Minimally invasive wearable animal blood pressure monitoring device based on multi-modal signal fusion

Abstract

The invention discloses a minimally invasive wearable animal blood pressure monitoring device based on multi-modal signal fusion, and relates to the technical field of animal blood pressure monitoring, the device comprises a necklace belt, a minimally invasive blood pressure sensing module and a main control module, the main control module is arranged on the side surface of the outer wall of the necklace belt; the main control module is connected with the auxiliary physiological sensing module and the motion state monitoring module through flexible connecting lines, the main control module is connected with a pressure sensor of the minimally invasive blood pressure sensing module through a flexible connecting line, and the minimally invasive blood pressure sensing module is arranged on the side face of the inner wall of the necklace belt. The pressure sensor is connected with the electrode array patch through the connecting rod, and the electrode array patch is connected with the microneedle electrode through the base. By installing the motion state monitoring module and the minimally invasive blood pressure sensing module, severe baseline drift and high-frequency jitter interference caused by body motion to blood pressure signals are eliminated in dynamic activities such as walking or running of animals.

Description

Technical Field

[0001] This invention relates to the field of animal blood pressure monitoring technology, specifically a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion. Background Technology

[0002] Animal blood pressure monitoring plays an important role in pharmacological research and animal health management. Traditional animal blood pressure monitoring methods mainly include invasive arterial cannulation and non-invasive oscillometric methods. Invasive arterial cannulation involves inserting a catheter directly into the artery for continuous blood pressure monitoring. Although it has high accuracy, it is complex to operate, requires anesthesia, and is prone to infection and animal stress response, making it unsuitable for long-term or daily monitoring. Non-invasive oscillometric methods usually measure blood pressure by inflating a cuff to compress the artery. However, this method is easily affected by the animal's fur, skin characteristics, movement status, and changes in body position, resulting in poor measurement stability. Moreover, it is almost impossible to obtain reliable data when the animal is active.

[0003] Patent CN109770879B discloses an intelligent animal blood pressure monitoring system and its monitoring method, which enables continuous and accurate measurement of animal blood pressure.

[0004] The signal processing module of the aforementioned patent calculates animal blood pressure in real time based on animal electrocardiogram and pulse wave signals. It is simple, fast, efficient, and low-cost. It can continuously calculate and record animal blood pressure through the signal processing module, achieving continuous and accurate measurement of animal blood pressure. However, there is still room for optimization in blood pressure measurement when animals are exercising and changing their body position.

[0005] Therefore, this application proposes a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion that can monitor blood pressure values ​​in animals during their active state. Summary of the Invention

[0006] The purpose of this invention is to provide a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion, so as to solve the technical problem that the traditional blood pressure monitoring method mentioned in the background art is easily affected by the animal's fur, skin characteristics, movement state and body position changes, resulting in poor stability of the measurement results.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion, comprising a collar, a multimodal sensing unit, and a main control module. The collar is made of medical-grade silicone. The main control module is disposed on the outer side of the collar and has a cuboid structure. The outer shell of the main control module is a thin metal shell. The main control module is internally connected to the multimodal sensing unit via flexible connecting lines. The multimodal sensing unit includes a minimally invasive blood pressure sensing module, an auxiliary physiological sensing module, and a motion state monitoring module. The minimally invasive blood pressure sensing module is disposed on the inner side of the collar. The main control module is connected to the pressure sensor of the minimally invasive blood pressure sensing module via a flexible connecting line embedded inside the collar. The pressure sensor is connected to an electrode array patch via a connecting rod. The electrode array patch is connected to a microneedle electrode via a base. The main control module is connected to the auxiliary physiological sensing module and the motion state monitoring module via flexible connecting lines.

[0008] Preferably, the minimally invasive blood pressure sensing module includes a pressure sensor, an electrode array patch, and a microneedle electrode. A connecting rod is provided at the top of the outer wall of the pressure sensor, and the other end of the connecting rod is connected to the electrode array patch. A base is provided at the top of the outer wall of the electrode array patch, and the base is connected to the microneedle electrode through an electrode interface. The electrode interface is located at the bottom of the outer wall of the microneedle electrode, and a microneedle tip is provided at the top of the outer wall of the microneedle electrode.

[0009] Preferably, the connecting rod is provided with a flexible connecting wire inside, one end of which is connected to the electrode array patch, and the other end of which is connected to the pressure sensor.

[0010] Preferably, the length of the microneedle is set between 0.5 and 1.5 mm, and it is used to penetrate the stratum corneum and part of the epidermis to detect pressure fluctuations in subcutaneous tissue fluid.

