3 mmol / L to achieve sustained-release regulation.

[0063] During the second hour of dialysis, the system detected elevated levels of bicarbonate (HCO3) in Mr. Liu's blood. - The bicarbonate concentration was 20.8 mmol / L, while the baseline target was set at 22.5 mmol / L, indicating a need for upward adjustment. However, to prevent short-term bicarbonate regulation from causing acid-base imbalance, the system first entered the metabolic compliance scoring module within the trend domain. Based on Mr. Liu's HCO3 concentration from his previous three dialysis sessions... - The regulatory response had an average response rate of +0.3 mmol / L / 30 min, classifying it as a slow-acting ion. Simultaneously, its resting metabolic rate before dialysis was low, with a body temperature of only 36.1℃ and a red blood cell distribution width (RDW) of 15.2% (slightly high), suggesting a low-speed, high-lag pattern in its overall metabolic activity. Based on historical regulatory response lag, basal metabolic activity, and immediate pH level (7.35), the system generated a metabolic compliance score of 0.47 (system threshold set at 0.65). Therefore, the current HCO32- content... - If the migration trend does not meet the synchronicity standard with the individual's metabolic capacity, the system will determine it as a non-compliant state, automatically delay the activation window of the regulatory pathway, and wait for the next pH monitoring cycle to confirm whether there are signs of upregulation before considering intervention.

[0064] At the same time, Mr. Liu's calcium ions (Ca 2+ ) Regulation of behavior and sodium (Na) + The adjustment was identified by the system as a high-interference combination. In the original dialysis mix, the sodium concentration was set at 141 mmol / L and the calcium concentration at 1.75 mmol / L. The system recorded that within 20 minutes after sodium adjustment, although sodium stability was good, calcium... 2+ A sudden drop in concentration to 1.59 mmol / L, accompanied by a decrease in blood pressure from 128 / 72 mmHg to 114 / 66 mmHg, suggests that sodium-induced changes in osmotic pressure may have induced changes in vascular tone, indirectly affecting calcium levels. 2+The membrane exchange rate changed. This effect was not shown in the initial planned trend coupling diagram. Based on the built-in induced association identification rules, the system marked this regulation as a non-obvious linkage phenomenon. At this time, the system automatically activated the interlocked regulation authorization mechanism, setting sodium and calcium as an interlocked pair and marking them as high interference level (score 0.87). It requires that in subsequent regulation actions, the two must maintain a response feedback time window of at least 15 minutes before they can exchange control, to avoid the risk of coordinated fluctuations in heart rhythm, tension, or bone calcium system caused by superimposed regulation.

[0065] Subsequently, at 2.5 hours of dialysis, the system detected potassium ions (K). + The regulatory behavior of [a specific substance] exhibited a high-frequency oscillation trend. Its concentration decreased from 4.6 mmol / L to 4.1 mmol / L within 30 minutes, but rebounded to 4.3 mmol / L within the following 5 minutes. The corresponding heart rate variability (HRV) curve showed a significant narrowing, suggesting a possible rhythm disturbance caused by high-frequency fine-tuning. According to real-time system calculations, the fluctuation period of this regulatory behavior was only 40 seconds per cycle. Mr. Liu's heart rate at this time was 72 bpm (≈0.83 Hz, i.e., a cycle of approximately 0.83 s). This fluctuation period was much smaller than his average single heart rate cycle (≈1.3 s), triggering the rhythm gating mechanism. Initially, the regulation was not directly blocked; instead, a delayed feedback observation period (10 minutes) was initiated, during which the system continuously assessed whether it was physiological oscillation interference or a spurious fluctuation. During this observation period, the system calculated the regulation energy density using a composite load index consisting of three factors: regulation amplitude (±0.2 mmol / L), regulation interval (less than 1 min), and regulation duration (4 consecutive times). The result was 1.78 (the system threshold was 1.5), confirming that the load exceeded the limit, and thus the gating was officially activated.

