Biochemical index internal environment homeostasis regulation system and method

Abstract

The application discloses a biochemical index internal environment homeostasis regulation system and method, the system includes data source module, data pretreatment module, intelligent algorithm decision module, actuator drive module, feedback monitoring module, main power supply and safety redundancy module; the data pretreatment module is used for washing and fusing multidimensional data; the intelligent algorithm decision module receives the output of the data pretreatment module and is based on setting parameter, and outputs execution instruction; the actuator drive module is used for responding execution instruction; the feedback monitoring module is connected with the actuator drive module and the data source module respectively, is used for mixing after liquid parameter, user's internal environment change, system state data real-time feedback to the data source module, forms closed loop control.The application solves the problems of traditional solution regulation lag, poor precision, insufficient adaptability through real-time monitoring, accurate regulation and dynamic balance of multiple biochemical indexes such as electrolyte, acid-base, osmotic pressure, and at the same time, guarantees treatment safety through redundancy design.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, specifically to a system and method for regulating the homeostasis of biochemical indicators, which is particularly suitable for individualized and precise regulation in critically ill patients with multiple electrolyte imbalances and acid-base imbalances. Background Technology

[0002] Continuous dialysis filtration is a core technology for treating critical illnesses such as multiple organ dysfunction syndrome in the intensive care unit (ICU). Its core objective is to remove metabolic waste products from the body and maintain water, electrolyte, and acid-base balance through extracorporeal circulation. The internal environment of critically ill patients is dynamic and variable. Biochemical indicators such as serum potassium, sodium, calcium, and bicarbonate often fluctuate with the patient's condition. It is necessary to adjust the composition of the dialysis fluid / replacement fluid in real time to ensure treatment safety. There are extremely high requirements for real-time adaptation of the osmotic pressure, electrolyte concentration, and pH of the replacement fluid / dialysis fluid.

[0003] However, existing technologies present at least the following significant clinical challenges: (1) Currently, the main clinical practice is to use pre-made finished drug solutions or to manually prepare large bags of finished solutions according to empirical formulas. This mode is "offline preparation". Once the preparation is completed, the concentration of its components is fixed. However, the electrolyte and acid-base status of critically ill patients are in a highly dynamic state. The existing "fixed prescription" cannot be dynamically adjusted in real time according to the fluctuation of the patient's biochemical indicators every hour or even every minute, which can easily lead to insufficient regulation or overcorrection.

[0004] (2) During the biochemical regulation process, high concentrations of calcium and magnesium ions and bicarbonate ions are prone to chemical reaction when mixed to produce calcium carbonate and magnesium carbonate precipitates. Existing equipment usually relies on physical barriers or simple alarm shutdowns to deal with this, lacking proactive risk prediction and process optimization at the physicochemical level. Frequent precipitation and crystallization can not only lead to pipeline blockage and sensor failure, but may also cause medical accidents, which seriously limits the application of online regulation technology for high concentration mother liquor. Summary of the Invention

[0005] To address the aforementioned problems, the purpose of this invention is to propose a biochemical indicator homeostasis regulation system and method. By real-time monitoring, precise regulation, and dynamic balance of multiple biochemical indicators (electrolytes, acid-base, osmotic pressure, etc.), this system solves the problems of lagging, poor accuracy, and insufficient adaptability in traditional solution preparation. At the same time, redundant design ensures treatment safety.

[0006] This was achieved through the following technical solutions: First, a biochemical indicator homeostasis regulation system is proposed to construct a closed-loop control for drug delivery to users, including: The data source module is used to collect multi-dimensional data from users, including clinical monitoring data, online sensor data, system status data, and set parameters. The data preprocessing module, connected to the data source module, is used to clean and fuse the collected multi-dimensional data; The intelligent algorithm decision-making module receives the output of the data preprocessing module and, based on the set parameters, outputs execution instructions through deviation calculation. The actuator drive module is used to respond to execution commands and complete the delivery, mixing and output of the mixture. The mixture consists of the drug solutions delivered by the independent drug solution mother bag group and the base liquid delivery unit respectively. The feedback monitoring module is connected to both the actuator drive module and the data source module. It is used to feed back the liquid parameters of the mixture, changes in the user's internal environment, and system status data to the data source module in real time, forming a closed-loop control. Among them, changes in the internal environment correspond to clinical monitoring data, and liquid parameters of the mixture correspond to online sensor data.

[0007] Optionally, clinical monitoring data includes blood gas analysis, biochemical indicators, and vital signs; online sensor data includes ion concentration, pH value, osmotic pressure, temperature, pressure, and bubble signals; system status data includes pump speed, valve status, pipeline duration, and remaining drug volume in individual drug master bags when delivering any drug solution; and set parameters include the target concentration range and concentration threshold corresponding to the mixed solution.