[0011] Preferably, the motion monitoring module is located on the outer side of the collar band. The motion monitoring module is connected to the main control module via a flexible connecting line. The motion monitoring module is used to transmit the collected motion signals to the main control module. The motion monitoring module includes a three-axis accelerometer and a three-axis gyroscope. The three-axis accelerometer is used to measure the linear acceleration of the animal in three-dimensional space, and the three-axis gyroscope is used to measure the angular velocity of the animal around the three coordinate axes.

[0012] Preferably, the main control module is connected to the auxiliary physiological sensing module via a flexible connecting line. The auxiliary physiological sensing module is located on the outer side of the collar band and includes a skin impedance sensor and a local temperature sensor. The skin impedance sensor is used to measure the impedance signal in the animal's skin, and the local temperature sensor is used to measure the local temperature of the animal's skin.

[0013] Preferably, the main control module includes a processor and a battery. The processor has an interface on its outside and a Bluetooth communication protocol and an individualized calibration model inside. The processor's interface is connected to a flexible connection cable connected to the main control module. The processor is connected to the battery via a power cable. The processor is used to synchronize and preprocess the minimally invasive blood pressure signal, motion signal, and auxiliary signal. The processor dynamically separates motion deviations from the minimally invasive blood pressure signal and extracts feature parameters from the processed blood pressure waveform to calculate the blood pressure value.

[0014] Preferably, the processor inputs the extracted feature parameters into an internally preset individualized calibration model. The individualized calibration model receives reference blood pressure values ​​input from an external terminal for incremental learning and dynamic optimization. The individualized calibration model is used to adapt to the physiological differences of different animals based on the extracted feature parameters, and calculates and outputs systolic blood pressure, diastolic blood pressure, and mean blood pressure in real time. The main control module is connected to an external smart terminal via Bluetooth communication protocol, which is used to wirelessly transmit the processed blood pressure data to the external terminal.

[0015] Preferably, the top of the outer wall of the main control module is provided with a connection port, which is connected to the battery via a connection cable for charging the battery. The front side of the outer wall of the main control module is provided with a display screen and indicator lights, which are connected to the processor via connection cables.

[0016] Preferably, the top of the outer wall of the collar strap is provided with a first connecting groove and a second connecting groove. The collar strap is connected to the auxiliary physiological sensing module through the first connecting groove and to the motion state monitoring module through the second connecting groove. The two ends of the collar strap are respectively provided with buckles and slots. A button is provided at the top of the end of the collar strap with the slot.

[0017] Compared with the prior art, the beneficial effects of the present invention are:

[0018] 1. This invention, by installing a collar, microneedles, pressure sensors, microneedle electrodes, and electrode array patches, enables long-term direct detection of fluid pressure fluctuations in animal subcutaneous tissue under minimally invasive conditions. It can acquire high-quality and interference-resistant original arterial blood pressure-related physiological signals, avoid interference from body hair and skin characteristics, and ensure the accuracy of measurement from the signal source. It solves the technical problem of inaccurate and unreliable measurement results caused by signal attenuation or distortion in traditional non-invasive methods.

[0019] 2. This invention, by installing an auxiliary physiological sensing module, achieves synchronous monitoring and temperature compensation of animal skin impedance and local temperature. It can evaluate the contact quality between the sensing patch and the skin tissue in real time, ensuring the reliability of the physical basis of signal acquisition and avoiding data failure caused by poor contact from the source. At the same time, through the temperature compensation mechanism, it eliminates the thermal drift effect caused by changes in environmental and body surface temperature on the pressure sensor readings, solves the problem of measurement error introduced by improper wearing or changes in contact state due to animal activity, and improves the reliability and physical authenticity of the data.

[0020] 3. This invention, by installing a motion state monitoring module, achieves real-time capture and motion compensation of the animal's three-dimensional linear acceleration and angular velocity, providing the main control module with accurate motion reference signals. This enables the processor to dynamically identify, separate, and eliminate motion artifacts from mixed blood pressure signals through an adaptive filtering algorithm. This solves the serious baseline drift and high-frequency jitter interference caused by the animal's body movement during dynamic activities such as walking or running, and breaks through the technical bottleneck of traditional equipment being unable to accurately monitor blood pressure when the animal is not stationary.

