Adjusting system applied to continuous kidney replacement treatment of acute kidney injury patient

Abstract

The invention relates to an adjusting system applied to continuous kidney replacement treatment of acute kidney injury patients. Comprising the following steps: capturing the blood flow velocity and vascular endothelial permeability of glomerular capillaries in real time by using a micro sensor or a nano-scale optical imaging technology; the method comprises the following steps: pre-judging a risk time point of insufficient tissue perfusion through dynamic modeling of a capillary structure and a functional relationship; combining the microcirculation data with conventional biochemical indexes of the patient to create an early intervention model; when it is monitored that microcirculation perfusion is greatly reduced or kidney oxygen supply is insufficient, continuous renal replacement therapy CRRT parameter fine adjustment is triggered in advance; the method comprises the following steps: collecting plasma toxins of different patients at clinical and molecular levels, and constructing a common toxin map of the acute kidney injury AKI; whether the specific toxin concentration exceeds the standard or not is judged by combining a real-time sensor, and the removal priority is output in real time; a membrane material with an adjustable aperture or a special coating is adopted, so that inflammatory mediators and bacterial endotoxin molecules are adsorbed or intercepted, and the permeability of micromolecular electrolyte is kept.

Description

Technical Field

[0001] The invention relates to a continuous renal replacement therapy adjustment system, in particular to a continuous renal replacement therapy adjustment system applied to patients with acute renal injury. Background Art

[0002] Although the continuous renal replacement therapy (CRRT) adjustment system currently used in clinical practice for patients with acute kidney injury (AKI) has made certain progress in hardware technology, parameter control and safety assurance, there are still many deficiencies and disadvantages in core aspects such as accuracy, individualization and intelligence, which limit its therapeutic effect and clinical response efficiency in the management of severe AKI. First of all, most of the existing systems still use systemic indicators such as blood pressure, urine volume, and heart rate as the main feedback basis. These indicators are slow to respond and have poor spatial representation. They cannot reflect the local microcirculation status of the kidneys in real time, and are prone to delayed response after insufficient perfusion or excessive filtration, and cannot achieve early identification and intervention in the early stage of disease deterioration.

[0003] Secondly, the existing CRRT equipment still mainly relies on static weight changes or empirical judgments in terms of volume adjustment, lacks accurate identification of the real distribution structure of body fluids, and cannot distinguish the proportional changes between intracellular fluid, extracellular fluid and plasma volume. As a result, when dealing with water and sodium retention, hypovolemia or interstitial fluid accumulation in clinic, it is impossible to make differentiated treatments based on physiological space, which is prone to problems such as "excessive fluid replacement" or "excessive dehydration", increasing the prerenal burden. In addition, the current system still mainly relies on manual adjustment in the control strategy of filtration rate, fluid replacement rate, and dialysate composition, with low adjustment frequency and long cycle, and most of them lack logical modeling and risk prediction mechanisms, and cannot achieve comprehensive judgment on the relationship between the patient's metabolic state, hemodynamic changes and microcirculatory perfusion level, resulting in the treatment plan often lagging behind the actual changes in the disease, affecting the optimal use of the treatment window. In terms of data fusion and modeling capabilities, existing systems generally lack unified modeling and logical integration of multidimensional physiological signals such as key pathological data such as lactate concentration, creatinine clearance rate, and inflammatory factor levels. They lack risk indexes or dynamic control functions and cannot achieve personalized parameter output and early intervention evaluation. At the same time, most devices have not yet integrated bedside BIA analysis, ultrasound volume monitoring, and microcirculation assessment modules, and still use the "single parameter single model" method for local regulation, lacking a system perspective, which can easily cause an imbalance in the regulation rhythm between filtration-volume-perfusion.

[0004] In addition, in terms of treatment strategies, most current systems operate in a "continuous mode" and lack the ability to regulate and judgment criteria for intermittent and pulsed filtration. It is difficult to dynamically switch the treatment rhythm according to the changes in the patient's condition, making it difficult for some patients with unstable volume or easily fluctuating hemodynamics to tolerate, increasing the risk of adverse reactions such as hypotension and arrhythmia. In terms of informatization and intelligent feedback, most CRRT systems still do not establish a truly closed-loop feedback control system. Even if they have monitoring functions, they lack algorithm support for converting the collected data into adjustment instructions, resulting in medical staff having to rely on experience or intermittent evaluation to decide whether to adjust parameters, which not only consumes manpower but also has a problem of response lag, and is not conducive to grasping the rescue window. More importantly, most current systems fail to incorporate the dynamic identification of toxin types into the treatment plan formulation process. For pathogenic molecules such as inflammatory factors, medium and large molecular toxins, and endotoxins, they have not established a clearance priority ranking model, nor have they matched appropriate filter membrane material strategies. This makes it difficult to achieve efficient and precise molecular clearance when dealing with complex AKI etiologies such as septic AKI or postoperative hyper-inflammatory states, with unclear filtration goals and fixed membrane materials. In addition, in terms of personalized treatment, most current systems do not have the ability to dynamically adjust treatment parameters based on the patient's basic condition, disease course progression, and metabolic load differences, and have not established a closed-loop system of "condition-driven - model decision - automatic execution". In clinical applications, standardized and templated programs are often used, making it difficult to meet the treatment needs of highly variable critically ill patients. Summary of the Invention

[0005] The purpose of the present invention is to provide an adjustment system for continuous renal replacement therapy for patients with acute kidney injury, so as to solve some of the drawbacks and deficiencies pointed out in the background technology.

[0006] The technical solutions adopted by the present invention to solve its above technical problems include the following steps:

[0007] S1. Dynamic assessment and intervention modeling of renal microcirculation:

[0008] S1.1. Use a micro sensor or nanoscale optical imaging technology to capture the blood flow velocity of glomerular capillaries and vascular endothelial permeability in real time; through dynamic modeling of the relationship between microvascular structure and function, predict the risk time point of insufficient tissue perfusion;

[0009] S1.2. Combine the microcirculation data with the patient's routine biochemical indicators to create an early intervention model; when a significant decrease in microcirculation perfusion or insufficient renal oxygen supply is detected, trigger fine-tuning of the parameters of continuous renal replacement therapy (CRRT) in advance;

[0010] S2. Multidimensional toxin identification and selective clearance:

[0011] S2.1. Collect the plasma toxin compositions of different patients at the clinical and molecular levels, construct a common toxin map for acute kidney injury (AKI); combine with a real-time sensor to determine whether the concentration of a specific toxin exceeds the standard, and output the clearance priority in real time;

[0012] S2.2. Use a membrane material with an adjustable pore size or a special coating to adsorb or retain inflammatory mediators and bacterial endotoxin molecules, while maintaining permeability to small molecule electrolytes; enable different filter materials to be switched or applied hierarchically in the same continuous renal replacement therapy (CRRT) process through the system;

[0013] S3. Personalized volume management and tissue perfusion coupling strategy:

[0014] S3.1. Use bioelectrical impedance or ultrasound evaluation techniques to divide the patient's body volume into three compartments: intracellular fluid, extracellular fluid, and plasma volume; according to the changes in the compartment ratio, treat different pathological types of blood volume and interstitial fluid excess differently;

[0015] S3.2. Integrate microcirculation evaluation, systemic blood pressure, and lactate value data to give a tissue perfusion rating of low, medium, and high grades, and moderately adjust the filtration rate or fluid replacement speed according to the rating;

[0016] S3.3. Generate a personalized fluid replacement formula based on the volume compartment and perfusion rating, and at the same time comprehensively consider the electrolyte level and toxin accumulation rate; adopt a small-step correction strategy to avoid hemodynamic shock caused by large-scale adjustments.