[0008] Optionally, the data preprocessing module includes: an outlier removal unit, used to establish a sliding window of length N, to sample and calculate the dynamic mean and standard deviation of multi-dimensional data within the window in real time, and to identify new sampling points that meet specific preset conditions as outliers and remove them; and a fusion unit, used to perform time alignment, confidence weighting based on data reliability, and format standardization output on the multi-dimensional data after removal.

[0009] Optionally, the independent drug solution mother bag group includes multiple independently sealed mother bags, each pre-filled with a drug solution including: NaCl, KCl, calcium salt, MgCl2, NaHCO3, and glucose.

[0010] Optionally, the intelligent algorithm decision module includes a PID fuzzy control unit, configured to: calculate the output control increment u(t) of each drug pump based on the deviation e(t) = r(t) - y(t) between the real-time state value y(t) of the mixture and the target setpoint r(t). ,in, k p To set specific proportions for each component in the drug solution, k i The integral coefficient is... k d is the differential coefficient.

[0011] Optionally, the intelligent algorithm decision-making module also includes a multi-parameter coordinated adjustment unit, configured to: simultaneously calculate and adjust the ratio of parallel components corresponding to the single ion concentration based on the principles of charge neutrality and constant osmotic pressure when adjusting the concentration of a single ion. Specifically, when it is necessary to increase the HCO3⁻ concentration, the flow rate of the NaHCO3 mother liquor is increased. At the same time, the flow rate of NaCl mother liquor was reduced proportionally. This is used to maintain a constant Na⁺ concentration and keep the charge balance. The calculation formula is: ,in, and These represent the solute concentrations of the corresponding mother liquors.

[0012] Optionally, the intelligent algorithm decision-making module also includes a precipitation risk warning unit, which is configured to: monitor in real time the product of Ca²⁺ and HCO₃⁻ or the product of Mg²⁺ and HCO₃⁻ in the mixture; when the product exceeds the preset corresponding concentration threshold, automatically execute an anti-precipitation command; the anti-precipitation command includes at least one of adjusting the delivery sequence of each drug solution, suspending the delivery of bicarbonate drug solution, performing pre-dilution of the base solution, or increasing the stirring speed.

[0013] Optionally, in the actuator drive module, the base fluid delivery unit includes two independent pipelines for delivering sterile pure water and physiological saline, respectively; the precision delivery unit uses a peristaltic pump I or a plunger pump group, and both the drug mother liquor and the base fluid are set with their respective corresponding flow rate adjustment ranges; the delivery pump unit uses a peristaltic pump II to draw any qualified liquid from the intelligent mixing chamber and output it.

[0014] Optionally, the intelligent mixing chamber includes: a chamber shell, a top cover, multiple liquid inlets and a bottom liquid outlet; a central shaft, a stirring impeller, and a spiral heating element for heating, all located at the bottom of the chamber; and multiple sensors, located on the chamber shell and communicating with the interior of the chamber, including an ion sensor, a pH sensor, and a temperature sensor.

[0015] Optionally, it also includes a safety redundancy module, which includes: a graded alarm unit for triggering an early warning when the remaining liquid is insufficient, the pipeline timeout occurs, or the concentration exceeds the limit; and a backup module, including a pre-installed backup finished liquid and a backup power supply.

[0016] Secondly, a method for regulating the homeostasis of biochemical indicators is proposed, which uses the aforementioned system for regulating the homeostasis of biochemical indicators and includes the following steps: S1. Real-time collection of patient biochemical indicators, online sensor data, and set parameters; S2. Clean and merge the collected data to generate standardized state values; S3. Calculate the deviation between the actual parameters and the set parameters, generate the execution instructions for each pump group through the PID control algorithm, and perform sedimentation risk warning and multi-parameter collaborative calculation. S4. Responding to the execution command, the drug solution and base solution are sent to the intelligent mixing chamber, mixed, and then output. S5. Collect feedback data in real time and return it to step S1 for dynamic adjustment of execution instructions; S6: Monitors the status of the entire process system, and provides tiered warnings, emergency shutdowns, or switches to backup power and finished liquid in case of abnormal situations.

[0017] The beneficial effects of this invention compared to the prior art are: (1) This invention constructs an integrated architecture of multi-component independent mother liquor and intelligent mixing chamber, and uses multi-parameter sensors integrated in the intelligent mixing chamber to form an ultra-fast response inner loop feedback loop. Compared with the traditional mode of "preset prescription - static configuration of whole bag of liquid" which requires waiting for the whole bag to be used up before adjustment, this system realizes the seamless connection of "real-time monitoring - real-time adjustment", which greatly reduces the risk of complications such as arrhythmia caused by continuous deviation of internal environment in critically ill patients.