[0066] Furthermore, to determine whether the current regulatory behavior has indeed caused system instability, the system performs regulatory path reconstruction and behavior entropy calculation. Potassium regulation is temporarily suspended, and the system simultaneously reorders the regulation sequence of sodium, calcium, and bicarbonate, so that Ca... 2+ In K + The adjustment was completed beforehand to prevent it from being indirectly influenced. The entropy calculation of the adjustment behavior showed that the current ratio complexity was 0.84 (entropy range 0–1), higher than the safe entropy threshold of 0.65. Therefore, the system decided to enter a dimensionality reduction adjustment process, temporarily freezing the magnesium ion adjustment path and setting K... + With HCO3 - The regulation is merged into a single, gradual adjustment channel to simplify system behavior and restore rhythmic stability.

[0067] Example 2:

[0068] Combined with appendix Figure 5In this embodiment, the patient, Mr. Zhang, is 62 years old and suffers from hypertensive nephropathy, requiring long-term hemodialysis treatment. During one dialysis session, the system targets sodium ions (Na+). + The dynamic fine-tuning path identified three consecutive rounds of adjustment, each time by +0.3 mmol / L every 45 seconds. At the same time, the system's electrocardiogram monitoring module detected a slight disturbance in the patient's heart rate rhythm, with a 12% decrease in heart rate variability (HRV), suggesting a possible risk of rhythm interference induced by the adjustment.

[0069] To confirm the authenticity of the risk, the system calculates the spectral interference exponential function Λ(t) based on the projection of the current adjustment behavior into the frequency domain. Settings:

[0070] The current calculation window is from t0 = 60s to t1 = 180s;

[0071] The system sampling frequency is 1Hz;

[0072] The fundamental frequency of heart rate rhythm is approximately ω HR =1.1Hz, common high-frequency harmonics are 2.2 and 3.3Hz;

[0073] Adjusting behavior frequency ω mod The frequency is approximately 0.022Hz (45 seconds per cycle), which is considered a very low frequency. However, the system records that the power in its high-frequency sidewaves (such as the 0.9 to 1.5Hz range) is abnormally high.

[0074] κ(t) = 2.3, which is the hypersensitive state index regulated by the current system based on the patient's metabolic stress index (e.g., elevated lactate, low body temperature);

[0075] The frequency domain energy density curve is fitted as follows:

[0076] E(ω)=5e ―0.5(ω―1.2)2 Ω(ω)=3e ―0.3(ω―1.1)2 ,ψ(ω,t)=1 in ω∈[0.9,1.5]

[0077] Substitute into the interference exponential function formula:

[0078]

[0079] Substitute into the above equation:

[0080]

[0081] The calculated result is: Λ(t)≈5.82, while the system threshold is θ. Λ =3.5, therefore, it is determined that this sodium regulation behavior has a significant potential to interfere with the heart rate spectrum. The system marks it as a candidate event for rhythm impact and prepares to initiate the gating mechanism.

[0082] Example of path overlap detection:

[0083] Next, the system analyzes the current sodium regulatory behavior and calcium ion (Ca) relationship. 2+ Adjusting the time distribution of the paths revealed:

[0084] Sodium adjustment window: τ Na = [60s, 180s];

[0085] Calcium regulatory window: τ Ca = [150s, 240s];

[0086] The overlapping interval between the two: τ Na ∩τ Ca = [150s, 180s] = 30s;

[0087] The cross coefficient of the adjustment path is then calculated as follows:

[0088]

[0089] The system-defined cross threshold ξ th =0.25, therefore Ξ is satisfied. ij >ξ th This is a high-risk path overlap scenario. The system judges that if the gating is started immediately, it may cause false synchronization and blocking. Therefore, the activation time of the rhythm gating is delayed to after [200s], that is, the control is executed only after all overlapping windows are released.