[0017] Further, the renal microcirculation dynamic evaluation and intervention modeling method includes:

[0018] Adopt a micro-optical sensing device integrated on a renal region-guided probe, and obtain three types of key characteristic data of glomerular capillaries in the target area within a sub-second time scale through optical coherence reflectometry and velocity scattering techniques: blood flow velocity v g (t) represents the blood flow velocity through the glomerular microvessels per unit time; endothelial permeability τ e (t) reflects the degree of permeability of the endothelial cells of the vascular wall to plasma components; perfusion density ρ c (t) is the density level of vascular perfusion per unit area or volume;

[0019] After obtaining the above three types of data, introduce a microvascular structure and function coupling model to quantify the immediate functional state of the microcirculation. The function is set as follows:

[0020]

[0021] Where:

[0022] M(t) is the glomerular microvascular function index at the current moment; α1 and α2 are empirical fitting weight parameters tuned from the training set of clinical samples; the first term is used to characterize the fluctuation effect of blood flow under the condition of decreased endothelial barrier stability. If the permeability increases, the blood flow change will be amplified; the second term ln(1 + ρ c (t)) models the perfusion density of the microvascular network on a logarithmic scale.

[0023] Furthermore, the dynamic assessment and intervention modeling method for renal microcirculation includes:

[0024] Extract key biochemical indicators from the patient's monitoring data and construct a multi-parameter metabolic imbalance index F b (t) to express the degree of deviation of the current internal environment. The formula is as follows:

[0025]

[0026] Where:

[0027] F b (t) is the current metabolic stress score in the patient's body. The lower the value, the more the metabolic state deviates from normal; β1, β2, and β3 are weight adjustment coefficients adjusted according to the medical condition complaint to strengthen the influence of a certain type of metabolic imbalance; B cr (t) is the serum creatinine concentration, indicating the kidney's detoxification ability; B ph (t) is the blood pH value, reflecting the acid-base balance. 7.4 is the physiological neutral reference value; B Na (t) is the current serum sodium concentration, and 140 mmol / L is the standard reference value. Through three normalization transformations, indicators with different dimensions are incorporated into a unified judgment framework, so that the state of high acidosis or high sodium will not be masked by a slight increase in serum creatinine.

[0028] Furthermore, the dynamic assessment and intervention modeling method for renal microcirculation includes:

[0029] Overall assessment of the microcirculation state and metabolic imbalance state, constructing a comprehensive risk scoring function R(t), and using it as the key reference basis for judging whether to trigger an intervention:

[0030] R(t) = γ1 · M(t)γ2 · F b (t)

[0031] Where:

[0032] The score result of R(t) reflects the comprehensive risk level of current kidney perfusion; γ1 and γ2 respectively represent the weight factors of microcirculation and metabolic imbalance in risk decision-making; if the value of R(t) continuously decreases and is lower than the intervention threshold T1 set by the system, the system will initiate early risk warning; if it further drops to T2 < T1, the system will automatically trigger the adjustment of CRRT parameters, including the reduction of filtration rate, the adjustment of fluid replacement components, the reset of transmembrane pressure, and the adjustment of dialysate buffer operation.

[0033] Furthermore, the method for constructing the personalized volume management and tissue perfusion coupling strategy:

[0034] Through bioelectrical impedance analysis (BIA) or bedside ultrasound technology, a dynamic model of the patient's body fluid state is established, and the total fluid is divided into three independent compartments: intracellular fluid (ICF), extracellular fluid (ECF), and plasma volume (PV); the ratio of the three reflects the real distribution state of the fluid in the body and provides the ability to identify volume compared with a single total volume; and an exponential function of the volume structure index is constructed to express the volume offset trend:

[0035]

[0036] Where:

[0037] C s (t) represents the volume structure index, which is used to evaluate the degree of abnormal body fluid distribution; ECF(t) represents the extracellular fluid volume of the patient at time t; PV(t) is the plasma volume, which is a component determining the circulatory perfusion pressure; ICF(t) is the intracellular fluid volume, which reflects the cell metabolic state; +1 is to avoid the denominator being zero and enhance the numerical stability of the model in the low fluid volume state; λ1 is an empirical adjustment coefficient, which is used to control the sensitivity of the function output.

[0038] Furthermore, the method for constructing the personalized volume management and tissue perfusion coupling strategy:

[0039] Fuse the microcirculation perfusion state and systemic indicators to form a perfusion level evaluation system; the blood flow velocity v g (t) is measured by a blood flow velocity sensor in the glomerular region, and systemic perfusion is represented by the mean arterial pressure (MAP) and metabolic perfusion efficiency markers such as lactate concentration (Lac); the three together determine whether the current tissue perfusion is effective; and an exponential function of the perfusion efficacy index is constructed as:

[0040]

[0041] Where:

[0042] P d (t) is the tissue perfusion efficacy index, which is used to judge the oxygen supply and perfusion quality at the tissue level; v g(t) is the blood flow velocity of the glomerulus per unit time, which is an important indicator reflecting the patency of microcirculation; MAP(t) is the mean arterial pressure at time t; the index part e -(MAP(t)-65) It reflects the nonlinear response of perfusion state to blood pressure changes; Lac(t) is the blood lactate concentration, which is used to assess the degree of tissue hypoxia; the logarithmic function of the second term ln(1+Lac(t) 2 ) can inhibit the effect of lactate surge on index fluctuation; λ2 and λ3 are the weight parameters of perfusion contribution and metabolic inhibition respectively, which are set individually according to the patient's basic status; through the bipolar pull structure of positive incentive of fast blood flow and high blood pressure and negative penalty of high lactate, the dynamic quantitative classification of perfusion status is achieved.