[0018] (2) This system breaks through the limitation of traditional equipment that can only make coarse adjustments to a single variable. It innovatively adopts a multi-parameter collaborative algorithm based on physical and chemical constraints. When adjusting the concentration of a single ion (such as correcting acidosis), the system can automatically decouple and synchronously adjust the related components according to the principle of charge neutrality and constant osmotic pressure. This achieves smooth and coordinated regulation of multiple biochemical indicators and effectively avoids the derivative medical risks of systemic osmotic pressure imbalance caused by the adjustment of a single indicator.

[0019] (3) In response to the clinical pain point that calcium and magnesium ions are prone to chemical precipitation with bicarbonate ions in continuous dialysis filtration (CRRT), this system introduces a dynamic ion product real-time monitoring mechanism. When the system predicts a high risk of precipitation, it actively implements a precipitation prevention strategy of pre-dilution of the base solution and increasing the stirring speed. This intervention from the dual dimensions of fluid dynamics and chemical solubility upgrades the traditional "passive alarm shutdown" to "active prevention", which greatly improves the stability of the system during long-term operation.

[0020] (4) To address the issue that online sensors are susceptible to interference from air bubbles or physical drift in intensive care environments, this invention designs a dynamic confidence assessment mechanism that integrates clinical benchmarks and online data in the preprocessing module. By eliminating transient false alarms and using offline clinical biochemical results to verify the online data exceeding limits in real time, it effectively avoids erroneous solution preparation instructions caused by "pseudo-abnormal" signals, providing highly reliable data assurance for closed-loop control. Attached Figure Description

[0021] Figure 1A schematic diagram of a biochemical indicator homeostasis regulation system; Figure 2 A flowchart of a method for regulating the homeostasis of biochemical indicators in the internal environment; Figure 3 This is a schematic diagram of a mixing cavity.

[0022] Figure label: 100-Top cover, 101-Liquid inlet, 102-Mixing chamber, 103-Central shaft, 104-Agitator impeller, 105-Spiral heating element, 106-Ion sensor, 107-pH sensor, 108-Temperature sensor, 109-Outer shell, 110-Bottom outlet. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0024] like Figure 1 The diagram shows a framework of a biochemical indicator homeostasis control system. This system is used to construct a closed-loop control for drug delivery to the user, including a data source module, a data preprocessing module, an intelligent algorithm decision-making module, an actuator drive module, a feedback monitoring module, a main power supply, and a safety redundancy module. The main power supply provides power to each module. Through closed-loop feedback control between the modules, error correction is achieved. The system continuously subtracts the patient's biochemical indicator values ​​from the target parameters set by the doctor to obtain the error. The PID algorithm calculates and generates control commands in real time based on this dynamically changing error, determining whether the pump should increase or decrease pressure. Furthermore, after changing the mixture ratio, the patient's biochemical indicators will fluctuate accordingly. This fluctuation is fed back to the system, forming another feedback loop, which is then used for further fine-tuning.

[0025] In this embodiment, the data source module is used to collect multi-dimensional data from the user. This multi-dimensional data includes clinical monitoring data (blood gas analysis, biochemical indicators, vital signs), online sensor data (ion concentration, pH, osmotic pressure, temperature, pressure, bubble signal), system status data (pump speed, valve status, tubing duration, and remaining drug volume in individual drug supply bags when delivering any type of drug solution), and set parameters (target concentration range, concentration threshold, adjustment rate threshold, and treatment mode for the mixed solution). The treatment modes in the set parameters include CVVH (Continuous Venous-Venous Hemofiltration), CVVHD (Continuous Venous-Venous Hemodialysis), and CVVHDF (Continuous Venous-Venous Hemodialysis), etc., which can be set by the user.

[0026] It should be noted that when this system is adjusted, the concentrations of various parameters of the final mixture (i.e., dialysate / replacement fluid) are based on the following table 1: Table 1:

[0027] As shown in Table 1 above, the core requirements are that the osmotic pressure of the replacement fluid and dialysate should be close to that of plasma (280–300 mOsm / L) to avoid cell edema or dehydration; electrolyte concentration should match physiological levels to quickly correct imbalances; acid-base balance should be maintained at pH 7.35–7.45 with target bicarbonate concentration; and blood glucose should be stabilized to avoid hyperglycemia or hypoglycemia, especially suitable for patients with diabetes and sepsis.

