Intelligent intravenous hypotensive drug administration control system based on target blood pressure

Abstract

The present application relates to a kind of intelligent intravenous antihypertensive drug dosing control systems based on target blood pressure, it includes: fluid interface unit, attitude sensing unit and control terminal host, control terminal host can be received signal analysis, generate the control instruction for adjusting the action of external infusion pump;Wherein, the control logic executed by control terminal host includes: based on optical characteristic signal determination whether there is drug compatibility contraindicated deposit in infusion pipeline, and when judging that there is deposit, trigger the blocking instruction for external infusion pump;Based on body position change signal and pipeline pressure signal, distinguish and filter out the blood pressure fluctuation interference caused by body position change in external blood pressure signal, activate trigger " high frequency monitoring observation window ", to dynamically adjust dosing control instruction;Based on sustained dosing history data and external blood pressure signal determination drug accumulation risk, and when evaluating that there is steady-state accumulation or delayed blood pressure sudden drop trend, advance output preventive speed limit instruction.

Description

Technical Field

[0001] This invention relates to the field of medical devices and automated drug delivery control technology, and in particular to an intelligent intravenous antihypertensive drug delivery regulation system based on target blood pressure. Background Technology

[0002] Cerebral hyperperfusion syndrome is one of the more serious and dangerous complications following carotid endarterectomy, carotid artery stenting, and endovascular treatment of intracranial atherosclerotic stenosis. If not detected and treated promptly, it can easily lead to cerebral hemorrhage or even death. Appropriate antihypertensive therapy after surgery is very effective in preventing cerebral hyperperfusion syndrome. For example, for patients undergoing endovascular treatment of the anterior circulation, postoperative blood pressure usually needs to be strictly controlled within the range of 100-120 mmHg. However, patients with cerebral vascular stenosis often have concurrent cardiac vascular stenosis, and low blood pressure can lead to insufficient myocardial perfusion, resulting in myocardial ischemia, injury, or even infarction. Therefore, stabilizing blood pressure within the target range and avoiding excessively high or low blood pressure is particularly crucial during intravenous antihypertensive drug infusion via a microinfusion pump. Urapidil, as an intravenous antihypertensive drug with both central and peripheral antihypertensive mechanisms, is widely used in the emergency department and perioperative period. Studies on its clinical administration methods, such as those by Gao Xin et al. in the article "The Influence of Different Administration Methods of Urapidil on Controlled Hypotension During Deep Brain Stimulation" (see Gao Xin, Li Jingsheng, Fu Huiqun, et al. The Influence of Different Administration Methods of Urapidil on Controlled Hypotension During Deep Brain Stimulation [J]. Foreign Medical Sciences: Anesthesiology and Resuscitation, 2022(4).), published that for Parkinson's disease patients undergoing deep brain stimulation, those using continuous infusion of urapidil for controlled hypotension had more stable mean arterial pressure and a lower incidence of perioperative hypertension and hypotension. This study fully confirms the great clinical value of continuous infusion, but this process relies on close monitoring by anesthesiologists and real-time manual adjustment of the drug infusion rate, without the application of an intelligent automated drug administration control system.

[0003] Current automated drug delivery control systems still have significant limitations. For example, CN117414497A discloses a blood pressure control method and system based on artificial intelligence deep learning. This system acquires blood pressure information in real time by obtaining infusion parameters from a micro-infusion pump and the patient's vital signs. It constructs a regression curve of blood pressure control based on the infusion rate and blood pressure information from a previous moment, and substitutes the current blood pressure information into this curve to obtain an updated infusion rate, thereby adjusting the micro-infusion pump to lower the patient's blood pressure. This patent application also discloses control logic that outputs an alarm signal when the blood pressure drop exceeds a set threshold or the rate of drop is too large. However, this existing technology has the following serious safety hazards and problems when dealing with complex and vulnerable neurocritical care clinical scenarios: First, the system in CN117414497A only focuses on the physiological parameters and infusion rate input into the patient, completely neglecting to monitor the physicochemical properties of the medication within the infusion tubing. In actual clinical practice, urapidil injection is acidic; if it accidentally comes into contact with alkaline liquids such as sodium bicarbonate in the infusion tubing, it will rapidly produce a white flocculent precipitate, leading to severe microvascular embolism. Existing infusion systems can only control the mechanical flow rate and cannot sense the physicochemical properties of the medication. The prevention of drug incompatibilities relies entirely on visual observation by medical staff, creating a significant blind spot in physical drug administration.

[0004] Secondly, while CN117414497A mentions acquiring various vital signs parameters such as patient respiration and intracranial pressure, it completely fails to consider the dual impact of patient positional changes on hemodynamics and the physical state of infusion tubing. In neurocritical care, patients frequently need to change position (e.g., from supine to sitting). Gravity causes not only a redistribution of blood, leading to transient physiological "blood pressure fluctuations," but also changes in trunk height inevitably alter the hydrostatic pressure of the fluid column within the infusion tubing connected to the vein. Current technology not only fails to sense the patient's spatial posture but also lacks real-time monitoring of minute hydrodynamic changes (hydrostatic pressure and tubing resistance) within the infusion tubing. In clinical practice, relying solely on external monitors for blood pressure fluctuations can easily misinterpret physiological fluctuations caused by positional changes as abnormal drug efficacy; even with a single surface motion sensor, false triggering often occurs due to localized chest twitching or turning disturbances. Existing technologies generally lack a cross-monitoring mechanism that cross-validates "body surface spatial motion" and "pipeline fluid dynamics (hydrostatic pressure)" in multiple dimensions, which can easily lead to the issuance of erroneous flow rate adjustment commands, significantly increasing the risk of stroke or iatrogenic hypotension in patients.

[0005] Finally, the control logic of CN117414497A relies heavily on data from a very short historical period to construct regression curves for feedback regulation. Clinical practice shows that drugs like urapidil are relatively safe in the initial stages of continuous infusion, but as the administration time increases (e.g., after 2-3 hours), the blood drug concentration gradually reaches a steady-state peak, resulting in cumulative effects. Existing infusion systems cannot calculate the dynamic "drug accumulation." At times when the patient's vascular sensitivity to the drug significantly increases (especially during postoperative nighttime sleep when vagal nerve excitation leads to a natural physiological drop in blood pressure), if the control system rigidly maintains the original pump rate or short-term regression prediction, it can easily cause a rapid drop in the patient's blood pressure within a short period, triggering severe or even fatal iatrogenic hypotensive shock.

[0006] Furthermore, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making this invention, but due to space limitations, not all details and contents were listed in detail. However, this does not mean that the present invention does not possess the features of these prior art. On the contrary, the present invention already possesses all the features of the prior art, and the applicant reserves the right to add relevant prior art to the background art. Summary of the Invention

[0007] In view of the shortcomings of the prior art, the present invention provides an intelligent intravenous antihypertensive drug administration regulation system based on target blood pressure to solve at least some of the above-mentioned technical problems.

[0008] This invention discloses an intelligent intravenous antihypertensive drug administration regulation system based on target blood pressure, comprising: a fluid interface unit configured in an infusion line for acquiring optical characteristic signals of the drug solution flowing through it and pressure signals within the infusion line; a posture sensing unit for acquiring patient position change signals; and a control terminal host, communicatively connected to the fluid interface unit and the posture sensing unit, and equipped with a communication interface for receiving external blood pressure signals; the control terminal host is configured to generate a regulation system for adjusting the external infusion based on the fusion analysis of the received optical characteristic signals, line pressure signals, position change signals, and external blood pressure signals. The control commands for pump operation include: determining whether there is drug incompatibility precipitation in the infusion tubing based on optical feature signals, and triggering a blocking command for the external infusion pump when precipitation is detected; identifying and filtering blood pressure fluctuation interference caused by body position changes in the external blood pressure signal based on body position change signals and tubing pressure signals, and activating the "high-frequency monitoring observation window" to dynamically adjust the drug administration control commands; determining the risk of drug accumulation based on continuous drug administration history data and external blood pressure signals, and outputting a preventive rate limiting command in advance when assessing the presence of steady-state accumulation or a delayed blood pressure drop trend.