[0043] Furthermore, the method for constructing the personalized capacity management and tissue perfusion coupling strategy is as follows:

[0044] Once the volume distribution including interstitial fluid accumulation shifts and the perfusion efficiency decreases, the system will control the CRRT parameters based on these two core indicators: dynamically adjust the filtration rate, fluid composition and treatment mode; and capture the linkage change trend and introduce the linkage adjustment trigger function:

[0045]

[0046] in:

[0047] A r (t) is the treatment regulation activation function. The larger the value, the more drastic the interactive changes in volume and perfusion are, and the treatment strategy should be adjusted immediately. represents the instantaneous coordinated change rate between volume structure and perfusion efficiency; the second term is the cosine adjustment factor, PV(t) / (ECF(t)+ICF(t)+1) represents the proportion of plasma volume to total body fluid; π is used to control the periodicity of the cosine function; λ4 modulation coefficient is used to set the weight of the cosine correction term in the overall control; function A r When the value of (t) exceeds a certain threshold value T1 or is lower than a certain preset interval T2, the corresponding intervention logic is triggered.

[0048] The present invention is applied to the adjustment system of continuous renal replacement therapy for patients with acute kidney injury, which has highly intelligent, individualized and prospective control capabilities, significantly improves the response efficiency and treatment safety of CRRT (continuous renal replacement therapy) under complex pathological conditions, and has the following beneficial effects:

[0049] First, it realizes the real-time dynamic acquisition of key parameters such as glomerular microcirculatory blood flow velocity, vascular endothelial permeability and perfusion density, breaking through the hysteresis limitation of traditional CRRT relying on systemic indicators (such as blood pressure and urine volume), and can detect microcirculatory precursors of insufficient renal tissue perfusion earlier;

[0050] Second, by constructing a microcirculation structure-function coupling model, a metabolic imbalance index, and a volume-structure index function, quantitatively evaluate the perfusion efficacy, degree of metabolic disorder, and trend of fluid distribution shift in patients, providing a precise decision-making basis for adjusting personalized treatment parameters;

[0051] Third, systematically integrate multi-source information such as hemodynamics, microcirculation perfusion, and body fluid distribution, establish a linkage regulation trigger function. When the system detects a trend of coordinated deterioration of volume-perfusion or the value exceeds the threshold, it can automatically control the filtration rate, fluid replacement components, transmembrane pressure, and treatment mode to achieve dual-closed-loop management of continuous regulation and risk warning. Brief Description of the Drawings

[0052] Figure 1 It is a flowchart of an adjustment system for the continuous renal replacement therapy of the present invention applied to patients with acute kidney injury.

[0053] Figure 2 It is a flowchart of a method for dynamically evaluating and intervening in the renal microcirculation and modeling of the present invention.

[0054] Figure 3 It is a flowchart of a method for constructing a personalized volume management and tissue perfusion coupling strategy of the present invention. Detailed Embodiments

[0055] The following provides a detailed description of the specific embodiments of the present invention in conjunction with the drawings.

[0056] Combined with the attach Figure 1As shown, this system addresses the condition fluctuations and organ perfusion risks during the continuous renal replacement therapy (CRRT) for patients with acute kidney injury. It proposes an adjustment system with prospective identification capabilities and an intelligent response mechanism. The core lies in achieving early prediction of deteriorating renal function and fine-tuning of dynamic treatment parameters through high-precision and real-time microcirculation monitoring methods, combined with physiological modeling and biochemical index fusion algorithms. First, the system relies on micro sensors or nanoscale optical imaging technology, deployed in the renal region or the proximal end of the blood pipeline. Through non-invasive or minimally invasive means, it continuously captures the blood flow velocity at the glomerular capillary level (reflecting the dynamic state of blood passing through the microcirculation system) and the changes in vascular endothelial permeability (reflecting the microvascular barrier function and the risk of inflammatory leakage). These two data can directly reflect the integrity of renal unit perfusion and the microcirculation function state. After obtaining the original data, the system will establish a dynamic relationship model of microvascular structure-function based on time series modeling. The core logic is to transform the coupling relationship between blood flow velocity and endothelial permeability into a continuous function change curve to identify the potential precursor of perfusion insufficiency, namely "decreasing flow velocity - increasing permeability", so as to judge whether the glomerular microperfusion is in a state of progressive deterioration. At the same time, to improve the accuracy of risk identification and clinical feasibility, the system introduces routine biochemical indicators of patients, such as serum creatinine, blood urea nitrogen, electrolyte levels (potassium, sodium), pH value, and lactate, etc., to fuse with microcirculation indicators. By constructing a multivariate aggregation function or an early intervention scoring model, it quantitatively evaluates the current renal metabolic load and blood supply of the patient, and then predicts the possible critical points of renal function in the next few hours.

[0057] When the model identifies that the microcirculation perfusion is significantly decreasing, or the systemic indicators indicate that the oxygen supply to the kidneys cannot meet the metabolic needs, that is, there is a high-risk trend of acute hypoperfusion, focal hypoxia, or inflammatory activation, the system will immediately trigger the fine-tuning mechanism of CRRT parameters, including reducing the filtration rate, optimizing the fluid replacement components, adjusting the electrolyte concentration in the dialysate, or changing the treatment rhythm (such as adjusting from continuous mode to intermittent pulse mode), so as to reduce the renal unit perfusion load, stabilize the microcirculation environment, and prevent further damage from occurring, thus achieving a technological leap from traditional "passive adjustment" to "active prediction + personalized micro-control", and improving the prospectiveness, safety, and effect accuracy of the treatment.

[0058] This system aims to improve the treatment efficiency and individualization level of continuous renal replacement therapy (CRRT) in patients with acute kidney injury (AKI) through a refined toxin identification and hierarchical clearance strategy. The key lies in breaking through the technical limitations of the traditional "broad-spectrum clearance and non-discriminatory filtration", and instead constructing an intelligent adjustment mechanism that combines toxin type identification, clearance priority ranking, and filter material selection logic. The system first collects plasma samples from different AKI patients on a large scale at the clinical and molecular levels. Through mass spectrometry analysis, biochemical detection, and molecular screening techniques, plasma toxins closely related to the progression of AKI are extracted and classified, including small molecule metabolic wastes (such as creatinine, urea), medium molecule inflammatory factors (such as IL-6, TNF-α), macromolecule bacterial endotoxins (such as lipopolysaccharide LPS), etc. Based on this, an "acute kidney injury toxin map" is established, forming a pathological molecular feature database that can be compared among patients. During the operation of the system, highly sensitive sensors installed in the CRRT pipeline are used to continuously monitor the concentration levels of specific toxins in the blood, and by comparing with the warning thresholds in the toxin map, it is judged which toxins have reached the clearance intervention standard, thereby dynamically outputting the "toxin clearance priority", that is, the basis for judging which molecules to clear first and by what means. On this basis, the system introduces new types of adjustable structure filter membrane materials, including filters with controllable pore size adjustment capabilities (achieved through pressure adjustment, temperature-sensitive materials, or intelligent polymers), and membrane materials covered with bioselective coatings (such as polymers with adsorption groups, heparin analogs, or high-affinity site structures), which are used to precisely adsorb or intercept specific medium and large molecule toxins such as inflammatory mediators or bacterial endotoxins, while maintaining good permeability to small molecule electrolytes such as sodium, potassium, and calcium, thus achieving "selective filtration" rather than traditional "broad-spectrum filtration". Furthermore, the system can control the switching or superimposed use of different types of filter materials in the CRRT process according to the current toxin concentration gradient and priority changes.