[0090] Comprehensive strategy adjustment: During the delay period before the gating is officially triggered, the system initiates fine-tuning constraints, reducing the sodium regulation range from 0.3 mmol / L to 0.1 mmol / L, lengthening the adjustment interval to every 90 seconds, and simultaneously freezing Ca. 2+ The channel is adjusted for 20 minutes. If subsequent fluctuations do not recur and Λ(t) drops below the threshold, the system automatically removes the gating label; if the index continues to rise, it immediately enters a restricted state. This is achieved by introducing the spectral energy focusing function Λ(t) and the path crossover coefficient Ξ. ij It can quantitatively analyze the risk of regulatory behavior interfering with the patient's heart rhythm, and adaptively adjust the gating trigger window and control level by combining information on the overlap of high-sensitivity component paths.

[0091] The system has entered a pre-triggered state of rhythm gating due to excessive spectral energy focusing caused by sodium ion regulation. To avoid misjudgment leading to regulatory interruption or over-control, the system enters a delayed feedback observation period according to the mechanism, and monitors the cross-link effects of this regulatory pathway on other physiological indicators in real time.

[0092] The first crucial step after entering the observation period is to identify the osmotic pressure and vascular tone responses induced by sodium regulation. In Mr. Zhang's case, three minutes after regulation, the system detected an increase in his plasma osmotic pressure from 295 mOsm / kg to 302 mOsm / kg, accompanied by a brief fluctuation in blood pressure (from 120 / 70 mmHg to 112 / 68 mmHg). This change appears superficially consistent with the direction of sodium concentration regulation, but then, six minutes later, the system detected an increase in HCO3-. - The concentration decreased by 0.6 mmol / L (from 21.2 mmol / L to 20.6 mmol / L), but sodium regulation does not directly couple positively with the bicarbonate metabolic pathway, and there were no pH fluctuations or abnormal respiratory compensation. Therefore, the system judged this synchronous fluctuation of the indicator without a clear physiological causal chain as a possible systemic oscillation rather than a true feedback of the regulatory target. Thus, a validation delay was added to the sodium regulation pathway to prevent the system from incorrectly correcting the ratio based on false signals.

[0093] To further determine whether this fluctuation was an occasional disturbance, the system activated a segmented response window mechanism with delayed feedback. A 3-minute window was set for rapid response indicators such as blood pressure and heart rate, while a window for HCO3 was set... - A 10-minute observation window was set for slowly changing indicators such as pH. Mr. Zhang's osmolality and blood pressure returned to baseline levels within 5 minutes, but HCO3... - The trend failed to recover and continued to decline, leading the system to initially identify a false fluctuation signal. However, to ensure safety before degating, the system entered a level-two safety confirmation mechanism, analyzing HCO3 levels from Mr. Zhang's seven previous dialysis sessions. - Backtesting of the fluctuation trend revealed that the slope, oscillation amplitude, and response speed of the fluctuation curve all deviated from the historical standard range. In addition, 12 minutes after adjustment, the patient's body temperature and heart rate variability indicators returned to normal. The system comprehensively judged that the fluctuation was not a real physiological feedback, and therefore decided to release the gating restriction, restore the sodium ion regulation path, but reduce the regulation amplitude and frequency.

[0094] With the gating released, the system continued to track the combined distribution of the amplitude and frequency of the regulatory pathway over the next 30 minutes. By establishing a time-amplitude coupled tension curve, the system found that from the 20th to the 40th minute, although the individual sodium regulation amplitude was small (±0.2 mmol / L), the regulation frequency increased from once every 45 seconds to once every 30 seconds, resulting in a dense, bimodal distribution of the local tension curve, with the regional tension density exceeding the preset "high stress regulation zone" threshold. At this point, the system did not immediately restrict regulation but instead entered the predictive resistance analysis window to assess whether Mr. Zhang possessed sufficient buffering capacity. Assessment parameters included: plasma volume (still within a good range), extracellular sodium buffer index (above average), pH stabilization rate, and residual renal sodium excretion capacity (eGFR residual value 5.6 ml / min). The overall score was 0.82 (system critical threshold was 0.7), indicating that he possessed physiological buffering capacity, and there was no need to trigger gating prematurely; regulation could continue.