[0009] Furthermore, the external blood pressure signal can originate from a non-invasive blood pressure (NIBP) monitoring device or an invasive arterial blood pressure (IBP) monitoring device. In clinical practice, when a patient changes position, if NIBP monitoring is used, gravity causing blood redistribution can trigger transient physiological blood pressure fluctuations; if IBP monitoring is used, the device is prone to outputting falsely high or low blood pressure values ​​before medical staff re-zero the arterial pressure sensor (level with the right atrium). The system of this invention, by triggering a "high-frequency monitoring observation window" and "temporarily suspending large-scale speed adjustment" during this period, not only effectively avoids the interference of physiological fluctuations in NIBP, but also perfectly compensates for the physical artifacts of the device caused by the lag in sensor zeroing in IBP monitoring, thereby eliminating dangerous drug administration caused by blindly following distorted blood pressure signals.

[0010] Existing technologies (such as CN117414497A) disclose a blood pressure control system based on artificial intelligence deep learning. This system focuses solely on the physiological parameters and infusion rate input into the patient, passively regulating blood pressure through a regression curve. This single-dimensional blood pressure feedback regulation logic has significant blind spots and cannot address multi-dimensional clinical risks such as tubing crystallization, postural changes, and drug accumulation. To address this objective technical problem, this application provides an intelligent intravenous antihypertensive drug delivery regulation system that integrates optical features, tubing pressure, postural changes, and external blood pressure signals. This system acquires underlying physical-dimensional optical and hydrodynamic signals through a fluid interface unit, and combines this with spatial inertial signals captured by a posture sensing unit to construct a multi-sensor cross-validation system. This application utilizes optical feature comparison to achieve physical interception of the infusion tubing, uses cross-validation of tubing hydrostatic pressure and body posture to filter out physiological interference in external blood pressure signals, and introduces pharmacokinetic assessment to proactively prevent the risk of drug accumulation. This technical solution overcomes the limitations of existing technologies that rely solely on data-driven approaches, enabling a leap from passively responding to changes in blood pressure to actively defending against physical embolism and physiological accumulation, significantly improving the safety and clinical reliability of the multidimensional intervention closed loop.

[0011] According to a preferred embodiment, the control terminal host includes a main control module, a pharmacokinetic accumulation early warning model, an early warning unit, and an actuator; the main control module is used to collect external blood pressure signals, intra-tubular pressure data converted from tubular pressure signals, pitch angle data calculated from body position change signals, and optical characteristic signals, so as to execute control logic at each level in sequence and output instructions; the signal input terminal of the actuator is electrically connected to the main control module to convert the generated instructions into control pulses or drive messages and transmit them to the external infusion pump.

[0012] To address the technical challenges of timing misalignment and command conflicts arising from multi-source heterogeneous sensor signals in complex control logic, this application's system incorporates a hardware architecture within the control terminal host that integrates a main control module with a collaborative early warning unit and actuators. The main control module aggregates external blood pressure signals, intra-line pressure data converted from tubing pressure signals, pitch angle data calculated from body position change signals, and optical characteristic signals, unifying the data alignment benchmarks of optical, fluid dynamics, and spatial inertial signals along the time axis. This architecture enables the system to strictly execute control logic at each level, including safety circuit breaker, physiological noise reduction, and steady-state protection, in a strictly sequential manner. Furthermore, the actuators directly convert digital commands generated by the host into control pulses or drive messages adapted to the external infusion pump, shortening the signal transmission path from anomaly detection to mechanical action and ensuring the high efficiency and electrical stability of the conversion of underlying control commands to downstream actuators.

[0013] According to a preferred embodiment, the optical feature signal includes a transmitted light signal and a scattered light signal; the logic executed by the main control module to determine whether drug incompatibility precipitation exists includes: firstly comparing the transmitted light signal and the scattered light signal; when a sudden drop in the amplitude of the transmitted light signal is detected and a sudden increase in the intensity of the scattered light signal is detected, and the duration of the sudden change is compared and matched with a preset duration threshold, it is determined that drug incompatibility precipitation has occurred, and the highest priority physical blocking command is issued in priority over conventional pressure reduction algorithms.

[0014] In clinical intravenous infusion scenarios, antihypertensive drugs such as urapidil are acidic. If they come into contact with alkaline liquids, they will rapidly produce white flocculent precipitates, leading to microvascular embolism. Existing technologies (such as CN117414497A) completely fail to monitor the physicochemical properties of the drug solution within the infusion tubing. To address this blind spot in physical drug administration, this application introduces an optical monitoring mechanism in the fluid interface unit, prioritizing the comparison of transmitted and scattered light signals. Based on the physical laws of optical scattering, when actual chemical precipitate particles appear in the drug solution, these particles block the straight-line propagation of the light beam, causing a sharp drop in the transmitted light amplitude, while simultaneously triggering diffuse reflection, resulting in a sharp increase in the scattered light intensity. The system utilizes this specific abrupt change in physical quantity, combined with a duration threshold to exclude conventional bubbles or occasional photoelectric interference, enabling objective determination of drug incompatibility precipitation. Once an anomaly is detected, the system issues the highest priority blocking command, prioritizing it over conventional antihypertensive algorithms, effectively cutting off the path of dangerous crystalline substances into the human bloodstream at the physical interface.

[0015] According to a preferred embodiment, the logic executed by the main control module to filter out blood pressure fluctuation interference caused by body position changes includes: when the change in pitch angle data calculated from the patient's body position change signal exceeds a set spatial threshold, extracting the hydrostatic pressure change characteristics in the pipeline pressure signal for cross-validation; if the cross-validation confirms that a change in spatial body position has occurred, triggering and entering a high-frequency monitoring observation window mode; during the period of the high-frequency monitoring observation window mode, temporarily suspending the execution of the large-scale flow rate change command triggered by the external blood pressure signal, and only performing micro-flow rate compensation based on the fluctuation trend of the external blood pressure signal within the observation window; after the external blood pressure signal drops and stabilizes, deactivating the observation window mode.

[0016] Postoperative changes in body position are common in patients after cerebrovascular stenosis surgery. Gravity causes blood redistribution, leading to transient physiological blood pressure fluctuations. However, existing technologies (such as CN117414497A) cannot detect the patient's spatial posture, easily misinterpreting these physiological fluctuations as insufficient or excessive medication, thus issuing incorrect flow rate adjustment commands. To address this issue, this application introduces noise reduction logic based on cross-validation of tubing pressure and body position changes. When a patient undergoes a genuine spatial positional change (such as changing from a supine to a sitting position), not only does it trigger peripheral blood redistribution, but the change in the patient's trunk height inevitably causes a physical step change in the hydrostatic pressure of the infusion tubing connected to the vein. This application extracts the hydrostatic pressure change characteristics from the tubing pressure signal and cross-validates them with the body position calculation data at the physical level, effectively filtering out single-sensor motion artifacts caused by local muscle twitching in the patient. After confirming the occurrence of a genuine change in body position, the system enters a high-frequency monitoring and observation window mode and temporarily slows down the significant change in flow rate, providing a buffer period for identifying the true efficacy of the drug and completely avoiding the oscillation of dosing instructions caused by relying solely on external blood pressure fluctuations.