[0059] In the issue of fluid management and perfusion regulation, a dynamic adjustment mechanism based on the three-compartment modeling of body fluids is proposed, which integrates perfusion assessment and individualized fluid replacement strategy. The core technology is to improve the precision of volume identification, the targeting of intervention response and the hemodynamic stability of the treatment process. The system first continuously evaluates the patient's body fluid status through bioelectrical impedance analysis (BIA) or bedside ultrasound technology, and divides the total fluid into three compartments with clear anatomical and functional significance: intracellular fluid (ICF), which represents the stability of the cell metabolic environment; extracellular fluid (ECF), including interstitial fluid, which is the area most prone to retention and edema; plasma volume (PV), which is the core component of the circulatory system and directly affects blood pressure and perfusion level. The system calculates the volume ratio and change trend between each compartment based on real-time monitoring data. When it is found that ECF increases significantly and PV is normal, it indicates that the interstitial fluid is excessive but the circulating volume is still acceptable. At this time, blind dehydration should be avoided; on the contrary, if PV increases, there may be a real risk of blood volume overload. Based on this capacity structure recognition, the system further integrates the microcirculatory perfusion indicators of the glomerular region (such as blood flow velocity), systemic blood pressure data (such as MAP), and indicators reflecting metabolic oxygen supply (such as blood lactate level Lac), and constructs a multidimensional perfusion scoring system, which divides the current patient tissue perfusion status into three levels: high perfusion, medium perfusion, and low perfusion. The high perfusion state is suitable for accelerating the filtration rate and improving toxin clearance, the medium perfusion state can maintain the current parameters, and the low perfusion state should slow down the filtration rate, prevent the perfusion from further decreasing, and evaluate whether fluid support is needed. Based on the volume compartment structure and perfusion rating, the system also comprehensively considers the patient's current electrolyte level (such as sodium, potassium, and calcium concentrations) and the toxin accumulation rate (calculated by the rising rate of creatinine and urea nitrogen) to generate a dynamically updated individualized fluid replacement formula, including water volume, electrolyte concentration, alkaline buffer ratio, etc., to ensure that fluid replacement can maintain electrolyte balance without causing additional capacity burden through fine proportioning. The entire regulation process adopts a small step strategy, that is, fine-tuning key parameters at short intervals (such as 10-15 minutes) rather than drastically changing the filtration rate or fluid infusion rate at one time, to avoid hypotension, sudden changes in volume or cardiovascular stress caused by drastic fluctuations in treatment parameters. Ultimately, a multi-dimensional integrated adjustment model of accurate volume identification, graded perfusion management, targeted electrolyte supplementation and gradual adjustment of CRRT parameters is achieved, effectively improving the treatment safety and individualized intervention capabilities of patients with acute kidney injury.

[0060] Embodiment 1:

[0061] Combined with Figure 2As shown below, through a clinical simulation example, the data acquisition, calculation methods, and the formation path of treatment decisions for each step are described in detail. Patient Wang, male, 62 years old, developed acute kidney injury (AKI) after severe pneumonia complicated with septic shock and was admitted to the intensive care unit. At the initial stage of CRRT treatment, the urine output continued to decrease, the serum creatinine increased to 412 μmol / L, the lactate concentration was 3.8 mmol / L, and the blood pressure was unstable. When evaluating whether to increase the filtration rate to remove toxins, the medical team decided to introduce the renal microcirculation function index modeling method described in the present invention for auxiliary decision-making.

[0062] First, the team wore an integrated kidney-guided sensing probe on Wang's kidney area. This device uses a micro-optical system based on the principles of optical coherence reflection and particle scattering to capture three dynamic parameters of glomerular capillaries per second. The data collected in one acquisition showed that the blood flow velocity v g (t) was 0.74 mm / s, the endothelial permeability τ e (t) was 0.48 (unit: dimensionless, relative permeability index), and the perfusion density ρ c (t) was 15.2 (unit: number of capillary densities per mm 2 of internal capillaries); the data was updated in a rolling manner with a 5-second sampling window. The system then substituted the data into the microcirculation function modeling function for calculation:

[0063]

[0064] Among them, the parameters α1 and α2 were obtained by regression fitting of the hospital's big data and were set to 1.2 and 0.85 respectively. Both were median values within the typical clinical general range, and the value ranges were α1 ∈ [1.0, 1.5] and α2 ∈ [0.7, 1.0]. Substituting the above data into the model, we get:

[0065]

[0066] M(t) = 1.2·sin(0.5) + 0.85·ln(16.2)

[0067] M(t) ≈ 1.2·0.4794 + 0.85·2.785

[0068] M(t) ≈ 0.5753 + 2.36725 ≈ 2.9426

[0069] The system showed that the patient's microcirculation function index at this moment was 2.94. According to the clinical threshold of the model established by the present invention (the normal reference range is 3.5 - 4.8), this value was in the risk interval of mild to moderate perfusion function decline, indicating mild perfusion insufficiency in the glomerular region and accompanied by an increase in blood lactate, reflecting that the oxygen supply to the renal tissue could no longer fully meet the metabolic needs.

[0070] At this time, the system links the CRRT main control module to start the parameter fine-tuning mechanism: first, the filtration rate is automatically reduced from the original 35mL / kg / h to 25mL / kg / h to avoid further reducing the glomerular pressure perfusion. Secondly, the fluid composition is adjusted to reduce the original sodium concentration from 140mmol / L to 135mmol / L, and the bicarbonate ratio is slightly increased to support acid-base balance to avoid acidosis and increase the perfusion load. The entire parameter adjustment process lasts for 6 hours, and the system automatically re-collects data and updates Mt every 15 minutes. After 4 hours of treatment, the blood flow velocity was measured again and rose to 0.95mm / s, the permeability dropped to 0.31, and the perfusion density increased to 18.1. The model was substituted again:

[0071]

[0072] M(t)=1.2·sin(0.725)+0.85·ln(19.1)

[0073] M(t)≈1.2·0.663+0.85·2.949≈0.7956+2.5067≈3.3023

[0074] At this time, the Mt value has risen to 3.30, which is a significant improvement. The system prompts that the microcirculation function has recovered and recommends gradually restoring the original filtration intensity. In the end, the patient's creatinine dropped to 289μmol / L within 48 hours, urine volume recovered to 800mL / d, blood pressure stabilized at 108 / 72mmHg, and there was no further low perfusion fluctuation. This example fully demonstrates that the system of the present invention can effectively identify the state of glomerular local perfusion in patients with acute kidney injury and intervene in time through real-time microcirculation data acquisition, function modeling, indicator quantification and regulation feedback, making CRRT treatment more proactive, individualized and predictable, enhancing the safety and targeting of treatment, and is particularly suitable for intelligent renal support management in intensive care units and high-risk CRRT scenarios.

