Search

CN-122004814-A - Implantable hypertension self-adaptive closed-loop treatment system and method based on multi-sensor fusion

CN122004814ACN 122004814 ACN122004814 ACN 122004814ACN-122004814-A

Abstract

The invention discloses an implantable hypertension self-adaptive closed-loop treatment system and method based on multi-sensor fusion, and relates to the crossing field of medical instruments and biomedical engineering. The sensing array is attached around the outer wall of the ascending aorta to collect multi-point pressure signals, the main processing unit executes a blood pressure reconstruction and control decision algorithm to drive the execution unit, the execution unit comprises a carotid sinus and renal artery region dual nerve stimulation electrode and a renal artery local trace drug delivery device, and the programmer is used for setting targets and calibration. The method comprises the steps of collecting multichannel signals, relatively aligning, weighting and fusing, reversely compensating and reconstructing blood pressure by a vascular wall transfer function, sequentially activating stimulation or drug administration according to a grading strategy, periodically calibrating and updating a model, further having a preventive self-maintenance function, infusing anti-fibrosis drugs, preventing package, monitoring electric quantity, impedance and the like, and degrading operation.

Inventors

  • LIN YANAN

Assignees

  • 林亚南

Dates

Publication Date
20260512
Application Date
20260320

Claims (10)