[0017] According to a preferred embodiment, the logic for the main control module to determine the risk of drug accumulation includes calling the internal pharmacokinetic accumulation early warning model: the pharmacokinetic accumulation early warning model performs integral calculations on the time axis based on historical dosing data and built-in pharmacokinetic parameters to predict the cumulative blood drug concentration in the patient; when the predicted cumulative blood drug concentration tends to the steady-state saturation threshold, or when the main control module detects that the downward trend of the external blood pressure signal is accelerating, it issues a preventive rate-limiting command to the actuator in advance to limit the upper limit of the pumping rate or to slightly reduce the speed.

[0018] The study "The Influence of Different Urapidil Administration Routes on Controlled Hypotension During Deep Brain Stimulation" indicates that continuous infusion can stabilize the patient's mean arterial pressure, confirming the clinical value of continuous administration via microinfusion pumps. However, continuous administration carries the risk of drug accumulation over time, and existing technologies (such as CN117414497A) rely solely on infusion parameters from a very short historical period to construct regression curves, failing to address the delayed blood pressure drop caused by steady-state drug accumulation. To address this deficiency, this application utilizes a built-in pharmacokinetic accumulation early warning model to achieve predictive intervention based on pharmacokinetic mechanisms. The system performs integral calculations based on historical dosing data and built-in pharmacokinetic parameters to quantify and predict the cumulative blood drug concentration in the patient. When the predicted blood drug concentration approaches the steady-state saturation threshold, or when the downward slope of the external blood pressure signal is detected to accelerate, the system issues a preventative rate-limiting command to the actuator before the absolute blood pressure value falls below the safe lower limit. This mechanism, through preemptive intervention by quantifying metabolic retention, effectively compensates for the lag inherent in purely data-driven models.

[0019] According to a preferred embodiment, the control terminal host is provided with a preset slot, and the fluid interface unit is at least partially embedded in the preset slot; the fluid interface unit integrates an independent optical turbidity acquisition unit and a pressure acquisition unit.

[0020] Regarding the mechanical connection structure of the system, this invention achieves a fixed assembly between the monitoring consumables and the processing host by setting a preset slot in the control terminal host and embedding at least part of the fluid interface unit within the preset slot. Inside the fluid interface unit, the optical turbidity acquisition unit and the pressure acquisition unit are structurally independent, concentrating the dual physical functions of light and force detection into a single tubing access module. This slot-type embedding structure meets the needs of rapid clinical installation and disassembly while using solid supports to limit the spatial displacement of the sensor acquisition interface relative to the host, reducing mechanical vibration interference caused by external tubing dragging or patient limb movement, and providing a structurally stable testing foundation for the underlying optical detection and pressure acquisition.

[0021] According to a preferred embodiment, the control terminal host is equipped with an optical turbidity analysis unit; the laser emitting module in the optical turbidity acquisition unit emits an initial detection beam, which passes through the optical coupling window and irradiates the flowing depressurizing liquid; the transmitted / scattered light signal carrying the physicochemical state of the liquid passes through the optical alignment window reserved on the side wall of the preset slot and is transmitted to the dual-channel photoelectric receiving module in the optical turbidity analysis unit, and is handed over to the signal feature extraction module to extract the turbidity and precipitation feature signals.

[0022] For the acquisition and transmission structure of optical signals, this invention utilizes an optical alignment window to achieve non-contact optical path transmission across interfaces. The laser emitting module on the fluid interface unit side emits an initial probe beam to irradiate the depressurizing drug solution. The transmitted and scattered light signals, carrying the physicochemical state of the drug solution, pass through the optical alignment window pre-reserved on the side wall of the preset slot and are transmitted to the optical turbidity analysis unit inside the control terminal host for photoelectric conversion and feature extraction. This structural design physically isolates the optical interface that directly contacts the fluid medium from the dual-channel photoelectric receiving module responsible for signal processing, blocking the physical path of accidental leakage from the pipeline that could cause an electrical short circuit inside the host. Simultaneously, this mechanism eliminates the need for a built-in photoelectric receiving and processing chip in the disposable fluid interface unit, reducing the manufacturing cost of clinical disposable consumables while maintaining optical feature detection functionality.

[0023] According to a preferred embodiment, the control terminal host is equipped with a pressure electrical signal processing module; the pressure acquisition unit includes a pressure transmission interface, a mechanical coupling interface module, and a force sensing probe module; the fluid pressure inside the pipeline acts on the pressure transmission interface to cause deformation, and the deformation force is converted into axial mechanical thrust through the mechanical coupling interface module and applied to the force sensing probe module to output the pipeline pressure electrical signal; the pipeline pressure electrical signal is transmitted to the pressure electrical signal processing module inside the control terminal host through a conductive contact terminal or spring provided in a preset slot in a pluggable electrical contact manner.

[0024] To address the common technical problems of fluid column damping attenuation and direct liquid contact with corrosive pressure sensor components in traditional long-distance invasive pressure measurement hoses, this application adopts a structure combining mechanical coupling transmission with pluggable electrical contact terminals. The actual fluid pressure inside the pipeline acts on the pressure transmission interface, causing deformation. This deformation force is directionally converted into axial mechanical thrust by the rear mechanical coupling interface module and applied to the force sensing probe. This physically isolated transmission method blocks the physical path of the core component sensing the corrosive force of the liquid. Subsequently, the generated pipeline pressure electrical signal is directly transmitted to the conversion module inside the host through conductive terminals or spring contacts within the slot. This electrical contact structure replaces the traditional extended hydraulic hose, eliminating the pressure signal hysteresis effect in fluid dynamics transmission and ensuring high-fidelity transmission of the pipeline pressure and hydrostatic pressure characteristic signals.

[0025] According to a preferred embodiment, the posture sensing unit is configured as a wireless patch that can be attached to the core torso area of ​​the patient to acquire the original body position change signal; the control terminal host integrates a posture data interaction unit, which receives the body position change signal through a wireless communication module, and the inertial data parsing module and the posture calculation module call the posture fusion algorithm to smooth and integrate the extracted inertial data, filter out the small tremors caused by breathing fluctuations, and calculate the pitch angle data.

[0026] In terms of the data acquisition and computing architecture of the posture sensing unit, this invention configures the posture sensing unit as a wireless patch that can be attached to the core torso region of the patient. Utilizing the spatial rigidity and linkage characteristics between the torso and the main bones, it avoids local physiological motion interference caused by irregular limb movements at the physical acquisition end. Simultaneously, the system employs a distributed data processing mechanism of sensor front-end acquisition and host back-end processing. The inertial data parsing module and posture calculation module within the control terminal host call the posture fusion algorithm to smooth and integrate the raw inertial data acquired by the patch. This processing mechanism filters out minute tremors caused by chest cavity breathing fluctuations and calculates pitch angle data reflecting the patient's torso spatial state. While ensuring the accuracy of posture determination, it transfers the complex computational load and reduces the data transmission bandwidth requirements and power consumption of the wireless patch front-end.

[0027] According to a preferred embodiment, along the direction of medium flow, the infusion tubing is sequentially inserted into and engaged with the external infusion pump and the fluid interface unit, and finally connected to the patient's infusion end (i.e., the venous end), so that the drug solution is forced to pass through the physical interface detection of the fluid interface unit before entering the human blood vessels.