[0075] In order to further judge the overall metabolic stability of the internal environment, this example decided to introduce the metabolic imbalance index F proposed by the present invention. b (t) Perform multi-dimensional cross-validation. The system automatically captures three key indicators from the patient's current CRRT interface and biochemical laboratory: serum creatinine concentration, B cr (t), blood pH value B ph (t), blood sodium concentration B Na (t), at this time, the patient's experimental data are as follows: blood creatinine 412μmol / L, pH 7.28, blood sodium concentration 147mmol / L, reflecting obvious metabolic load and mild acidosis. The system starts the metabolic index function calculation path based on this, and the formula used is:

[0076]

[0077] Among them, each weight coefficient is set based on the chief complaint of the condition. At this time, the patient is in a state of internal environmental disorder dominated by "metabolic acidosis + hypernatremia". Therefore, the set coefficients are: β1 = 0.15, β2 = 0.55, β3 = 0.30, all within the recommended range, which are: β1 ∈ [0.1, 0.2], β2 ∈ [0.4, 0.6], β3 ∈ [0.2, 0.4]. Substitute the values for calculation as follows:

[0078] First, process the serum creatinine item:

[0079]

[0080] The second item is the pH normalization processing item:

[0081]

[0082] The third item is the serum sodium deviation ratio processing item:

[0083]

[0084] Therefore, the final calculation result is:

[0085] F b (t) ≈ 0 + 0.4911 + 0.315 = 0.8061

[0086] The system sets the reference range according to the exponential judgment logic: F b (t) ≥ 1.5 indicates a good metabolic state, 1.0 - 1.5 indicates moderate metabolic stress, and F b (t) < 1.0 indicates obvious metabolic imbalance. The calculation result this time is 0.8061, which is significantly lower than 1.0, indicating that the current internal environment is in a state of metabolic stress. The main problem lies in the combined effect of the decline in acid-base buffering ability and high sodium level, rather than simply caused by renal detoxification failure. Therefore, the system does not regard the creatinine item as the main source of punishment, but highlights the dominant position of electrolyte and acid-base imbalance through the second and third items.

[0087] Based on the double-exponential results of M(t) = 2.94 and F b (t) = 0.8061, the CRRT intelligent control system calculates the overall perfusion-metabolism combined risk function R(t) = γ1·M(t) - γ2·F b (t). Set the parameters γ1 = 1.0, γ2 = 1.2 (the recommended coefficient range is γ1 ∈ [0.8, 1.2], γ2 ∈ [1.0, 1.5]), then substitute for calculation:

[0088] R(t) = 1.0·2.94 - 1.2·0.8061 = 2.94 - 0.9673 = 1.9727

[0089] The system sets a threshold range for R(t): when R(t) ≥ 2.5, it is the perfusion metabolism stable area; when 1.5 ≤ R(t) < 2.5, it is the transition warning area; when R(t) < 1.5, it is the immediate intervention area. At this time, the result is 1.9727, which falls within the transition warning area, indicating that although it is not in a highly dangerous state at present, the metabolic stress has weakened the perfusion repair ability.

[0090] Based on this, the system suggests fine-tuning the filtration parameters and adopting the strategy of "deceleration + alkaline dialysate buffering": the filtration rate is adjusted from 25 mL / kg / h to 20 mL / kg / h, the bicarbonate concentration in the dialysate is increased from 32 mmol / L to 35 mmol / L, the sodium concentration is decreased from 135 mmol / L to 132 mmol / L, and the filter membrane maintains medium flux to reduce the impact of increased blood sodium on vascular tension and buffer the state of metabolic acidosis at the same time. After 6 hours of intervention, the re-sampled data of the patient is as follows: the pH rises to 7.35, the sodium concentration drops to 142 mmol / L, and the creatinine drops to 368 μmol / L. Substitute into the formula:

[0091] The first item:

[0092]

[0093] The second item:

[0094]

[0095] The third item:

[0096]

[0097] Therefore:

[0098] F b (t) ≈ 0 + 0.5236 + 0.3042 = 0.8278

[0099] Although there is a slight improvement, F b (t) is still in the metabolic stress range of < 1. At this time, on the premise that the perfusion has increased (the previous Mt rose to 3.30), the system maintains the current buffer parameters unchanged but does not continue to decelerate to avoid reducing the toxin clearance efficiency due to excessive conservatism. The whole process shows that the model has a logical closed-loop in the evaluation, judgment, and decision-making path, and has clinical real-time feasibility and individual difference adaptability.

[0100] In this embodiment, to further determine whether the state of renal perfusion and internal environment imbalance has reached the intervention critical point, the treatment team continued to apply the integrated risk scoring function R(t) proposed in the present invention as a unified intervention decision variable running through the microcirculation and metabolic dimensions. The purpose of designing this function is to achieve the weighted integration of multiple pathological factors, so as to accurately identify the trend of disease deterioration and avoid misleading or delayed response caused by single indicators. At this time, the previous output result of the system was Mt = 3.30 and F_by = 0.8278, which were substituted into the following integration function:

[0101] R(t)=γ1·M(t)-γ2·F b (t)

[0102] Among them, the parameters γ1 and γ2 represent the weight distribution factors of the system for microcirculation indicators and metabolic pressure in the evaluation. According to the current disease characteristics of Wang (blood pressure fluctuates but the metabolic load is heavy), γ1 = 0.9 and γ2 = 1.3 are set, and both are within the recommended range: γ1 ∈ [0.8, 1.2], γ2 ∈ [1.0, 1.5]. Substitute each value into the calculation:

[0103] R(t)=0.9·3.30 - 1.3·0.8278 = 2.97 - 1.0761 = 1.8939

[0104] The risk grading interval of the system for R(t) is set as follows: if R(t) ≥ 2.5, it represents that the two-dimensional state of microcirculation and metabolism is good, and the current treatment strategy can be continued; if 1.5 ≤ R(t) < 2.5, it is the early warning area, indicating a potential deterioration trend, and it is recommended to enhance monitoring and consider fine-tuning; if R(t) < 1.5, it is the high-risk intervention area, and immediate intervention in the reconstruction of CRRT parameters is required to prevent metabolic collapse or sudden drop in perfusion. The score obtained from this calculation is 1.8939, which is clearly within the "early warning area". Therefore, the system automatically issues a yellow warning label and prompts medical staff to enter the "parameter fine-tuning plan". The system first adjusts the filtration rate with the lowest intervention intensity, reducing it from the original 20 mL / kg / h to 18 mL / kg / h to ensure that the perfusion pressure is not affected by sudden changes;