[0028] The data source module is equipped with multiple sensors, including an ion sensor, a pH sensor, a temperature sensor, and a bubble sensor. Each sensor is used to transmit online sensor data. Each sensor is installed on the outer shell of the intelligent mixing chamber and connected to the mixing chamber, thereby acquiring data of the mixed liquid in the mixing chamber and transmitting it to the software system (such as the LIS system commonly used in hospitals).

[0029] The data preprocessing module connects to the data source module and is used to clean and fuse the collected multi-dimensional data. The data preprocessing module includes an outlier removal unit and a fusion unit. The outlier removal unit establishes a sliding window of length N, samples and calculates the dynamic mean μ and standard deviation σ of the multi-dimensional data within the window in real time, and selects new sampling points x that meet specific preset conditions. new Identify and remove outliers (e.g.) The fusion unit performs time alignment, confidence weighting based on data reliability, and format standardization (unifying the output format) on the multi-dimensional data after data removal. Time alignment eliminates time-series bias by synchronizing data with different sampling frequencies (e.g., second-level sensor results versus minute-level clinical results) through interpolation. Confidence weighting dynamically allocates weights based on data source reliability (e.g., when sodium concentration sensor values ​​exceed the 130-150 mmol / L range, clinical test results are introduced for confidence comparison, and sensor calibration prompts are triggered when the deviation exceeds 5%) to suppress drift or bubble interference. The unified output format encapsulates the aligned and weighted data into a standardized vector with timestamps, source labels, and confidence scores for direct use by intelligent algorithms.

[0030] The intelligent algorithm decision-making module receives the output from the data preprocessing module and, based on the set parameters, outputs execution instructions through deviation calculation, PID fuzzy control, multi-parameter coordinated adjustment, and sedimentation risk early warning. The intelligent algorithm decision-making module includes a PID fuzzy control unit, a sedimentation risk early warning unit, and a multi-parameter coordinated adjustment unit.

[0031] In the PID fuzzy control unit, the system can determine the real-time state value corresponding to multi-dimensional data. y(t) Compared with the target set value r(t)Calculate the deviation e(t) = r(t) - y(t) PID control is used to calculate the output control increment of each medicine pump. u(t) : .

[0032] In the above formula, u(t) To control the increment, k p To set specific proportions for different components of the drug solution, such as the proportion of electrolytes to metal ions. k p =2.0, HCO3⁻ proportional system k p =1.5, glucose ratio coefficient k p =1.0, k i The integral coefficient is... k d is the differential coefficient. k i The value ranges from 0.1 to 0.5 min⁻¹, and it is used to eliminate the steady-state error of the system; k d The value ranges from 0.01 to 0.1 min and is used to suppress solubility fluctuations and overshoot.

[0033] Table 2 below shows the closed-loop control and related parameter data and their corresponding functions: Table 2:

[0034] In this embodiment, based on the calculation of the PID fuzzy control unit, incremental control logic is used in conjunction with the proportional coefficient of specific components to ensure that the adjustment of different biochemical indicators is both rapid and stable, effectively avoiding the risk of integral saturation that is easily generated by positional PID, and greatly improving the safety of critical care.

[0035] For the precipitation risk warning unit, when Ca is detected in the mixture... 2+ With HCO3 - The product of Mg²⁺ and HCO₃ - When the product of the two concentrations exceeds the preset corresponding concentration threshold, the system will automatically adjust the drug delivery sequence or suspend the delivery of bicarbonate mother liquor.

[0036] Table 3 below shows the relevant data for monitoring indicators and adjustment thresholds: Table 3:

[0037] In the system's settings, several threshold data reference settings are as follows: blood gas analysis (pH 7.35-7.45, HCO3⁻ 20-40 mmol / L, BE±3 mmol / L), biochemical indicators (Na⁺ 130-150 mmol / L, K⁺ 3.5-5.5 mmol / L, Ca²⁺ 0.8-1.6 mmol / L, Mg²⁺ 0.4-1.0 mmol / L, blood glucose 3.9-6.1 mmol / L), and vital signs (blood pressure 90-140 / 60-90 mmHg, heart rate 60-100 beats / min, blood oxygen ≥95%). These threshold data are used to ensure the patient's physiological safety. (It should be noted that the upper and lower limits of the thresholds can be slightly adjusted compared to the standards in the corresponding tables.) In this embodiment, when Ca 2+ Concentration > 1.6 mmol / L and HCO3 - Concentration > 40 mmol / L, or Mg²⁺ concentration > 1.0 mmol / L and HCO₃ - When the concentration is >40 mmol / L, the system will automatically adjust the drug delivery sequence or pause HCO3 administration. - The instructions for mother liquor delivery, including adjusting the order of addition, involve first pre-dilution of the mixing chamber with the base liquid and increasing the stirring speed, followed by the delivery of bicarbonate. By proactively implementing physicochemical interventions such as base liquid pre-dilution, adjusting the dosing sequence, and increasing the stirring speed, the system mitigates risks before crystallization occurs, significantly improving the stability of the system during long-term operation.