[0028] In terms of the system's piping arrangement, this invention establishes a physical end-point detection mechanism for the medication before it enters the body by defining the sequential placement of the external infusion pump and fluid interface unit along the flow direction of the infusion tubing, ultimately connecting them to the patient's infusion end. Positioning the fluid interface unit downstream of the infusion pump and upstream of the patient's infusion end forces the medication to flow through a dual optical and pressure detection interface before entering the patient's blood vessels. This positional constraint ensures that drug crystals generated by mixing in the upstream tubing, as well as particulate matter that may be generated by the mechanical extrusion of the infusion pump, are all within the monitoring channel of the fluid interface unit. This guarantees that when the control terminal issues a physical blocking command, it can cut off the inflow path of abnormal substances at the tubing section adjacent to the patient's infusion end, improving the objective effectiveness of the system's physical interception.

[0029] This invention also discloses an intelligent control method for an intravenous antihypertensive drug infusion device, applied to a control terminal host. The method includes the following steps: continuously acquiring optical characteristic signals of the drug solution flowing through the infusion tubing and tubing pressure signals, acquiring the patient's body posture signals, and receiving external blood pressure signals; based on the received optical characteristic signals, tubing pressure signals, body posture signals, and external blood pressure signals, generating control commands according to preset priorities to regulate the external infusion pump. The steps of generating control commands include: prioritizing the analysis of optical characteristic signals, and when precipitation characteristics indicating drug incompatibility are identified, outputting a blocking command to stop the operation of the external infusion pump; if the risk of precipitation is ruled out, cross-validating based on the body posture signals and tubing pressure signals to identify and filter blood pressure fluctuation interference caused by body position changes in the external blood pressure signals, so as to temporarily slow down or dynamically adjust the drug delivery rate; during continuous infusion, inferring the drug accumulation state based on historical drug delivery data and external blood pressure signals, and outputting preventive rate limiting or deceleration commands in advance when the predicted accumulation risk threshold is reached. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the clinical application scenario and physical connection structure of an intelligent intravenous antihypertensive drug administration control system according to a preferred embodiment of the present invention. Figure 2 This is a diagram showing the overall module principle, signal flow direction, and dynamic constraint relationship of an intelligent intravenous antihypertensive drug administration control system according to a preferred embodiment of the present invention. Figure 3 This is a schematic diagram of the logic architecture of a three-level deep control logic according to a preferred embodiment of the present invention; Figure 4 This is a flowchart of an intelligent intravenous antihypertensive drug administration control process according to a preferred embodiment of the present invention.

[0031] List of reference numerals 100: Intelligent intravenous antihypertensive drug administration control system; 200: Control terminal host; 210: Main control module; 220: Pharmacokinetic accumulation early warning model; 230: Early warning unit; 240: Actuator; 250: Optical turbidity analysis unit; 300: Fluid interface unit; 310: Optical turbidity acquisition unit; 311: Laser emission module; 312: Optical coupling window; 313: Dual-channel photoelectric receiving module; 314: Signal feature extraction module; 320: Pressure acquisition unit; 321: Pressure transmission interface; 322: Force sensing probe Modules: 323: Pressure electrical signal processing module; 3131: Low-frequency hydrostatic pressure envelope extraction subunit; 3232: High-frequency mechanical dynamic pressure frequency domain filtering subunit; 3233: High-precision analog-to-digital conversion front end; 324: Mechanical coupling interface module; 400: Attitude sensing unit; 410: Inertial data parsing module; 420: Attitude calculation module; 430: Attitude data interaction unit; 440: Wireless communication module; 500: Infusion pump; 510: Infusion stand; 600: Infusion tubing; 610: Storage bag; 700: Patient infusion end. Detailed Implementation

[0032] The following is a detailed explanation with reference to the accompanying drawings.

[0033] like Figure 1 and Figure 2 As shown, the present invention provides an intelligent intravenous antihypertensive drug administration regulation system 100 based on target blood pressure. The intelligent intravenous antihypertensive drug administration regulation system 100 includes a control terminal host 200, a fluid interface unit 300, and an attitude sensing unit 400. The control terminal host 200 is also equipped with a communication interface for receiving external blood pressure signals.

[0034] Figure 1 The illustration shows a clinical application scenario and physical connection structure of an intelligent intravenous antihypertensive drug administration control system 100 in one exemplary embodiment. Exemplarily, the infusion pump 500 can be suspended and fixed on an infusion stand 510 beside the bed to provide gravitational potential energy and initial conditions for mechanical pumping; the control terminal host 200 is independently placed on a bedside table beside the bed, serving as a central processing platform; the fluid interface unit 300 is at least partially embedded in a preset slot of the control terminal host 200 to achieve data and energy interaction with the host; the posture sensing unit 400, as an independent module, is directly attached to the patient's chest surface area in the form of a wireless patch.

[0035] During operation, the physical flow and force path of the antihypertensive drug solution are as follows: the drug solution is drawn from the storage bag 610 and flows downstream along the infusion line 600; along the direction of flow, the infusion line 600 first passes through and engages with the infusion pump 500, and the mechanical actuator of the infusion pump 500 applies pressure to the line to drive the liquid forward; subsequently, the main body of the infusion line 600 passes through the fluid interface unit 300 at the front of the control terminal host 200 and finally connects to the patient infusion end 700 (i.e., the venous end) at the patient's arm. By arranging the fluid interface unit 300 downstream of the infusion pump 500 and upstream of the patient infusion end 700, the drug solution must be forced to pass through the unit's dual physical interface detection of optical and pressure before entering the human blood vessels.

[0036] Furthermore, the fluid interface unit 300 can collect the physicochemical signals of the drug solution flowing through it and transmit them to the control terminal host 200 via hardwired transmission. The posture sensing unit 400 can move spatially with the patient's body movements and rotations, dynamically collecting the patient's inertial data and transmitting it to the control terminal host 200 in the form of wireless signals. Based on the multi-source converged sensor data and external blood pressure signals, the control terminal host 200 performs logical operations and then adjusts the pumping speed of the infusion pump 500 or executes forced pump shutdown via issued electrical signal control commands.

[0037] like Figure 2 As shown, this embodiment details the overall module principle, signal flow, and dynamic constraint relationships of the intelligent intravenous antihypertensive drug administration control system 100. During operation, the intelligent intravenous antihypertensive drug administration control system 100 acquires and transmits multidimensional data through multiple parallel physical and signal links, ultimately forming a closed-loop intervention at the control center.

[0038] In the first optical monitoring link, the optical turbidity acquisition unit 310 located within the fluid interface unit 300 performs physical signal detection. Specifically, the laser emission module 311 emits a detection beam, which penetrates into the fluid medium through the optical coupling window 312. After the drug solution interacts with the beam, it generates a transmitted / scattered light signal, which is transmitted downstream to the optical turbidity analysis unit 250 within the control terminal host 200. Subsequently, the dual-channel photoelectric receiving module 313 captures these two light signals and processes them through the signal feature extraction module 314, thereby extracting turbidity and precipitation characteristic signals, which are then input to the main control module 210.

[0039] Furthermore, in the optical monitoring link, the fluid interface unit 300 and the control terminal host 200 cooperate to transmit optical signals through an optical window. During operation, the laser emitting module 311, located inside the optical turbidity acquisition unit 310, emits an initial detection beam. This beam passes through the highly transparent optical coupling window 312 and directly irradiates the internally flowing depressurizing drug solution. When the drug solution is in normal condition, the beam is mainly transmitted in a straight line; if urapidil reacts with alkali and causes incompatibility resulting in flocculent precipitate, the beam will be violently scattered within the fluid.