[0105] Secondly, based on the two-way pressure judgment of sodium and bicarbonate, the sodium concentration in the dialysate was adjusted from 132 mmol / L to 130 mmol / L, and at the same time, the bicarbonate concentration was fine-tuned from 35 mmol / L to 36 mmol / L to further improve the acid-base buffering efficiency. In addition, the transmembrane pressure setting was adjusted from 210 mmHg to 190 mmHg to reduce the shear stress at the membrane end and lower the risk of inflammatory factor release. The system re-evaluates R(t) every 30 minutes. If the values are stable or increasing for three consecutive times, the intervention can be confirmed as effective. If R(t) < T2 for any one time, where T2 is set to 1.2 (the default limit threshold of the system), the red intervention level will be immediately triggered, and the following operations will be automatically executed: the filtration rate will be reduced to the lowest safe value of 15 mL / kg / h; the dialysate will be switched to a high-alkaline buffer solution (HCO3 - to 38 mmol / L); at the same time, quickly evaluate whether vasoactive drugs need to be combined or additional volume replacement is required. After about 90 minutes of fine-tuning, data was collected again. At this time, the creatinine decreased to 345 μmol / L, the pH increased to 7.37, the sodium decreased to 140 mmol / L, the blood flow velocity increased to 1.03 mm / s, the endothelial permeability decreased to 0.26, and the perfusion density increased to 19.4. Substitute these values back into the model for calculation:

[0106] First, update M(t):

[0107]

[0108] Then update F b (t):

[0109] Creatinine = 345 μmol / L, pH = 7.37, Na = 140 mmol / L

[0110] Serum creatinine item ≈ 0, pH difference = 0.03, normalized value = 1 / 1 + 0.03 ≈ 0.9709, × 0.55 = 0.534

[0111] Sodium item = 140 / 140 = 1, × 0.3 = 0.3

[0112] Total F b (t) = 0 + 0.534 + 0.3 = 0.834

[0113] Substitute into the new R(t):

[0114] R(t) = 0.9·3.4364 - 1.3·0.834 = 3.0928 - 1.0842 = 2.0086

[0115] Although the value is still within the warning range, it is significantly higher than the previous 1.89, and the trend is upward. The system evaluates it as "the intervention is effective and tends to be stable", and recommends that if the trend continues after 6 hours, the filtration rate can be slowly restored to 20 mL / kg / h while maintaining the current fluid replacement plan unchanged.

[0116] Example 2:

[0117] Combined with the attached Figure 3 As shown, after continuous evaluation of the microcirculation index M(t), metabolic imbalance index F b (t) and comprehensive scoring function R(t) in Example 1, the treatment of Wang entered the stage of fine volume management. Since the patient had underlying cardiovascular diseases and showed manifestations of hypoperfusion combined with elevated lactate during CRRT, the team decided to further introduce the personalized volume management and tissue perfusion coupling strategy proposed in the present invention. By establishing the volume structure index function C s (t), the true state of fluid distribution was clarified, and it was used to assist in judging the intensity and direction of CRRT filtration and fluid replacement operations.

[0118] First, the system obtained Wang's three-compartment body fluid data in real time through a bioelectrical impedance analysis (BIA) device. The measurement data on the 3rd day of treatment showed that the intracellular fluid volume ICF(t) = 18.5 LL, the extracellular fluid ECF(t) = 16.2 LL, and the plasma volume PV(t) = 4.4 L, indicating that the extracellular fluid was significantly higher than the plasma volume. Clinically, there was mild to moderate lower limb edema, suggesting the presence of interstitial fluid retention, which might affect perfusion and filtration efficiency. According to the calculation formula of the volume structure index function of the present invention:

[0119]

[0120] Substitute the data, select the empirical adjustment coefficient λ1 = 2.5 within the recommended range, and the value range is λ1 ∈ [2.0, 3.0]. The calculation is as follows:

[0121]

[0122] The system identified the current volume structure index as 1.36. Referring to the risk demarcation range set in the present invention, if C s (t) < 0.8 indicates that the volume structure distribution is normal, 0.8 ≤ C s (t) < 1.2 is the mild deviation area, 1.2 ≤ C s (t) < 1.6 indicates moderate deviation, and if ≥ 1.6, it is a highly abnormal distribution. Wang's current value falls in the moderate deviation area, indicating relatively retained interstitial fluid, while the plasma volume is only moderate, and too fast filtration may cause true hypovolemia.

[0123] Combined with the previous microcirculatory function value M(t) = 3.43, the metabolic imbalance index F b (t)=0.83, fusion score R(t)=2.01, the system judged that microcirculation has gradually recovered, but the risk of volume distribution imbalance still needs intervention, so the volume coupling plan is automatically called out: 1) The ultrafiltration rate is reduced from the original 18mL / kg / h to 16mL / kg / h; 2) 1.5% albumin is added to the fluid replacement composition to increase plasma osmotic pressure, guide interstitial fluid back to the blood vessels, and optimize the PV ratio; 3) The dialysate sodium concentration is adjusted from 130mmol / L to 132mmol / L to slightly improve the vascular filling state. The system remeasures the volume compartment data every 4 hours. The second measurement result is: ECFt=15.4L, PVt=5.2L, ICFt=18.4L, recalculated:

[0124]

[0125] At this time, the capacity structure index dropped to 1.215, and it obviously tended to stabilize and entered the mild deviation zone, indicating that the treatment strategy was effective. The system recommends maintaining the current filtration rate unchanged, retaining the fluid replacement structure, and extending the monitoring cycle to once every 6 hours. From this example, it can be seen that the capacity structure index function of the present invention is not only more hierarchical in identifying "too much fluid" or "insufficient volume", but also can specifically analyze the state of "fluid distribution dislocation", so that the filtration and fluid replacement decisions can truly be connected with the patient's physiological state in real time, avoiding the risks of hypotension, microcirculation collapse, etc. caused by excessive dehydration or incorrect fluid replacement.

[0126] This embodiment further uses the "perfusion efficiency index function" proposed by the present invention. d (t), through the glomerular blood flow velocity v g The fusion model of three key indicators, namely, mean arterial pressure (t), mean arterial pressure MAP (t), and lactate concentration Lac (t), dynamically evaluates the true effectiveness of tissue perfusion. At this time, the real-time monitoring data shows: Wang's glomerular blood flow velocity v g (t) = 1.08 mm / s, MAP = 62 mmHg, blood lactate Lac(t) = 4.2 mmol / L, the system substitutes this into the perfusion efficiency function:

[0127]

[0128] The individualized parameter values ​​λ2=1.5 and λ3=0.9 within the recommended range are selected, where λ2∈[1.2,1.8] and λ3∈[0.7,1.1], reflecting that the current patients need to focus on perfusion dynamic support, and the lactate level should have a moderate inhibition weight. Substitute into the calculation as follows:

[0129] Calculate the positive part first:

[0130]

[0131] Calculate the negative part again:

[0132]

[0133] Finally, it is obtained that:

[0134] P d (t) = 0.0768 - 2.6325 = -2.5557

[0135] The system determines that the current perfusion efficiency index is -2.56, which is significantly low. Combining with the previous volume structure index C s (t) = 1.215 belongs to mild distribution abnormality, indicating that Wang has a complex perfusion disorder state of "microcirculation perfusion insufficiency + high metabolic load". At this time, the system of the present invention triggers a "perfusion level classification response": according to the set partition, if P d (t) ≥ 1.0 is good perfusion, 0 < P d (t) < 1.0 is moderate perfusion, if P d (t) < 0 is a state of low perfusion and high metabolic stress, and immediate intervention is required.