  1. 1. An implantable hypertension self-adaptive closed-loop treatment system based on multi-sensor fusion is characterized by comprising a distributed blood pressure sensing array, an implantable main processing unit, a multi-target execution unit and an external programmer; The distributed blood pressure sensing array is used for acquiring pressure signals from the outer space of the wall of the ascending aorta at multiple points; The implanted main processing unit is used for processing the pressure signal, executing a blood pressure reconstruction algorithm and a control decision algorithm, and driving the multi-target execution unit; The multi-target execution unit at least comprises a first nerve stimulation electrode, a second nerve stimulation electrode and a micro drug delivery device; The external programmer is used for setting a blood pressure target value, receiving calibration data and system state information and wirelessly communicating with the implanted main processing unit; The wireless communication between the implanted main processing unit and the external programmer adopts a dual-band hybrid architecture, wherein a low band is used for awakening, near-field safe pairing and emergency instruction transmission, and a high band is used for sensing data, treatment record transmission and algorithm parameter updating; The energy management module of the implanted main processing unit adopts a hybrid energy supply architecture of rechargeable battery and super capacitor buffer, energy is supplemented through a resonant wireless charging technology, and the charging frequency is selected from a medical special frequency band of 6.78MHz or hundreds of kHz.
  2. 2. The system of claim 1, wherein the distributed blood pressure sensing array comprises 4-8 micro-fiber pressure sensors, the sensors are attached and fixed on a flexible support of the outer wall of an ascending aorta blood vessel in a surrounding mode, the flexible support is made of medical titanium alloy, a biocompatible film is coated on the surface of the flexible support, and the support is provided with a porous structure to promote adventitia tissue ingrowth.
  3. 3. The system of claim 1, wherein the implantable main processing unit is a hermetically sealed MEMS micro-system containing a microprocessor, a sensing signal demodulation circuit, a stimulus generation circuit, a wireless communication module, and an energy management module, and wherein the control algorithm executed by the system comprises: Blood pressure reconstruction sub-algorithm based on multi-sensor signal dynamic weighted fusion and inverse compensation of vascular wall transfer function, wherein the transfer function is in the form of The system comprises a static gain coefficient K, a time delay T1 and a time constant T2, and a hierarchical intervention decision sub-algorithm based on the deviation between the real-time blood pressure and a target value, wherein the trigger threshold of the sub-algorithm is set based on clinical guidelines and neural stimulation/drug efficacy time literature data and supports individual adjustment.
  4. 4. The system of claim 1, wherein the first nerve stimulation electrode is a flexible electrode wrapped around an outer wall of the carotid artery Dou Xieguan for low-intensity electrical stimulation, and wherein the second nerve stimulation electrode is a microelectrode attached to a nitinol stent near the renal artery for targeted electrical inhibition of renal sympathetic nerves; The micro drug delivery device comprises a micro drug storage bag and a precision pump, wherein the output end of the micro drug delivery device is positioned in a renal artery area and used for local micro drug infusion, the drug is at least one of rapamycin, paclitaxel and everolimus, preferably rapamycin, the drug concentration is 0.5-2.0 mug/mL, and the single drug administration dosage is 0.5-2.0mL; The miniature medicine storage bag is made of medical silica gel or polyurethane, has a volume of 1.0-2.0mL and a self-sealing diaphragm structure, and is of a passive diffusion type or active pumping type, and the flow is controlled to be 0.1-1.0 mu L/day.
  5. 5. The system of claim 1, wherein the external programmer is a mobile device with wireless communication capability for setting personalized blood pressure target values and time phase parameters, receiving non-invasive blood pressure meter measurements to trigger system calibration, and displaying system operating status and alarm information.
  6. 6. An implantable hypertension self-adaptive closed-loop treatment method based on multi-sensor fusion is characterized by comprising the following steps: s1, acquiring multichannel pressure signals through a distributed sensing array surrounding the outer wall of an ascending aorta blood vessel; S2, aligning, dynamically weighting and fusing the multichannel pressure signals, and inversely compensating a vascular wall transfer function to reconstruct an intravascular real-time blood pressure waveform and value; S3, comparing the reconstructed blood pressure value with a target value, and sequentially or selectively activating a first nerve stimulation, a second nerve stimulation and local drug infusion according to a hierarchical intervention strategy; S4, periodically receiving a establish-free standard blood pressure value, and updating parameters of a vascular wall transfer function model and a blood pressure reconstruction algorithm model; S5, performing preventive self-maintenance drug infusion, namely infusing anti-fibrosis/anti-proliferation drugs into tissues around the electrode and the stent through a micro drug infusion device at the initial stage of implantation and at periodic time points (every 3-6 months) so as to prevent the electrode point and the stent from being wrapped; In the hierarchical intervention strategy in the step S3, the first preset time window is 15-60 minutes, and the second preset time window is 45-120 minutes.
  7. 7. The method according to claim 6, wherein in step S2, the inverse compensation of the vessel wall transfer function specifically comprises: in an off-line modeling stage, an individuation transfer function initial model is established by synchronously acquiring an extravascular sensor signal and a noninvasive sphygmomanometer reference value and adopting a system identification method; and in the online compensation stage, the initial model is utilized to carry out inverse filtering treatment on the real-time fusion signal, and a Kalman filter is combined to separate a blood pressure rapid change component and a drift component, so as to output a reconstructed blood pressure value.
  8. 8. The method of claim 6, wherein the hierarchical intervention strategy in step S3 is: when the reconstructed blood pressure value continuously exceeds the upper limit of the target value (the systolic pressure is more than or equal to 140mmHg or the diastolic pressure is more than or equal to 90 mmHg), the first nerve stimulation is started preferentially; If the blood pressure does not reach the standard within the first preset time window (15-60 minutes, preferably 30 minutes), the second nerve stimulation is superposed or switched on; if the local drug infusion is not reached within a second preset time window (45-120 minutes, preferably 60 minutes) then starting the local drug infusion; The upper limit of the target value is set based on the guidelines for prevention and treatment of hypertension in China (2023 edition), and ACC/AHA in the United states.
  9. 9. The method according to claim 6, wherein in step S4, the updating parameters of the vessel wall transfer function model specifically includes: establishing a mapping relation between sensor waveform characteristics and calibration blood pressure values by adopting a transfer function model or a machine learning model; when a transfer function model is adopted, optimizing parameters by a least square method or a gradient descent method; when a machine learning model is adopted, a support vector regression or a 3-layer BP neural network is selected, and the model is deployed in an implanted main processing unit after offline training and model compression; and periodically receiving the measured value of the noninvasive sphygmomanometer, and adjusting a transfer function model through a recursive parameter updating algorithm to adapt to the long-term change of the mechanical properties of the blood vessel wall.
  10. 10. The method of claim 6, further comprising a system self-maintenance step of continuously monitoring the implantable main processing unit power, electrode impedance, and sensor signal consistency, triggering a fail-safe mode when an anomaly is detected, degrading to a conservative stimulation mode and issuing a body surface alarm; The self-maintenance step further includes triggering an immediate preventive drug infusion to alleviate tissue encapsulation when a sustained increase in electrode impedance or a decrease in sensor signal uniformity is detected.