[0040] Along the optical path transmission direction, since the fluid interface unit 300 is embedded in a preset slot in the control terminal host 200, the transmitted / scattered light signal carrying the physicochemical state of the drug solution is output from the fluid interface unit 300 to the control terminal host 200, and then transmitted to the optical turbidity analysis unit 250 inside the control terminal host 200. Inside the optical turbidity analysis unit 250, the dual-channel photoelectric receiving module 313 captures the transmitted and scattered light signals and converts them into initial electrical signals. Subsequently, the signal feature extraction module 314 performs differential comparison and feature extraction on the above signals, calculates the turbidity and precipitation characteristic signals, and inputs the signals to the main control module 210.

[0041] To address the issues of CN107998488A's reliance on passive feedback based on regression curves and KR20210023133A's reliance on a single phototube to determine the presence of drug incompatibility precipitation in the infusion tubing 600, this application's main control module 210 determines the presence of drug incompatibility precipitation within the infusion tubing 600 based on multi-dimensional optical characteristic signals. Specifically, the optical turbidity analysis unit 250 within the control terminal host 200 is connected to the fluid interface unit 300 via a spatially integrated positioning method involving electrical plug-in and optical path physical isolation. This unit forms a constraint relationship with the optical alignment window of the laser emission module 311 within the fluid interface unit 300, enabling a non-contact optical detection relationship between the fluid interface unit 300 and the depressurizing drug flowing through the infusion tubing 600. The laser emission module 311 guides the generated initial detection beam of a specific narrowband wavelength to the laminar flow interface of the depressurizing drug within the fluid interface unit 300. The dual-channel photoelectric receiving module 313 is equipped with a transmission light sensing array for... The system captures transmitted light signals penetrating the drug solution medium along the original straight propagation direction of the probe beam. The independently configured scattered light sensing array in the dual-channel photoelectric receiving module 313 is used to capture scattered light signals generated by suspended particles at a specific side angle position deviating from the straight propagation direction of the probe beam. When the depressurizing drug solution remains clear and laminar flow exists, the initial probe beam generated by the laser emitting module 311 maintains photoelectric coupling with the transmitted light sensing array. When the depressurizing drug solution is incompatible and crystallization occurs, the probe beam generated by the laser emitting module 311 forms a sudden photoelectric coupling response with the scattered light sensing array under the condition of spatial diffuse reflection and attenuation of the number of transmitted photons. Furthermore, when there are degassed bubbles inside the infusion line 600, the signal generated by the scattered light sensing array is compared and filtered by sliding comparison with the preset morphological change duration threshold inside the signal feature extraction module 314, so that the main control module 210 filters out transient photoelectric noise, thereby identifying incompatibility and eliminating bubble interference.

[0042] In the second fluid dynamics monitoring link, the fluid pressure in the pipeline directly acts on the pressure transmission interface 321 of the pressure acquisition unit 320, and the mechanical thrust is directly transmitted to the force sensing probe module 322 through the mechanical coupling interface module 324. After the probe undergoes elastic deformation under force, it outputs a pipeline pressure electrical signal to the pressure electrical signal processing module 323 in the control terminal host 200. After conversion and processing, the pipeline internal pressure data, representing the internal pressure of the pipeline, is sent to the main control module 210.

[0043] Furthermore, in the fluid dynamics monitoring link, the pressure acquisition unit 320 acquires pressure through the contact and linkage of mechanical transmission components. Specifically, the actual fluid pressure inside the pipeline first acts directly on the pressure transmission interface 321 (such as a flexible diaphragm interface), forcing it to expand or bulge outward. The deformation force generated by the pressure transmission interface 321 is directionally guided by the mechanical coupling interface module 324 located at its rear end, converting the irregular hydraulic pressure into axial mechanical thrust, which is ultimately applied to the force sensing probe module 322.

[0044] The force sensing probe module 322 undergoes elastic deformation after being subjected to the mechanical thrust, and outputs a corresponding proportional pipeline pressure electrical signal accordingly. For example... Figure 2 As shown in the signal flow diagram, the pipeline pressure signal is led out from the fluid interface unit 300 and transmitted to the pressure signal processing module 323 inside the control terminal host 200 via a pluggable electrical contact terminal / spring provided in a preset slot. The pressure signal processing module 323 amplifies and compensates the weak signal to generate pipeline internal pressure data characterizing the pipeline fluid flow and hydrostatic pressure, and sends it to the main control module 210, thereby providing fluid dynamics data support for subsequent body position monitoring logic.

[0045] To address the issue in CN107998488A where the inability to sense patient spatial posture leads to misinterpretation of physiological blood pressure fluctuations caused by gravity as a drug response, this application constructs a noise reduction logic that cross-validates patient surface spatial motion with the underlying fluid dynamics of the medical tubing. The posture sensing unit 400, serving as the motion signal extraction source, comprises a microelectromechanical inertial sensor component, a wireless communication module 440, and a flexible medical-grade skin-friendly patch substrate. The microelectromechanical inertial sensor component and wireless communication module 440, in a state where the flexible medical-grade skin-friendly patch substrate is attached and follows the patient's core torso in three-dimensional spatial linkage, perform six-axis spatial inertial data acquisition and radio frequency transmission. The posture sensing unit 400 is fixed to the core chest undulation measurement site, reflecting the spatial elevation angle displacement characteristics of the patient's main skeletal structure. When the patient's torso is in a supine position, the posture sensing unit 400 is positioned at the reference gravity zero-point horizontal position, allowing the inertial data analysis module 410 to perform initial dynamic calibration of the zero-point reference measurement plane. When the patient's torso changes from a supine to a semi-recumbent posture, the posture... The sensing unit 400 is positioned at a spatial tilt to allow the main control module 210 to acquire pitch angle data representing three-dimensional body rotation. To extract low-frequency hydrostatic pressure changes in the context of pulsed dynamic pressure generated by the external infusion pump 500, the pressure-electric signal processing module 323 integrated within the main control module 210 includes a low-frequency hydrostatic pressure envelope extraction subunit 3231, a high-frequency mechanical dynamic pressure frequency domain filtering subunit 3232, and a high-precision analog-to-digital conversion front-end 3233. The low-frequency hydrostatic pressure envelope extraction subunit 3231 and the high-frequency mechanical dynamic pressure frequency domain filtering subunit 3232... The mechanical pressure frequency domain filtering subunit 3232 and the high-precision analog-to-digital conversion front-end 3233 together with the computing microprocessor of the main control module 210 form a frequency domain differential isolation conversion structure, which is used to filter out the high-frequency mechanical harmonic noise generated by the external infusion pump 500 and strip out the low-frequency hydrostatic pressure step characteristic signal representing the patient's elevation change. Then, the characteristics of the purified pipeline pressure data change are compared with the trunk pitch angle flip data in the time domain based on the timestamp to identify and filter the physiological displacement artifacts contained in the external blood pressure signal input to the system.

[0046] In the third posture monitoring link, the attitude sensing unit 400 captures the spatial motion state as the patient's body rises and falls, and sends the original posture change signal to the attitude data interaction unit 430 in the control terminal host 200. Inside the attitude data interaction unit 430, after the signal is received by the wireless communication module 440, the inertial data analysis module 410 and the attitude calculation module 420 work together to calculate the pitch angle data and send it to the main control module 210.

[0047] Furthermore, such as Figure 1As shown, the posture sensing unit 400 can be configured as a miniaturized wireless patch. During clinical fitting, the posture sensing unit 400 is directly attached to the core trunk area of ​​the patient, such as the chest or below the collarbone. By positioning the posture sensing unit 400 in this location, it can maintain rigid linkage with the patient's main skeleton, thereby effectively avoiding physiological noise interference caused by irregular limb movements.