[0038] In the multi-parameter coordinated adjustment unit, based on the principles of charge neutrality and constant osmotic pressure, it is used to simultaneously calculate and adjust the ratio of parallel components corresponding to any single ion concentration when adjusting the concentration of any single ion. This is to maintain the total osmotic pressure of the mixture within a set range and maintain charge balance. For example, when it is necessary to increase the HCO3⁻ concentration, the required increase in the NaHCO3 mother liquor flow rate is calculated. Simultaneously calculate the required reduction in NaCl solution flow rate in the base solution. Simultaneously, the Cl⁻ concentration is reduced to maintain a constant Na⁺ concentration, preserve charge balance, and stabilize osmotic pressure (e.g., 270-310 mOsm / L). Taking the synergistic effect of HCO₃⁻ and Cl⁻ as an example, when a patient's acid-base balance shifts, it is necessary to increase HCO₃⁻ concentration. - At a given concentration, the algorithm performs the following cooperative decoupling calculations: First, based on HCO3... - deviation e Bic The required increase in NaHCO3 mother liquor flow rate was calculated using a PID algorithm. Then the system recognized An increase in Na⁺ concentration leads to a rise in the total osmotic pressure of the mixed solution. To maintain a constant Na⁺ concentration and keep the charge balanced, the algorithm automatically solves the compensation matrix. Based on the principle of equal replacement of anions, the system calculates the required reduction in the NaCl mother liquor flow rate. The calculation formula is: .

[0039] In the above formula, and These represent the solute concentrations of the two mother liquors. Furthermore, if the total osmotic pressure fluctuation after substitution exceeds ±2% of the target value, the flow rate of the base solution (pure water) will be further adjusted for volume compensation.

[0040] The actuator drive module includes an independent drug solution mother bag assembly, a base liquid delivery unit, a precision delivery unit, an intelligent mixing chamber, and a delivery pump unit. It responds to execution commands and accordingly completes the delivery, mixing, and output of the mixed solution. The mixed solution consists of drugs delivered separately by the independent drug solution mother bag assembly and the base liquid delivery unit. Each mother bag is connected to the precision delivery unit via pharmaceutical-grade silicone tubing, which in turn connects to the inlet of the intelligent mixing chamber. The independent drug solution mother bag assembly includes multiple independently packaged mother bags, each pre-filled with a drug solution including: NaCl, KCl, calcium salts, MgCl2, NaHCO3, and glucose.

[0041] In the actuator drive module, the base fluid delivery unit contains two independent pipelines, delivering sterile pure water (resistivity ≥18.2 MΩ·cm, sterile and pyrogen-free) and physiological saline (Na⁺ 154 mmol / L, Cl⁻ 154 mmol / L) respectively, adapting to different dilution requirements. The precision delivery unit uses a peristaltic pump I or a plunger pump assembly, with corresponding flow rate adjustment ranges set for both the drug stock solution and the base fluid. For example, the flow rate adjustment range for the drug stock solution is 0-500 mL / h (accuracy ±0.1 mL / min), and the flow rate adjustment range for the base fluid is 500-4000 mL / h (accuracy ±1 mL / min). The delivery pump unit uses a peristaltic pump II to aspirate any qualified liquid from the intelligent mixing chamber and output it (i.e., output the final replacement fluid / dialysis fluid).

[0042] The feedback monitoring module, connected to both the actuator drive module and the data source module, feeds back the liquid parameters of the mixed solution, changes in the user's internal environment, and system status data to the data source module in real time, forming a closed-loop control. Changes in the internal environment correspond to clinical monitoring data, while the liquid parameters of the mixed solution correspond to online sensor data. After a patient receives treatment with the prepared mixed solution or the finished solution, changes in blood gas analysis and biochemical indicators are fed back to the data source module through the clinical monitoring system. The system's operating status (pump speed, pressure, and remaining drug volume) is simultaneously fed back, and the intelligent algorithm decision-making module dynamically adjusts the execution commands based on the feedback data, thus forming a closed-loop control.