[0021] 4. This invention, equipped with a display screen, indicator lights, a main control module, and a processor, enables local data display, synchronous processing of multimodal signals, calculation of blood pressure values ​​based on an individualized calibration model, and wireless transmission. It provides an intuitive user interface and dynamically optimizes the model through incremental learning to adapt to physiological differences between different species and individuals, ultimately outputting continuous, accurate, and personalized blood pressure values. This solves the problem that general-purpose devices struggle to handle large measurement deviations caused by individual differences. Furthermore, it enables remote wireless monitoring and local intuitive display of monitoring data, greatly improving the convenience and experience of pet health management and scientific research. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0023] Figure 2 This is a schematic diagram of the minimally invasive blood pressure sensing module structure of the present invention;

[0024] Figure 3 This is a schematic diagram of the microneedle electrode structure of the present invention;

[0025] Figure 4 This is a side view of the minimally invasive blood pressure sensing module of the present invention.

[0026] Figure 5 This is a schematic diagram of the collar strap structure of the present invention;

[0027] Figure 6 This is a schematic diagram of the multimodal sensing unit of the present invention;

[0028] Figure 7 This is a flowchart of the execution steps of the main control module of the present invention;

[0029] Figure 8 This is a flowchart illustrating the overall steps of the present invention.

[0030] In the diagram: 1. Main control module; 2. Neckband; 3. Connection port; 4. Minimally invasive blood pressure sensor module; 5. Auxiliary physiological sensor module; 6. Motion status monitoring module; 7. Display screen; 8. Indicator light; 9. Buckle; 10. Button; 11. Base; 12. Microneedle electrode; 13. Connecting rod; 14. Pressure sensor; 15. First connecting slot; 16. Second connecting slot; 17. Slot; 18. Electrode interface; 19. Microneedle tip; 20. Electrode array patch. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," "outer," "front end," "rear end," "both ends," "one end," and "the other end," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0033] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installed," "equipped with," "connected," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0034] Please see Figure 1 , Figure 2 , Figure 3 , Figure 5 and Figure 6This invention provides an embodiment of a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion. A main control module 1 is disposed on the outer side of the collar 2. The main control module 1 is internally connected to a multimodal sensing unit via a flexible connecting line. The multimodal sensing unit includes a minimally invasive blood pressure sensing module 4, an auxiliary physiological sensing module 5, and a motion state monitoring module 6. The minimally invasive blood pressure sensing module 4 is disposed on the inner side of the collar 2. The main control module 1 is connected to a pressure sensor 14 of the minimally invasive blood pressure sensing module 4 via a flexible connecting line embedded inside the collar 2. The pressure sensor 14 is connected to an electrode array patch 20 via a connecting rod 13. The electrode array patch 20 is connected to a microneedle electrode 12 via a base 11. The main control module 1 is connected to the auxiliary physiological sensing module 5 and the motion state monitoring module 6 via flexible connecting lines.

[0035] Next, the operator first wraps the collar 2 around the dog's neck, aligning the microneedles 19 of the minimally invasive blood pressure sensing module 4 on both sides of the inner wall of the collar 2 with the carotid artery skin of the dog. Then, the buckle 9 is inserted into the slot 17, and the depth of insertion of the buckle 9 into the slot 17 is adjusted so that a finger can be inserted between the dog's neck and the collar 2. At this point, the microneedle 19 just passes through the short hair on the dog's neck and contacts the skin. The length of the microneedle 19 is set to 0.8mm. Then, the skin impedance sensor of the auxiliary physiological sensing module 5 on the collar 2 starts to work. The skin impedance sensor applies a small and safe alternating current signal to the skin tissue, usually with a frequency of 1kHz to 100kHz. Then, the impedance value is calculated by measuring the voltage drop exhibited by the tissue. The contact quality between the microneedle 19 and the skin tissue is evaluated by measuring the dog's impedance. The processor reads the impedance value on the skin impedance sensor in real time. If the impedance value is in the range of 20kΩ to 50kΩ and the wave... If the amplitude of the movement is less than 10% and the stable time is more than 3 seconds, the processor determines that the contact is good. The processor controls the indicator light 8 to light up green through the connection line, ensuring that the physical basis of signal acquisition is reliable and avoiding signal distortion or data failure caused by poor contact from the source. Then, the minimally invasive blood pressure sensing module 4 is activated. The electrode array patch 20 synchronously collects the fluctuation signal of the subcutaneous tissue hydraulic pressure in the dog's neck at a sampling rate of 1kHz. The microneedle 19 detects the small pressure fluctuation of the subcutaneous interstitial fluid and transmits the physiological pressure signal to the pressure sensor 14 through the microneedle electrode 12. The pressure sensor 14 receives the physical pressure signal transmitted from the microneedle electrode 12 and converts the physical pressure signal into the corresponding electrical signal. The pressure sensor 14 is finally transmitted to the processor of the main control module 1 through the flexible connection line, directly acquiring the physiological signal related to the animal's arterial blood pressure, avoiding interference from body hair and skin characteristics, and providing a high-quality signal source for high-precision measurement.