[0136] The system immediately enters the following strategies: 1) Lower the filtration rate to the lowest safe value of 15 mL / kg / h; 2) Actively increase the MAP target to ≥ 70 mmHg, transfer 3% hypertonic saline through CRRT-assisted fluid replacement and evaluate the necessity of using norepinephrine in combination; 3) Adjust the lactate metabolic buffering capacity of the dialysate and increase the bicarbonate concentration to 38 mmol / L; 4) To prevent dehydration from aggravating tissue perfusion insufficiency, suspend part of the dehydration plan and switch to mild isotonic fluid replacement. After 3 hours of intervention, the rechecked data: v g (t) = 1.35 mm / s, MAP = 68 mmHg, Lact = 2.8 mmol / L, substitute and recalculate:

[0137] Positive term:

[0138]

[0139] Negative term:

[0140]

[0141] Final result:

[0142] P d (t) = 1.929 - 1.962 = -0.033

[0143] The perfusion efficacy index improved significantly, rising from the original -2.56 to -0.033. Although it is still at the "low perfusion edge", it is already close to the medium perfusion area. The system evaluated it as "effective intervention", indicating an increase in perfusion efficiency. Subsequently, the filtration intensity can be slightly restored and the trend of lactate clearance can be continuously observed. This example clearly demonstrates the "non-linear sensitivity" and "sub-item controllability" of the perfusion efficacy index function of the present invention when dealing with complex states (such as low blood pressure, high lactate, and slow blood flow).

[0144] On the 5th day after Wang received CRRT treatment, although the indicators of the volume structure index C s (t) and the perfusion efficacy index P d (t) tended to be stable through separate modeling and intervention in the early stage, cyclic fluctuations and a slight decrease in urine output occurred again in the early morning. The system monitored a new dynamic deviation in the ratio of plasma volume PV to extracellular fluid ECF, and the perfusion efficacy fluctuated with a slight decrease in MAP and a slight increase in lactate. Therefore, the "coupled regulation trigger function" A r (t) proposed by the present invention was officially activated to quantify the volume-perfusion interaction trend and determine whether to enter a new round of parameter adjustment.

[0145] According to the previous evaluation data and the latest body fluid measurement results, at this time, Wang's C s (t) = 1.215, and the perfusion efficacy index P d (t) = -0.033. The system continuously calculated the change rate of C s (t)·P d (t) through a 5-minute sliding window, and obtained the derivative term indicating that the volume-perfusion coupling state is deteriorating slightly. At the same time, BIA measured that the plasma volume PV(t) = 5.1L, ECFt = 15.6L, and ICFt = 18.2L. Substituting into the cosine adjustment term and selecting the recommended modulation coefficient λ4 = 1.2, with the coefficient range λ4 ∈ [1.0, 1.5], then:

[0146]

[0147] A r (t) = -0.043 + 1.0752 = 1.0322

[0148] The system sets the safety interval of the coupled regulation trigger function as: T2 = 0.5, T1 = 1.0, that is, when A r (t) > 1.0 or A rWhen (t) < 0.5, the adjustment mechanism is triggered. The current value of 1.0322 exceeds the upper limit, indicating that the coupling state between volume perfusion is undergoing significant disturbances. Therefore, the system enters the "active micro-control mode". First, the filtration rate is lowered from 16 mL / kg / h to 14 mL / kg / h again to prevent rapid blood volume extraction from inducing hypoperfusion;

[0149] Secondly, the dialysate sodium concentration is lowered from the original 132 mmol / L to 130 mmol / L, while maintaining bicarbonate at 37 mmol / L to avoid excessive alkali load; Thirdly, the system activates the "treatment rhythm switching mechanism", switching the CRRT operation mode from continuous mode to intermittent pulse filtration every 4 hours, each lasting 90 minutes, during which partial basic fluid replacement is retained to stabilize the circulatory state; Fourthly, a small dose of high molecular weight dextran is added to the fluid replacement composition to enhance the plasma colloid osmotic pressure and guide the return of interstitial fluid.

[0150] After 4 hours of adjustment, the system re-evaluates the latest parameters: C s (t) = 1.145, P d (t) = 0.215, the derivative within the sliding window The plasma volume rises to 5.4 L, the ECF drops to 15.1 L, and the ICF remains at 18.0 L. Recalculate the cosine term:

[0151]

[0152] Although A r (t) still slightly exceeds 1.0, the derivative term has changed from negative to positive, indicating a favorable trend. The system evaluates that the current intervention is initially effective and recommends maintaining the intermittent mode for 2 more rounds and then re-evaluating whether to resume continuous treatment. This example shows that the linkage adjustment function A r (t) in the present invention can accurately capture the coupling fluctuations between volume offset and perfusion efficiency, especially showing high sensitivity and trend orientation near the critical value. Combining with the cosine term periodic response structure can smooth the fluctuation prediction, dynamically trigger reasonable parameter adjustment and treatment rhythm switching mechanism, significantly improving the volume-perfusion stability management ability in acute kidney injury CRRT, and having strong clinical adaptability and prospective control advantages.

[0153] The above shows and describes 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 by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. An adjustment system for continuous renal replacement therapy applied to patients with acute kidney injury, characterized in that It includes the following steps: S1. Dynamic assessment and intervention modeling of renal microcirculation: S1.

1. Use micro sensors or nano - scale optical imaging technology to capture the blood flow velocity of glomerular capillaries and vascular endothelial permeability in real - time; through dynamic modeling of the relationship between microvascular structure and function, predict the risk time points of tissue perfusion deficiency. S1.

2. Combine microcirculation data with the patient's routine biochemical indicators to create an early intervention model; when a significant decrease in microcirculation perfusion or insufficient renal oxygen supply is detected, trigger fine - tuning of the parameters of continuous renal replacement therapy (CRRT) in advance. S2. Multidimensional toxin identification and selective clearance: S2.

1. Collect the plasma toxin composition of different patients at the clinical and molecular levels, construct a common toxin atlas for acute kidney injury (AKI); combine with real - time sensors to determine whether the concentration of specific toxins exceeds the standard, and output the clearance priority in real - time. S2.