[0048] During operation, when a patient changes position due to nursing needs or sleep habits (such as changing from a supine position to a semi-recumbent position), the posture sensing unit 400 attached to the body surface undergoes three-dimensional spatial displacement along with the torso. Sensors inside the posture sensing unit 400 (such as microelectromechanical inertial sensors) respond to this physical displacement, generate an original posture change signal, and transmit the signal out of the physical boundary in the form of radio electromagnetic waves via a low-power radio frequency protocol.

[0049] Combination Figure 2 As shown, the aforementioned body position change signal is transmitted to the control terminal host 200 via wireless electromagnetic signals. Specifically, the control terminal host 200 integrates an attitude data interaction unit 430. Along the signal transmission path, the wireless communication module 440 inside the attitude data interaction unit 430 first receives the wireless body position change signal and converts it into a digital baseband signal readable by the internal bus. Subsequently, the signal is transmitted to each module sequentially according to a preset timing sequence. The inertial data analysis module 410 first unpacks and converts the received data packets into physical quantities, extracting structured inertial data containing triaxial acceleration and angular velocity; then, the attitude calculation module 420 calls an attitude fusion algorithm (such as Kalman filtering) to smooth and integrate the aforementioned inertial data, filtering out minor tremors caused by respiratory fluctuations, and finally calculating the pitch angle data characterizing the patient's spatial state. Figure 2 As shown, the pitch angle data is ultimately input to the main control module 210, thereby providing a spatial attitude reference for the system to perform hydrostatic cross-validation by combining pipeline pressure data and filtering out positional blood pressure fluctuation interference.

[0050] like Figure 2 As shown, the control terminal host 200 serves as a central processing device for information aggregation and instruction generation. It aggregates multiple parsed internal sensor signals and external access signals to the main control module 210 through an internal bus and communication interface.

[0051] During the control execution phase, the main control module 210, as the system core, aggregates the aforementioned turbidity and precipitation characteristic signals, external blood pressure signals, tubing pressure data, and pitch angle data. Internally, the main control module 210 calls the pharmacokinetic accumulation early warning model 220 to perform pharmacokinetic integral calculations and triggers an alarm by linking the early warning unit 230 when an abnormal threshold is detected. Finally, the main control module 210 generates physical control commands based on the three-level deep control logic and drives the actuator 240 to operate. The actuator 240 transmits control outputs (such as physical blocking, rate limiting, or pause commands) to the external infusion pump 500 to adjust the flow rate of the antihypertensive drug solution pumped to the patient's infusion terminal 700. This forms a complete dynamic closed loop from the patient's physiological / physicochemical perception to the adjustment of the infusion terminal.

[0052] At the signal receiving level, the input terminals of the main control module 210 are connected to the optical turbidity analysis unit 250, the pressure-electric signal processing module 323, the posture data interaction unit 430, and the external physiological data communication interface. During operation, the main control module 210 receives turbidity and precipitation characteristic signals, tubing pressure data, pitch angle data, and external blood pressure signals transmitted by the aforementioned modules and interfaces. By aggregating these four independent signals, the system establishes a comprehensive judgment benchmark for the physicochemical state of the drug solution, tubing fluid dynamics, patient spatial posture, and hemodynamic indicators.

[0053] At the logical processing level, such as Figure 2 As shown, the main control module 210 integrates an early warning unit 230 and a pharmacokinetic accumulation early warning model 220. When the main control module 210 determines that the turbidity and precipitation characteristic signals meet the threshold conditions for incompatibility, it sends a trigger electrical signal to the early warning unit 230, driving it to output an audible and visual alarm. When the change in pitch angle data combined with the hydrostatic pressure change in pipeline pressure data meets the cross-validation conditions, the system triggers and enters the high-frequency monitoring observation window mode. Simultaneously, the pharmacokinetic accumulation early warning model 220 performs integral calculations based on the input external blood pressure signal and historical drug administration data to deduce the cumulative blood drug concentration curve in the patient's body. When the model predicts that the blood drug concentration tends to the steady-state saturation threshold, or when the main control module 210 detects that the downward slope of the external blood pressure signal accelerates, the system generates a preventive rate-limiting command.

[0054] To address the issue in CN110869074A that lacks a parallel early warning intervention mechanism targeting specific drug pharmacokinetic pathways, potentially leading to delayed hypotensive shock in patients, and to eliminate the regulatory blind spots in basic closed-loop control based on external target blood pressure under normal conditions and to resolve timing inconsistencies in different control processes, the control terminal host 200 of this application adopts a dual-track parallel mechanism of a PID main control thread and a single-compartment pharmacokinetic asynchronous daemon thread in its underlying operating system computing power scheduling architecture. Specifically, the early warning unit 230 within the main control module 210 receives real-time external blood pressure information input from external monitoring equipment. The system generates a basic PID flow rate dynamic adjustment message for routine feedback compensation based on the preset target blood pressure safety range; the pharmacokinetic accumulation early warning model 220 embedded in the main control module 210 is used to accurately accumulate the total mass of antihypertensive drug solution pumped in based on the continuously stored historical data of drug administration, and to subtract the total amount of drug eliminated by the body through calculus calculation based on the built-in drug single-compartment model biological elimination half-life parameter and the patient-specific apparent distribution volume parameter, thereby generating a dynamic cumulative blood drug concentration characteristic value that characterizes real-time physiological residues; the actuator 240 is used to connect the main control module... The logic instructions generated by block 210 are converted into low-level hardware duty cycle control pulses to drive the mechanical stepper motor of the external infusion pump 500. Specifically, when the main control module 210 accurately determines that no postural disturbance event has been triggered and the internal physicochemical characteristics of the tubing are normal, the early warning unit 230, when the external blood pressure signal deviates from the target blood pressure expectation value, forms a dynamic speed regulation relationship with the actuator 240 based on a PID-based blood pressure reduction feedback control algorithm. When the main control module 210 confirms a change in patient position through cross-validation by the aforementioned attitude calculation module 420 and enters a high-frequency monitoring observation window mode to temporarily slow down significant changes in flow rate,… Under the condition of instruction issuance, the pharmacokinetic accumulation early warning model 220 maintains a synchronous integral calculation and deduction relationship with the hardware clock of the main control module 210 in an asynchronous background independent operation state; and when the pharmacokinetic accumulation early warning model 220 predicts that the characteristic value of dynamic cumulative blood drug concentration in the patient tends to the preset steady-state pharmacological saturation threshold, the pharmacokinetic accumulation early warning model 220 forms a preventive rate limiting intervention relationship with the actuator 240 through an interrupt preemptive instruction issuance method to limit the upper limit of the delivery rate of the external infusion pump 500 or slightly reduce it, so as to ensure the continuity of the calculation of drug accumulation metabolism integral deduction during the clinical interruption observation window.

[0055] At the mechanical actuation level, the signal input terminal of actuator 240 is electrically connected to the main control module 210, and its control output terminal is connected to the external infusion pump 500 via a communication cable. When any of the above intervention conditions are met, the main control module 210 sends the generated command to actuator 240, which then converts the command into a control pulse or drive message adapted to the external device and transmits it to infusion pump 500. After receiving the control output, infusion pump 500 directly changes the physical squeezing frequency applied to the infusion tubing by changing the speed of its internal pumping motor or performing an emergency stop action, thereby dynamically adjusting the flow rate of the antihypertensive drug solution pumped into the patient's infusion end 700. Thus, this embodiment achieves closed-loop control from physiological / physicochemical perception to mechanical pumping regulation through explicit electrical signal transmission and mechanical action conversion.