[0036] Simultaneously, the motion monitoring module 6 starts operating. This module contains a three-axis gyroscope and a three-axis accelerometer. When the accelerometer's output amplitude stabilizes at 1.0G and the gyroscope's output angular velocity approaches zero, the processor determines the dog is asleep or stationary. After the contact quality verification is passed, the auxiliary physiological sensing module 5 switches to periodic monitoring, sampling every 10 seconds. The local temperature sensor continuously measures skin temperature at a frequency of 1Hz, providing ambient temperature monitoring. Then, the processor in the main control module 1 preprocesses the minimally invasive blood pressure signal. The processor contains a bandpass filter to filter out high-frequency electronic noise. The processor extracts characteristic parameters from the processed blood pressure waveform, including systolic peak, diastolic trough, and dicrotic notch. The value corresponds to the highest pressure, the diastolic trough corresponds to the lowest pressure, and the dicrotic notch corresponds to the characteristic indentation caused by aortic valve closure. The processor inputs the extracted feature parameters into the individualized standard model to calculate the values ​​of systolic blood pressure, diastolic blood pressure, and mean arterial pressure (MAP), MAP=(1 / T)∫P(t)dt, where T is a complete cardiac cycle, the cardiac cycle is the time interval between adjacent systolic peaks, and P(t) is a function of blood pressure change over time. The calculated blood pressure value is then sent to the paired smartphone APP via the built-in Bluetooth communication protocol. At the same time, the processor transmits the blood pressure value to the display screen 7 via the connection cable for easy viewing by the user. This realizes wireless remote monitoring and local intuitive display of data, greatly improving the convenience of managing pet cardiovascular health and the user experience.

[0037] Please see Figure 1 , Figure 2 , Figure 4 , Figure 5 and Figure 7This invention provides an embodiment of a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion. A main control module 1 is disposed on the outer side of the collar 2. The main control module 1 is internally connected to a multimodal sensing unit via a flexible connecting wire. The multimodal sensing unit includes a minimally invasive blood pressure sensing module 4, an auxiliary physiological sensing module 5, and a motion state monitoring module 6. The minimally invasive blood pressure sensing module 4 is disposed on the inner side of the collar 2. The main control module 1 is connected to a pressure sensor 14 of the minimally invasive blood pressure sensing module 4 via a flexible connecting wire embedded inside the collar 2. The pressure sensor 14... The connecting rod 13 is connected to the electrode array patch 20, and the electrode array patch 20 is connected to the microneedle electrode 12 through the base 11. The main control module 1 is connected to the auxiliary physiological sensing module 5 and the motion state monitoring module 6 through a flexible connecting line. The motion state monitoring module 6 is set on the outer side of the collar band 2. The motion state monitoring module 6 is used to transmit the collected motion signals to the main control module 1. The motion state monitoring module 6 includes a triaxial accelerometer and a triaxial gyroscope. The triaxial accelerometer is used to measure the linear acceleration of the animal in three-dimensional space, and the triaxial gyroscope is used to measure the angular velocity of the animal around the three coordinate axes.

[0038] Furthermore, when the dog begins to walk or jog, its neck will sway rhythmically. The triaxial accelerometer in the motion monitoring module 6 installed on the collar 2 captures the periodic acceleration changes in the forward and backward directions in real time, with the acceleration amplitude stabilizing within the range of 0.5-1.5g. The triaxial gyroscope captures the regular pitch angular velocity, with the amplitude stabilizing within the range of 50-150°. o / s, the processor of the main control module 1 analyzes the motion parameters transmitted from the motion state monitoring module 6. When continuous and regular low-frequency motion characteristics are detected, it is determined that the animal has entered a mild motion state. Then, the minimally invasive blood pressure sensing module 4 continuously collects the pressure fluctuation signal of subcutaneous tissue fluid at a sampling rate of 1kHz. Under the influence of motion, interference appears in the original signal. The interference is low-frequency baseline drift with body undulation and high-frequency jitter superimposed on the regular blood pressure waveform. Low-frequency baseline drift refers to the slow and non-physiological up and down fluctuation of the baseline of the collected blood pressure waveform over time. At this time, the main control module 1 starts the spatial division multiplexing technology. Spatial division multiplexing technology is a method to process multiple signals simultaneously through the independence of spatial distribution. Spatial division multiplexing technology can analyze the signal quality of each microneedle electrode 12 in the electrode array patch 20 in parallel. By calculating the cross-correlation between signals, the unit with the least influence from muscle movement and the highest signal-to-noise ratio is selected as the main signal source, and the remaining units are used as redundant backups. At the same time, the original signal passes through a 0.5-20Hz bandpass filter to initially filter out some high-frequency noise and extremely low-frequency drift.