[0095] However, after 60 minutes, the system detected that the energy density trend of the regulatory behavior did not decrease with the restriction of the regulatory frequency. Instead, it fluctuated upward due to the synergistic intervention triggered by the reopening of potassium and calcium channels. As a result, the system triggered a regulatory pathway reconstruction mechanism, reordering the existing sodium-potassium-calcium pathway, postponing potassium regulation, limiting calcium regulation, and introducing CO2 to adjust the acid-base balance center.

[0096] After path reconstruction, the system calculates the entropy value of the adjustment behavior in this round, reflecting the stability of the current ratio adjustment in the strategy distribution. The original adjustment behavior entropy value was 0.91 (indicating complexity and instability), which decreased to 0.63 after reconstruction, close to the system stability threshold of 0.6, but still relatively high. The system determines that the current ratio structure is redundant, and thus enters the dimensionality reduction mechanism, temporarily freezing Mg. 2+ The regulation loop merges the sodium-potassium pathway into a group of slow-release control groups, achieving behavioral convergence through variable compression and path simplification.

[0097] From identifying suspicious feedback signals triggered by regulation and tracking physiological segmented responses during the delayed observation period, to predictive resistance analysis, tension assessment, pathway reconstruction, and entropy determination, this study demonstrates how a systematic and hierarchical intelligent judgment and regulation mechanism can achieve dynamic and safe ratio optimization in hemodialysis fluid preparation. This strategy not only avoids erroneous regulation due to false feedback but also improves the timeliness and stability of regulatory decisions, making it particularly suitable for dialysis scenarios with high-frequency fluctuations in multiple ion channels.

[0098] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic optimization of hemodialysis fluid ratio, characterized in that: Based on the migration directionality, intensity and temporal trend of target ions in transmembrane exchange, combined with the ion exchange stability at the beginning of dialysis, the buffer response hysteresis characteristics in the patient's body and the exchange inertia between blood and dialysate, a migration trend domain reflecting the active behavior trend of ions is formed, and the migration trend domain is used as the core criterion for dynamic adjustment of dialysate ratio. Based on the physiological inertia and membrane permeability of different ionic components during dialysis, and considering the migration trend domain, a regulatory hysteresis response time window for each ion is established. The timing, intensity, and duration of intervention of dialysate components are dynamically set to enable time-sensitive differentiated formulation of each component. To address the indirect cross-interference caused during the adjustment of ion components, a pre-defined mapping rule for inter-ion interactions is established. When dynamically adjusting the dialysate ratio, sensitive ion pathways are preferentially isolated before adjustments are implemented through priority sorting, time-sequential separation, and the introduction of buffer components. The frequency and amplitude of the real-time ratio adjustment process of each ionic component are dynamically monitored. When a cumulative fluctuation trend caused by high-frequency small-amplitude adjustments is detected, a rhythm gating restriction mechanism is triggered to limit the adjustment threshold and adjustment frequency of the components in order to suppress the blood microenvironment oscillation caused by excessive adjustment.

2. The method for dynamically optimizing the ratio of hemodialysis fluid according to claim 1, characterized in that... The migration trend domain establishes a contradiction judgment mechanism based on the consistency between the rate of change of ion concentration difference and the direction of the osmotic gradient of dialysate components, which is used to identify the deviation between surface migration and substantial metabolic traction. After the migration trend domain is formed, the logic of ion behavior inertia is introduced. If an ion frequently reverses direction in the historical trend, the current trend needs to be confirmed with a delay before it can be used for regulation.

3. The method for dynamically optimizing the ratio of hemodialysis fluid according to claim 2, characterized in that... The trend domain incorporates a metabolic compliance score to measure whether the current migration behavior is consistent with the patient's basal metabolic pattern. If not, the activation of the regulation window is delayed. During the establishment of the response time window, if there is no significant in vivo feedback response to the regulation of an ion within a preset time, the regulation authority of that component will be temporarily frozen. The response time window will adaptively scale based on fluctuations in parameters reflecting metabolic activity, such as patient body temperature, pH, or red blood cell distribution width, to synchronize with the metabolic rate.