[0056] like Figure 3 The logical architecture shown is Figure 4 The execution flow shown in this embodiment provides an intelligent intravenous antihypertensive drug administration control workflow with three levels of depth.

[0057] In the hardware collaboration and data input phase, such as Figure 4 As shown, after the system starts, the main control module 210 collects multiple sensor signals: namely, it acquires the transmitted and scattered light signals of the drug solution and the pipeline pressure signal through the fluid interface unit 300, acquires the patient's posture pitch angle data through the posture sensing unit 400, and receives external blood pressure signals. Based on the above-mentioned converged data stream, the main control module 210 follows... Figure 4 The defined defense priority executes the following three levels of control logic in sequence: the first level is the highest priority safety circuit breaker logic based on optical characteristics; the second level is the noise reduction logic based on cross-validation of body position and hydrostatic pressure; and the third level is the steady-state protection and pre-intervention logic based on pharmacokinetics.

[0058] Preferably, for the first-level control logic (i.e., the highest-priority safety fuse logic based on optical characteristics), such as Figure 3 Left logical branch AND Figure 4 As shown, the main control module 210 sets the prevention of physicochemical incompatibilities as the highest priority judgment condition of the system. During operation, the main control module 210 prioritizes the comparison and analysis of the input optical data.

[0059] When the main control module 210 detects a sudden drop in the amplitude of the transmitted light signal and a sudden increase in the intensity of the scattered light signal, and compares the duration of this abrupt change with a preset duration threshold, the main control module 210 determines that a drug incompatibility has occurred inside the infusion tubing (such as urapidil crystallizing and precipitating in alkaline liquids). Once precipitation is detected, the main control module 210 prioritizes the conventional pressure-reducing algorithm and triggers the highest-priority physical blocking command to the actuator 240. The actuator 240 then stops the mechanical operation of the external infusion pump 500 and triggers an alarm in the early warning unit 230, thereby ensuring physical disconnection before the dangerous flocculent drug solution enters the human vein, preventing microvascular embolism.

[0060] Preferably, for the second-level control logic (i.e., the noise reduction logic based on cross-validation of body position and hydrostatic pressure), such as Figure 4 As shown, if the main control module 210 determines that the optical characteristics of the liquid medicine are normal (i.e., eliminating the physical risk of precipitation), the system then executes the body position interference detection logic. Combined with... Figure 3 In the middle logic branch, the main control module 210 compares the patient's pitch angle change with the hydrostatic pressure change characteristics in the pipeline internal pressure data converted from pipeline pressure signals.

[0061] When the change in pitch angle data calculated from the body position change signal exceeds a set spatial threshold (e.g., >30°), the main control module 210 extracts the hydrostatic pressure change characteristics from the pipeline pressure data for cross-validation. If the cross-validation confirms a change in body position, the system triggers and enters the "high-frequency monitoring observation window" mode. During this observation window period, the system temporarily suspends the execution of any commands that cause significant changes in flow rate at the control output end to avoid physiological blood pressure fluctuations (transient fluctuations) caused by gravity changes being misjudged as insufficient or excessive drug efficacy. Subsequently, the system dynamically adjusts the drug administration command based on the fluctuation trend of the external blood pressure signal within the observation window; after the external blood pressure signal drops and stabilizes, the system deactivates the observation window mode and returns to the normal monitoring and drug administration state.

[0062] Preferably, for the third-level control logic (i.e., the steady-state protection and pre-intervention logic based on pharmacokinetics), such as Figure 4 As shown, after successively eliminating the risk of bottom-layer sedimentation and mid-layer positional disturbances, the system enters the highest-level steady-state protection stage. Combined with... Figure 3 On the right-hand logic branch, the main control module 210 calls the internal pharmacokinetic accumulation early warning model 220 to run.

[0063] During continuous infusion, the pharmacokinetic accumulation early warning model 220 calculates the metabolism and residue curve of the antihypertensive drug in the patient's body by integrating on the time axis based on the built-in pharmacokinetic parameters, thereby deriving the "cumulative blood drug concentration" in the patient's body. The main control module 210 monitors this calculated value and checks whether the continuous infusion duration has entered the preset high-risk period (e.g., continuous infusion > 2 hours).

[0064] When the pharmacokinetic accumulation early warning model 220 predicts that the cumulative blood drug concentration is approaching the steady-state saturation threshold, or when the main control module 210 detects an accelerated downward slope in the external blood pressure signal, the system triggers the "steady-state protection logic." At this time, the main control module 210 sends a preventative rate-limiting command to the actuator 240, limiting the upper limit of the pumping rate or performing a slight deceleration. The actuator 240 then limits the upper limit of the pumping rate of the infusion pump 500 or performs a slight deceleration operation. Thus, by outputting the rate-limiting command in advance, the system avoids delayed hypotensive shock caused by the superposition of drug concentration accumulation and the patient's nighttime sleep cycle.

[0065] In one alternative implementation, the specific sensing structure within the fluid interface unit 300 can be replaced with an equivalent one. For example, the pressure acquisition unit 320 can be replaced with a non-contact ultrasonic flow / pressure composite sensor; or, the dual-channel photoelectric receiver module 313 can be replaced with a multispectral array sensor with higher spectral resolution. An equivalent replacement must at least satisfy the following conditions: consistent external physical installation interface, consistent dimensional constraints of the sensing data, and the same conversion mechanism from pipeline pressure to electrical signal or from optical signal to electrical signal.

[0066] Preferably, the specific parameters involved in this embodiment (such as the threshold of pitch angle >30°, the duration of the high-frequency monitoring observation window, and the steady-state critical point of continuous infusion >2 hours) are all preferred parameter ranges suitable for urapidil and post-neurosurgery patients. In actual clinical applications, the above parameters are dynamically configurable variable parameters. The upper and lower limits of such parameters correspond to the safety threshold boundaries suitable for different patients, covering the dynamic range from the worst physiological tolerance condition to the optimal stable condition, thereby taking into account individual differences and the safety of blood pressure reduction.

[0067] When defining the state changes of the system for recognizing sedimentation or blood pressure fluctuations, the so-called "basically stable" or "uniform fluctuation" of the characteristic value means that the deviation of the data angle or value within adjacent sampling periods does not exceed the safety tolerance threshold set by the system, so as to avoid false triggering caused by the patient's local limb twitching / uneven force or electromagnetic noise.

[0068] Without departing from the concept of this invention, the shape, quantity, material, and arrangement of each acquisition unit can be adaptively adjusted according to the tubing specifications or the patient's body shape. The judgment conditions and execution sequence in the three-level depth control logic constitute the core mechanism of this invention, but the specific electrical disconnection method of the actuator 240 intervening in the infusion pump can be implemented in multiple ways.

[0069] In the actual operation of the system described in this application, the PID control algorithm executed by the main control module 210 through the control terminal host 200 is highly dependent on the preset "target blood pressure" parameter to achieve closed-loop feedback regulation. As a preferred embodiment, for neurocritical care scenarios after carotid endarterectomy, carotid artery stenting, and endovascular treatment of intracranial atherosclerotic stenosis, in order to prevent cerebral hyperperfusion syndrome while taking into account myocardial perfusion needs, the preset target blood pressure range within the system is usually set to 100~120 mmHg.