[0043] Table 4 below shows the alarm category settings and functions in the safety redundancy module: Table 4:

[0044] As shown in Table 4 above, the safety redundancy module is connected to the intelligent algorithm decision-making module. The safety redundancy module includes a tiered alarm unit, a backup finished solution unit, and a backup power supply unit, which are used sequentially for abnormal situation warnings (insufficient drug solution remaining volume, tubing timeout, concentration exceeding limits, etc.), switching to backup mode, and backup power supply. For abnormal situation warnings, such as when any remaining stock solution is <10%, or when tubing usage exceeds a preset time, or when abnormal concentrations (Na⁺ > 150 or < 130 mmol / L, K⁺ > 5.5 or < 3.5 mmol / L, pH < 7.25 or > 7.45), a warning alarm is triggered and stock solution adjustment is suspended; abnormal pressure (> 300 or < 0 mmHg) or bubble detection triggers an emergency shutdown, closing the intravenous tubing. Switching to backup mode involves using pre-stored finished solution; in case of mixing errors in the intelligent mixing chamber or insufficient stock solution remaining, the pre-stored finished solution can be directly used to treat the patient. Backup power supply refers to the power supply provided to each module when the main power supply fails, so as to ensure that the system can operate normally.

[0045] like Figure 3 The diagram shows a schematic of a mixing chamber. The intelligent mixing chamber is cylindrical and includes: a chamber shell, a top cover, multiple inlets, and a bottom outlet; a central shaft at the bottom of the chamber, a stirring impeller, and a spiral heating element on the inner wall of the chamber for heating; and multiple sensors mounted on the chamber shell and communicating with the interior of the chamber, including an ion sensor, a pH sensor, and a temperature sensor (used to transmit corresponding data to the data source module). The top cover covers the top of the shell; the multiple inlets are for inputting mother solution or base solution; the central shaft is located at the center of the stirring impeller and fixed to the bottom center of the intelligent mixing chamber; and the bottom outlet is for outputting to the patient.

[0046] like Figure 2 The diagram shows a flowchart of a method for regulating the homeostasis of biochemical indicators within the internal environment. This method utilizes the aforementioned system for regulating the homeostasis of biochemical indicators within the internal environment and includes the following steps: S1. Real-time acquisition of patient biochemical indicators, online sensor data, and set parameters. Online sensor data includes ion concentration, pH value, osmotic pressure, temperature, pressure, and bubble signals; system status data includes pump speed, valve status, pipeline duration, and remaining drug volume in individual drug master bags when delivering any drug solution; set parameters include the target concentration range and concentration threshold for the mixed solution.

[0047] S2. Clean and fuse the collected data to generate standardized status values. During cleaning, outliers are removed to ensure the accuracy and validity of the data. During fusion, the cleaned data is time-aligned, then weighted based on the confidence level of data reliability and output in a standardized format.

[0048] S3. Calculate the deviation between actual and set parameters, generate execution instructions for each pump group using a PID control algorithm, and perform precipitation risk warning and multi-parameter collaborative calculation. In this step, mandatory adjustment rate constraints are set for different biochemical indicators, where Na⁺ adjustment rate ≤ 10 mmol / L / 24h, K⁺ adjustment rate ≤ 1.0 mmol / L / h, and HCO₃⁻ adjustment rate ≤ 1.0 mmol / L / h. - Adjustment rate ≤ 5 mmol / L / h.

[0049] S4. Each pump unit delivers the medicine and base liquid to the intelligent mixing chamber according to the instructions. In the intelligent mixing chamber, the liquid is heated by the spiral heating element and stirred by the stirring impeller before being output.

[0050] S5, the feedback monitoring module collects mixed fluid parameters, changes in the patient's internal environment, and system status data in real time. All the real-time collected data is fed back to the data source module, enabling the intelligent algorithm decision-making module to adjust the execution instructions based on this real-time feedback data, and to control the pump speed, etc., thereby forming an effective closed-loop control.

[0051] S6. Monitor the entire process system status, and issue tiered warnings, emergency shutdowns, or switch to backup power and liquids in case of abnormalities. If Na⁺, K⁺, or pH values ​​exceed the safe preset range, a warning alarm is triggered and mother liquor adjustment is suspended. If abnormal pipeline pressure or bubbles are detected, an emergency shutdown is triggered and the intravenous end pipeline is closed. If a power failure is detected, the system switches to UPS backup power. The system architecture and beneficial effects of this method can be found in the description of the aforementioned control system, and will not be repeated here.