4. The method for dynamically optimizing the ratio of hemodialysis fluid according to claim 3, characterized in that... In the mapping rule for multi-ion interference, induced associations are dynamically identified based on the dialysis process. The second-order reaction triggered by the adjustment of one ion indirectly pulls on another ion, and non-dominant linkages are identified first. For ion combinations with high interference levels, an interlocking regulation authorization mechanism is adopted, requiring that the minimum response feedback time between the adjustments of two ions must be met before they can be rotated for regulation.

5. The method for dynamically optimizing the hemodialysis fluid ratio according to claim 4, characterized in that... The rhythm gating mechanism is preferentially applicable to regulatory behaviors with fluctuation periods shorter than the average period of a patient's single heart rate. After the regulation frequency self-suppression mechanism is triggered, the regulation is not blocked, but enters a delayed feedback observation period to determine whether it is a false fluctuation caused by physiological oscillation interference. The rhythm gating threshold is dynamically set by the system's cumulative regulation energy density, and is a composite load index composed of three factors: regulation amplitude, regulation interval, and regulation duration.

6. The method for dynamically optimizing the ratio of hemodialysis fluid according to claim 5, characterized in that... The regulatory behavior with a fluctuation period shorter than the average heart rate period must satisfy the requirement that the fluctuation spectrum energy is concentrated in the heart rate fundamental frequency and high-frequency harmonic range, so as to determine the interference with the heart rate rhythm and to preferentially activate the rhythm gating mechanism. Before the rhythm gating mechanism is activated, it is necessary to verify whether there is a time distribution overlap in the regulatory behavior of the target ion and the regulatory path of other high-risk components, and adjust the gating delay or advance window.

7. The method for dynamically optimizing the ratio of hemodialysis fluid according to claim 6, characterized in that... The delayed feedback observation period is used to monitor the regulation results and its cross-index transduction effects, including osmolarity fluctuations, vascular tone parameters and metabolic reflex delays, in order to comprehensively determine whether the regulation is triggered by a false signal. If, during the observation period, an indicator that has no direct physiological link to the regulation target exhibits synchronous fluctuations, the system classifies this as a systemic oscillation and adds a verification delay to that segment of the regulation path.

8. The method for dynamically optimizing the ratio of hemodialysis fluid according to claim 7, characterized in that... The delayed feedback observation period is divided into segmented response windows based on different physiological feedback types, which are used for rapid response indicators including blood pressure and those including HCO3. - The slow-changing indicators are detected synchronously with a delay; when the regulatory behavior is determined to be a false fluctuation and the gating restriction is about to be released, a secondary safety confirmation mechanism must be entered first, which includes trend stability backtesting and post-wave recovery analysis of patient vital signs.

9. The method for dynamically optimizing the ratio of hemodialysis fluid according to claim 8, characterized in that... The amplitude and frequency factors of the regulation are combined in the assessment to introduce time and amplitude coupling tension curves to evaluate whether the regulation path constitutes a high-stress regulation zone within a certain time window. Before judging that the regulation energy density threshold exceeds the limit, a predictive resistance analysis window is run to assess whether the regulation response zone that the patient is about to enter has physiological buffering potential. If not, gating is triggered in advance.

10. The method for dynamically optimizing the ratio of hemodialysis fluid according to claim 9, characterized in that... If the energy density trend continues to rise rapidly after the gating is triggered, the regulation path reconstruction mechanism is executed to re-plan the regulation timing, frequency and amplitude distribution of each ion component. The rhythm gating mechanism generates regulation behavior entropy value after regulation path reconstruction, which is used to evaluate the stability of the current ratio mode and the complexity of regulation behavior, so as to determine whether further dimensionality reduction processing is needed.

Images