[0070] The main control module 210 continuously calculates the deviation between the real-time received external blood pressure signal and the median of the target blood pressure range, such as 110 mmHg. The PID main control thread determines the PID control based on the deviation. Deviation integral and deviation differential The system calculates the rate correction command for the external infusion pump 500 according to preset proportional, integral, and derivative coefficients. For example, when the external blood pressure signal rises from 115 mmHg to 125 mmHg and shows an accelerating upward trend, the PID main control thread generates a command to increase the urapidil infusion rate, which is converted into a control pulse by the actuator 240 to increase the frequency of the stepper motor of the external infusion pump 500. If the aforementioned optical incompatibility, postural disturbances, or pharmacokinetic accumulation risks are triggered during this process, the system will override or correct the basic PID command according to the corresponding priority logic, thereby stabilizing the patient's blood pressure within the clinical target range of 100-120 mmHg while ensuring safety.

[0071] It should be noted that the specific embodiments described above are exemplary. Those skilled in the art can devise various solutions inspired by the disclosure of this invention, and these solutions all fall within the scope of this invention and its protection. Those skilled in the art should understand that this specification and its accompanying drawings are illustrative and not intended to limit the scope of the claims. The scope of protection of this invention is defined by the claims and their equivalents. This specification contains multiple inventive concepts; terms such as "preferredly," "according to a preferred embodiment," or "optionally" indicate that the corresponding paragraph discloses an independent concept. The applicant reserves the right to file divisional applications based on each inventive concept.

Claims

1. An intelligent intravenous antihypertensive drug administration regulation system based on target blood pressure, characterized in that, It includes: A fluid interface unit (300) is configured in an infusion line (600) to acquire optical characteristic signals of the drug solution flowing through it and pressure signals within the infusion line (600); A posture sensing unit (400) is used to acquire patient position change signals; The control terminal host (200) is communicatively connected to the fluid interface unit (300) and the attitude sensing unit (400), and is equipped with a communication interface for receiving external blood pressure signals; The control terminal host (200) is configured to generate control commands for regulating the operation of the external infusion pump (500) based on the fusion analysis of received optical feature signals, pipeline pressure signals, body position change signals and external blood pressure signals. The control logic executed by the control terminal host (200) includes: The presence of drug incompatibility precipitates in the infusion tubing (600) is determined based on optical feature signals, and a blocking command is triggered for the external infusion pump (500) when precipitates are detected. Based on body position change signals and tubing pressure signals, the system identifies and filters out blood pressure fluctuation interference caused by body position changes in external blood pressure signals in order to dynamically adjust the drug administration control instructions. Based on continuous dosing history data and external blood pressure signals, the risk of drug accumulation is determined, and preventive rate limiting instructions are issued in advance when there is a steady-state accumulation or a delayed blood pressure drop trend.

2. The system according to claim 1, characterized in that, The control terminal host (200) includes a main control module (210), a pharmacokinetic accumulation early warning model (220), an early warning unit (230), and an actuator (240). The main control module (210) is used to collect the external blood pressure signal, the intra-line pressure data converted from the pipeline pressure signal, the pitch angle data calculated from the body position change signal, and the optical feature signal, so as to execute the control logic at each level in sequence and output instructions; the signal input terminal of the actuator (240) is electrically connected to the main control module (210) to convert the generated instructions into control pulses or drive messages and transmit them to the external infusion pump (500).

3. The system according to claim 1 or 2, characterized in that, The optical characteristic signals include transmitted light signals and scattered light signals; The logic executed by the main control module (210) for determining whether drug incompatibility precipitation exists includes: first comparing the transmitted light signal and the scattered light signal; when a sudden drop in the amplitude of the transmitted light signal is detected and a sudden increase in the intensity of the scattered light signal is detected, and the duration of the sudden change is compared and matched with a preset duration threshold, it is determined that drug incompatibility precipitation has occurred, and the highest priority physical blocking command is issued in priority over the conventional blood pressure reduction algorithm.

4. The system according to any one of claims 1 to 3, characterized in that, The logic executed by the main control module (210) to filter out blood pressure fluctuations caused by changes in body position includes: When the change in pitch angle data calculated from the body position change signal exceeds a set spatial threshold, the hydrostatic pressure change characteristics in the pipeline pressure signal are extracted for cross-validation. If the cross-validation confirms a change in body position, a high-frequency monitoring observation window mode is triggered and entered. During the period of the high-frequency monitoring observation window mode, the execution of the large-scale flow rate change command triggered by the external blood pressure signal is temporarily suspended, and only a small flow rate compensation is performed based on the fluctuation trend of the external blood pressure signal within the observation window. After the external blood pressure signal drops and stabilizes, the observation window mode is deactivated.

5. The system according to any one of claims 1 to 4, characterized in that, The logic by which the main control module (210) determines the risk of drug accumulation includes calling the internal pharmacokinetic accumulation early warning model (220) to run: The pharmacokinetic accumulation early warning model (220) performs integral calculations on the time axis based on historical dosing data and built-in pharmacokinetic parameters to estimate the cumulative blood drug concentration in the patient's body. When the predicted cumulative blood drug concentration tends to the steady-state saturation threshold, or when the main control module (210) detects that the downward trend slope of the external blood pressure signal is accelerating, it issues a preventive speed-limiting command to the actuator (240) in advance to limit the upper limit of the pumping rate or to slightly reduce the speed.

6. The system according to any one of claims 1 to 5, characterized in that, The control terminal host (200) is provided with a preset slot, and the fluid interface unit (300) is at least partially embedded in the preset slot; The fluid interface unit (300) integrates an independent optical turbidity acquisition unit (310) and a pressure acquisition unit (320).

7. The system according to any one of claims 1 to 6, characterized in that, The control terminal host (200) is equipped with an optical turbidity analysis unit (250). The laser emitting module (311) in the optical turbidity acquisition unit (310) emits an initial detection beam, which passes through the optical coupling window (312) and irradiates the flowing depressurizing drug solution; The transmitted / scattered light signal carrying the physicochemical state of the drug solution passes through the optical alignment window reserved on the side wall of the preset slot and is transmitted to the dual-channel photoelectric receiving module (313) in the optical turbidity analysis unit (250), and is handed over to the signal feature extraction module (314) to extract the turbidity and precipitation feature signals.

8. The system according to any one of claims 1 to 7, characterized in that, The control terminal host (200) is equipped with a pressure electrical signal processing module (323). The pressure acquisition unit (320) includes a pressure transmission interface (321), a mechanical coupling interface module (324), and a force sensing probe module (322). The fluid pressure inside the pipeline acts on the pressure transmission interface (321) to cause it to deform. The deformation force is converted into an axial mechanical thrust through the mechanical coupling interface module (324) and applied to the force sensing probe module (322) to output a pipeline pressure electrical signal. The pipeline pressure signal is transmitted to the pressure signal processing module (323) inside the control terminal host (200) through the conductive contact terminal or spring provided in the preset slot in a pluggable electrical contact manner.

9. The system according to any one of claims 1 to 8, characterized in that, The posture sensing unit (400) is configured as a wireless patch that can be attached to the core torso region of the patient to acquire raw positional change signals. The control terminal host (200) integrates an attitude data interaction unit (430). The attitude data interaction unit (430) receives the body position change signal through the wireless communication module (440), and the inertial data parsing module (410) and attitude calculation module (420) call the attitude fusion algorithm to smooth and integrate the extracted inertial data, filter out the small tremors caused by breathing fluctuations, and calculate the pitch angle data.

10. The system according to any one of claims 1 to 9, characterized in that, Along the direction of medium flow, the infusion line (600) is sequentially inserted into and engaged with the external infusion pump (500) and the fluid interface unit (300), and finally connected to the patient infusion end (700), so that the drug solution is forced to pass through the physical interface detection of the fluid interface unit (300) before entering the human blood vessels.

Images