[0052] For example, taking a critically ill patient with hyperkalemia (serum potassium 5.8 mmol / L) as an example, the implementation process of this method is as follows: 1) Data collection: The data source module collects the patient's biochemical indicators (serum potassium 5.8 mmol / L), other electrolyte concentrations, and blood gas analysis results. At the same time, the doctor enters the target parameters (serum potassium 3.5-5.0 mmol / L, sodium 138-140 mmol / L), safety threshold (K⁺ regulation rate ≤1.0 mmol / L / h) and treatment mode (CVVHDF). 2) Data preprocessing: Remove outlier data, merge and standardize multi-source data, and output reliable data; 3) Algorithm decision: The deviation calculation unit obtains a blood potassium deviation of 0.8 mmol / L, the PID + fuzzy control unit sets the adaptation parameters, the multi-parameter collaborative adjustment unit plans the adjustment scheme (pause the delivery of KCl mother liquor and appropriately increase the flow rate of the base liquid), and the precipitation risk warning unit monitors in real time that there is no precipitation risk and outputs the execution command; 4) Precise execution: The actuator drive module responds to the command, stops the KCl mother liquor transfer pump, increases the flow rate of sterile pure water to 1200 mL / h, and transfers other drug solutions to the intelligent mixing chamber in proportion. After heating (37℃) and stirring, the solution is output to the CRRT host by the transfer pump unit. 5) Real-time feedback: The mixing chamber sensor detected a gradual decrease in potassium concentration. Subsequent biochemical tests showed that the patient's serum potassium was 5.2 mmol / L. This information was fed back to the data source module, and the algorithm decision module adjusted the basal fluid flow rate to 1000 mL / h. 6) Safety assurance: No abnormalities throughout the process, and the safety redundancy module is in standby mode; if the pipeline pressure is detected to rise to 320 mmHg, an emergency shutdown will be triggered immediately, and the system will be switched to the backup finished liquid system.

[0053] In summary, this invention constructs an integrated architecture of multi-component independent mother liquor and intelligent mixing chamber, integrating multi-parameter sensors to form a rapidly responding inner loop feedback, achieving seamless connection between "real-time monitoring and real-time adjustment," significantly reducing the risk of complications such as arrhythmia caused by continuous deviation of the internal environment in critically ill patients. Secondly, it innovatively adopts a multi-parameter collaborative algorithm based on physicochemical constraints, adhering to the principles of charge neutrality and constant osmotic pressure, automatically decoupling and synchronously adjusting related components when adjusting a single ion, achieving smooth and coordinated regulation of multiple biochemical indicators, effectively avoiding the derivative risks of systemic osmotic pressure imbalance caused by the adjustment of a single indicator. Furthermore, it addresses the issue that calcium and magnesium ions easily chemically precipitate with bicarbonate ions during continuous dialysis filtration. To address the pain points, dynamic ion product real-time monitoring is introduced. When a high risk is predicted, the system proactively pre-dilutes the base solution and increases the stirring speed. This intervention, based on both fluid dynamics and chemical solubility, upgrades the traditional "passive alarm shutdown" to "proactive prevention," significantly improving the stability of the system during long-term operation. Furthermore, to address the issue of online sensors being susceptible to interference from bubbles or physical drift in intensive care environments, a dynamic confidence assessment mechanism integrating clinical benchmarks and online data is designed in the preprocessing module. By eliminating transient false alarms and using offline clinical biochemical results to verify out-of-limit data in real time, this effectively avoids erroneous solution preparation instructions caused by false abnormal signals, providing highly reliable data assurance for closed-loop control, demonstrating significant progress.

[0054] The above embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solutions based on the technical concept proposed in this invention shall fall within the scope of protection of this invention.

Claims

1. A biochemical indicator homeostasis control system for constructing a closed-loop control system for drug delivery to users, characterized in that, include; The data source module is used to collect multi-dimensional data from users, including clinical monitoring data, online sensor data, system status data, and set parameters. The data preprocessing module, connected to the data source module, is used to clean and fuse the collected multi-dimensional data; The intelligent algorithm decision-making module receives the output of the data preprocessing module and, based on set parameters, outputs an execution command through deviation calculation. The actuator drive module is used to respond to execution commands and complete the delivery, mixing and output of the mixture, which consists of the drug solutions delivered by the independent drug solution mother bag group and the base liquid delivery unit respectively; The feedback monitoring module is connected to both the actuator drive module and the data source module. It is used to feed back the liquid parameters of the mixture, changes in the user's internal environment, and system status data to the data source module in real time, forming a closed-loop control. The changes in the internal environment correspond to clinical monitoring data, and the liquid parameters of the mixture correspond to online sensor data.

2. The biochemical indicator homeostasis regulation system according to claim 1, characterized in that, The clinical monitoring data includes blood gas analysis, biochemical indicators, and vital signs; the online sensor data includes ion concentration, pH value, osmotic pressure, temperature, pressure, and bubble signal; the system status data includes pump speed, valve status, pipeline duration, and remaining drug volume in the independent drug mother bag group when delivering any drug solution; the set parameters include the target concentration range and concentration threshold corresponding to the mixed solution.