Description

Implantable hypertension self-adaptive closed-loop treatment system and method based on multi-sensor fusion Technical Field The invention relates to the crossing field of medical instruments and biomedical engineering, in particular to an implantable self-adaptive closed-loop hypertension treatment system and method based on multi-sensor fusion. Background Hypertension is one of the most common chronic diseases worldwide, and long-term control failure can lead to various serious complications such as cardiovascular and cerebrovascular diseases, renal failure, fundus diseases and the like. The current mainstream treatment of hypertension mainly comprises life style intervention, oral medication and interventional or instrumental therapies which are emerging in recent years. Although the drug treatment can effectively reduce the blood pressure, the problems of poor compliance, large individual difference, frequent dosage adjustment, difficult tolerance of partial patients for long-term drug administration and the like exist. Interventional therapy such as renal artery sympatholytic (RDN) can reduce sympathetic nerve activity to a certain extent and realize continuous depressurization, but the effect is affected by factors such as anatomy of a patient, operation accuracy, nerve regeneration after operation and the like, the curative effect is not stable enough, and dynamic regulation can not be realized as required. In addition, most of the existing instrument therapies are controlled in an open loop or semi-closed loop, lack of real-time accurate sensing and self-adaptive feedback capability on the blood pressure state, and are difficult to cope with blood pressure fluctuation caused by daily activities, mood fluctuation, body position change and the like. In the aspect of blood pressure monitoring, a non-invasive cuff type electronic sphygmomanometer or an arterial catheter is commonly used in clinic for direct pressure measurement, wherein the non-invasive cuff type electronic sphygmomanometer cannot realize continuous monitoring and is easy to be interfered by the outside, and the arterial catheter is used for invasive operation, is only suitable for short-term critical monitoring and cannot be used as long-term implantation. In recent years, some implanted pressure sensors are used for detecting the pressure in or out of a cardiovascular system, but most of the implanted pressure sensors adopt a single-point measurement mode and are easily affected by local blood flow disturbance, vascular wall motion artifacts and sensor position difference, so that the accuracy of blood pressure reconstruction is insufficient. Particularly in large vessels such as the aorta, because the vessel wall is not a rigid structure, complex transmission characteristics exist between the pressure signal measured by the adventitia and the internal real blood pressure, and the characteristics can be changed due to individual physiological differences, pathological changes (such as arteriosclerosis) and tissue wrapping and remodeling after implantation, and if no targeted compensation is performed, the accuracy of blood pressure measurement and the safety of control can be directly affected. In the field of neuromodulation, low-intensity electrical stimulation to the carotid sinus may reduce sympathetic tone by enhancing baroreceptor reflex, while neural inhibition of the renal artery region may reduce renal-derived boost signals, which in combination may theoretically produce a synergistic hypotensive effect. However, most of the existing implantable nerve stimulation devices are single-target and fixed-parameter working modes, and lack the capability of grading, time sharing and on-demand adjustment according to the real-time state of blood pressure, so that excessive inhibition or response lag is easily caused. Meanwhile, although a slow release pump can be used for long-term administration in the aspect of drug delivery, in the treatment of hypertension, systemic administration is easy to cause hypotension and other side effects, and the problems of uneven drug distribution and insufficient dosage control precision in the aspect of local accurate delivery are required to be solved. In addition, the existing closed-loop treatment system generally depends on external program control equipment to carry out parameter setting and data reading, has limited interaction modes, and is difficult to acquire blood pressure changes in daily activities of patients in time and carry out quick calibration. The mechanical properties of the vessel wall change slowly with age, disease progression and fibrosis of tissue surrounding the implant, and if the blood pressure reconstruction model cannot be updated in time, the accuracy of long-term monitoring and control is reduced. More importantly, the implantable system needs to have certain self-maintenance and fault tolerance capabilities to cope with common faults such as power attenuation, electro