2. Use membrane materials with adjustable pore size or special coatings to adsorb or intercept inflammatory mediators and bacterial endotoxin molecules, while maintaining permeability to small - molecule electrolytes; enable different filter materials to be switched or applied hierarchically in the same CRRT process through the system. S3. Personalized volume management and tissue perfusion coupling strategy: S3.

1. Use bioelectrical impedance or ultrasonic evaluation techniques to divide the patient's body volume into three compartments: intracellular fluid, extracellular fluid, and plasma volume; according to the changes in the compartment ratio, treat different pathological types of excess blood volume and interstitial fluid differently. S3.

2. Integrate microcirculation assessment, systemic blood pressure, and lactate value data, give tissue perfusion ratings in three levels: low, medium, and high, and moderately adjust the filtration rate or fluid replacement speed according to the ratings. S3.

3. Generate a personalized fluid replacement formula based on the volume compartment and perfusion rating, and at the same time comprehensively consider the electrolyte level and toxin accumulation rate; adopt a small - step correction strategy to avoid hemodynamic shocks caused by large - scale adjustments.

2. The adjustment system for continuous renal replacement therapy applied to patients with acute kidney injury according to claim 1, characterized in that The method for dynamic assessment and intervention modeling of renal microcirculation includes: Adopt a micro - optical sensing device integrated on a renal region - oriented probe, and through optical coherence reflection and velocity scattering techniques, obtain three key characteristic data of glomerular capillaries in the target area within the sub - second time scale: blood flow velocity represents the blood flow velocity through glomerular microvessels per unit time; endothelial permeability reflects the degree of penetration of plasma components by endothelial cells of the vascular wall; perfusion density is the density level of vascular perfusion per unit area or volume.

3. The adjustment system for continuous renal replacement therapy applied to patients with acute kidney injury according to claim 2, wherein The method for dynamic assessment and intervention modeling of renal microcirculation includes: Extract key biochemical indicators from patient monitoring data and construct a multi-parameter metabolic imbalance index F b (t) to represent the degree of deviation of the current internal environment. The formula is as follows: Among them: F b (t) The current metabolic stress score in the patient's body. The lower the value, the more deviated the metabolic state is from normal; β1, β2, β3 are weight adjustment coefficients, adjusted according to the chief complaint of the disease condition to strengthen the influence of a certain type of metabolic imbalance; B cr (t) is the serum creatinine concentration, indicating the kidney's detoxification ability; B ph (t) is the blood pH value, reflecting the acid-base balance. 7.4 is the physiological neutral reference value; B Na (t) is the current blood sodium concentration, and 140 mmol / L is the standard reference value. Through three-item normalization transformation, indicators with different dimensions are incorporated into a unified judgment framework, so that the state of high acidosis or high sodium will not be masked by a slight increase in serum creatinine.

4. The adjustment system for continuous renal replacement therapy applied to patients with acute kidney injury according to claim 3, characterized in that The method for dynamic assessment and intervention modeling of renal microcirculation includes: Overall assessment of the microcirculation state and metabolic imbalance state, construct a comprehensive risk scoring function R(t), and use it as the key reference basis for judging whether to trigger an intervention: R(t) = γ1·M(t)γ2·F b (t) Among them: The scoring result of R(t) reflects the comprehensive risk level of current renal perfusion; γ1 and γ2 respectively represent the weight factors of microcirculation and metabolic imbalance in risk decision - making; if the value of R(t) continuously decreases and is lower than the system - set intervention threshold T1, the system starts early risk warning; if it further decreases to T2 < T1, the system will automatically trigger CRRT parameter adjustment, including reducing the filtration rate, adjusting the fluid replacement components, resetting the transmembrane pressure, and adjusting the dialysis fluid buffer.

5. The adjustment system for continuous renal replacement therapy applied to patients with acute kidney injury according to claim 1, characterized in that The personalized volume management and tissue perfusion coupling strategy construction method: Through bioelectrical impedance analysis (BIA) or bedside ultrasound technology, the patient's body fluid status is dynamically modeled and the total fluid is divided into three independent compartments: intracellular fluid (ICF), extracellular fluid (ECF) and plasma volume (PV); the ratio of the three reflects the actual distribution status of the fluid in the body.

6. The adjustment system for continuous renal replacement therapy applied to patients with acute kidney injury according to claim 5, characterized in that The personalized volume management and tissue perfusion coupling strategy construction method: Integrate the microcirculation perfusion status with systemic indicators to form a perfusion level assessment system; measure v g (t) using the blood flow velocity sensor in the glomerular region, and systemic perfusion is represented by the mean arterial pressure MAP and metabolic perfusion efficiency markers such as lactate concentration Lac; the three jointly determine whether the current tissue perfusion is effective; and the perfusion efficacy index function is constructed as follows: in: P d (t) is the tissue perfusion efficacy index, which is used to judge the oxygen supply and perfusion quality at the tissue level; v g (t) is the blood flow velocity of the glomerulus per unit time, which is an important indicator reflecting the patency of microcirculation; MAP(t) is the mean arterial pressure at time t; the exponential part e -(MAP(t)-65) reflects the non-linear response of the perfusion state to blood pressure changes; Lac(t) is the blood lactate concentration, which is used to evaluate the degree of tissue hypoxia; the logarithmic function ln(1 + Lac(t) 2 ) can inhibit the influence of sudden increase in lactate on the index fluctuation; λ2 and λ3 are the weight parameters of the perfusion contribution term and the metabolic inhibition term respectively, which are set individually according to the patient's basic condition; through the structure of positive excitation with fast blood flow and high blood pressure and negative punishment with high lactate, the dynamic quantitative grading of the perfusion state is achieved.

7. The adjustment system for continuous renal replacement therapy applied to patients with acute kidney injury according to claim 6, wherein The personalized volume management and tissue perfusion coupling strategy construction method: Once the volume distribution including interstitial fluid accumulation shifts and the perfusion efficiency decreases, the system will control the CRRT parameters based on these two core indicators: dynamically adjust the filtration rate, fluid composition and treatment mode; and capture the linkage change trend and introduce the linkage adjustment trigger function: in: A r (t) is the treatment regulation activation function. The larger the value, the more drastic the interactive changes in volume and perfusion are, and the treatment strategy should be adjusted immediately. represents the instantaneous coordinated change rate between volume structure and perfusion efficiency; the second term is the cosine adjustment factor, PV(t) / (ECF(t)+ICF(t)+1) represents the proportion of plasma volume to total body fluid; π is used to control the periodicity of the cosine function; λ4 modulation coefficient is used to set the weight of the cosine correction term in the overall control; function A r When the value of (t) exceeds a certain threshold value T1 or is lower than a certain preset interval T2, the corresponding intervention logic is triggered.

Images