3. The biochemical indicator homeostasis regulation system according to claim 1, characterized in that, The data preprocessing module includes: The outlier removal unit is used to establish a sliding window of length N, sample and calculate the dynamic mean and standard deviation of multi-dimensional data within the window in real time, and identify new sampling points that meet specific preset conditions as outliers and remove them. The fusion unit is used to perform time alignment, confidence weighting based on data reliability, and format standardization output on the multi-dimensional data after the removal process.

4. The biochemical indicator homeostasis regulation system according to claim 1, characterized in that, The independent drug solution mother bag group includes multiple independently packaged mother bags, each pre-filled with a drug solution including: NaCl, KCl, calcium salt, MgCl2, NaHCO3, and glucose.

5. The biochemical indicator homeostasis regulation system according to claim 1, characterized in that, The intelligent algorithm decision module includes a PID fuzzy control unit, configured to: calculate the output control increment u(t) of each drug pump based on the deviation e(t) = r(t) - y(t) between the real-time state value y(t) of the mixture and the target setpoint r(t). ,in, k p To set specific proportions for each component in the drug solution, k i The integral coefficient is... k d is the differential coefficient.

6. The biochemical indicator homeostasis regulation system according to claim 4, characterized in that, The intelligent algorithm decision-making module also includes a multi-parameter collaborative adjustment unit, configured as follows: When adjusting the concentration of a single ion, the ratio of parallel components corresponding to the single ion concentration is calculated and adjusted simultaneously based on the principles of charge neutrality and constant osmotic pressure. Specifically, when it is necessary to increase the HCO3⁻ concentration, the flow rate of the NaHCO3 mother liquor is increased. At the same time, the flow rate of NaCl mother liquor was reduced proportionally. This is used to maintain a constant Na⁺ concentration and keep the charge balance. The calculation formula is: ,in, and These represent the solute concentrations of the corresponding mother liquors.

7. The biochemical indicator homeostasis regulation system according to claim 4, characterized in that, The intelligent algorithm decision-making module also includes a precipitation risk early warning unit, which is configured to: monitor in real time the product of Ca²⁺ and HCO₃⁻ or the product of Mg²⁺ and HCO₃⁻ in the mixture; when the product exceeds the preset corresponding concentration threshold, automatically execute an anti-precipitation command; the anti-precipitation command includes at least one of adjusting the delivery order of each drug solution, pausing the delivery of bicarbonate drug solution, performing pre-dilution of the base solution, or increasing the stirring speed.

8. The biochemical indicator homeostasis regulation system according to claim 1, characterized in that, In the actuator drive module, the basic fluid delivery unit contains two independent pipelines, which respectively deliver sterile pure water and physiological saline; The precision delivery unit uses a peristaltic pump I or a plunger pump group, and both the mother liquor and the base liquor are equipped with their own corresponding flow rate adjustment ranges. The delivery pump unit uses a peristaltic pump II to draw in any qualified liquid from the intelligent mixing chamber and output it.

9. The biochemical indicator homeostasis regulation system according to claim 1, characterized in that, The intelligent mixing chamber includes: a chamber shell, a top cover, multiple liquid inlets, and a bottom liquid outlet; A central shaft, a stirring impeller, and a spiral heating element for heating are located at the bottom of the cavity; Multiple sensors are mounted on the outer shell of the cavity and communicate with the inside of the cavity, including an ion sensor, a pH sensor, and a temperature sensor.

10. The biochemical indicator homeostasis regulation system according to claim 1, characterized in that, It also includes a security redundancy module, which includes: The graded alarm unit is used to trigger an early warning when the remaining liquid is insufficient, the pipeline timeout is exceeded, or the concentration exceeds the limit. The backup module includes a pre-installed backup finished liquid and a backup power supply.

11. A method for regulating the homeostasis of biochemical indicators, comprising operating a homeostasis regulation system for biochemical indicators as described in any one of claims 1 to 10, characterized in that, Includes the following steps: S1. Real-time collection of patient biochemical indicators, online sensor data, and set parameters; S2. Clean and merge the collected data to generate standardized state values; S3. Calculate the deviation between the actual parameters and the set parameters, generate the execution instructions for each pump group through the PID control algorithm, and perform sedimentation risk warning and multi-parameter collaborative calculation. S4. In response to the execution command, the drug solution and the base solution are sent to the intelligent mixing chamber, mixed, and then output. S5. Collect feedback data in real time and return it to step S1 for dynamic adjustment of execution instructions; S6: Monitors the status of the entire process system, and provides tiered warnings, emergency shutdowns, or switches to backup power and finished liquid in case of abnormal situations.

Images