[0039] Then, the processor of the main control module 1 calls the built-in adaptive filtering algorithm. The adaptive filtering algorithm takes the motion feature vector provided by the motion state monitoring module as the reference input. The motion feature vector contains motion frequency, amplitude and direction information. It performs real-time noise reduction processing on the preprocessed original mixed blood pressure signal. The filter can estimate and subtract motion interference components from the mixed signal and output a processed blood pressure waveform. The filter can adjust the compensation strategy in real time for different motion intensities, so as to retain the real blood pressure change information to the maximum extent in complex dynamic environments.

[0040] Simultaneously, the physiological sensing module 5 operates, and the skin impedance sensor continuously detects and transmits the detected signal to the processor. If the impedance fluctuation caused by slight skin stretching is within the preset threshold, the processor determines that the contact quality remains reliable. If it exceeds the threshold, the processor marks the data for that period as "reduced reliability". At the same time, the local temperature sensor detects the slow rise in skin temperature caused by movement and corrects the reading of the pressure sensor 14 in real time through enhanced temperature compensation. The compensation formula is P=Praw*[1+α*(T-T0)+β*(dT / dt)], where β is the temperature change rate coefficient and dT / dt is the temperature change rate over time. Temperature compensation eliminates the influence of thermal drift on measurement accuracy and ensures the stability of the sensing interface and the physical authenticity of the data even under dynamic conditions.

[0041] The processor uses a feature recognition algorithm to identify and capture the systolic peak and diastolic trough of the blood pressure waveform after motion compensation and temperature restoration. Then, the processor inputs the extracted feature parameters into the individualized calibration model to calculate the systolic pressure, diastolic pressure and mean pressure in real time. The processor then transmits the calculated blood pressure value to the display screen 7 through the connection cable, and at the same time transmits the blood pressure value to the external paired mobile APP through the Bluetooth communication protocol.

[0042] Please see Figure 1 , Figure 2 and Figure 8This invention provides an embodiment of a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion. A main control module 1 is disposed on the outer side of the collar 2. The main control module 1 is internally connected to a multimodal sensing unit via flexible connecting lines. The multimodal sensing unit includes a minimally invasive blood pressure sensing module 4, an auxiliary physiological sensing module 5, and a motion state monitoring module 6. The minimally invasive blood pressure sensing module 4 is disposed on the inner side of the collar 2. The main control module 1 is connected to the pressure sensor 14 of the minimally invasive blood pressure sensing module 4 via flexible connecting lines embedded inside the collar 2. The main control module 1 is also connected to the auxiliary physiological sensing module 5 and the motion state monitoring module 6 via flexible connecting lines. The main control module 1 includes a processor and a battery. The processor has an interface on its outside and a Bluetooth communication protocol and an individualized calibration model inside. The processor's interface is connected to a flexible connection cable in the main control module 1. The processor is connected to a battery via a power cord. The processor dynamically separates motion deviations from the minimally invasive blood pressure signal. The processor extracts feature parameters from the processed blood pressure waveform to calculate the blood pressure value. The processor inputs the extracted feature parameters into the internally preset individualized calibration model. The individualized calibration model is used to adapt to the physiological differences of different animals based on the extracted feature parameters, and calculates and outputs systolic blood pressure, diastolic blood pressure, and mean blood pressure in real time. The main control module 1 is connected to an external smart terminal via the Bluetooth communication protocol. The Bluetooth communication protocol is used to wirelessly transmit the processed blood pressure data to the external terminal.

[0043] Furthermore, when the dog begins high-intensity exercise, the motion monitoring module 6 detects data in real time. The three-axis accelerometer continuously detects accelerations as high as 3-6G, accompanied by instantaneous peak impacts reaching 8G. The three-axis gyroscope detects rapid attitude changes and axial rotation, with angular velocities consistently exceeding 300°. o / s, the motion state monitoring module 6 transmits the detected data to the processor, while the minimally invasive blood pressure sensing module 4 preserves the signal. This is because the dog's violent body swaying and muscle vibration can cause the subcutaneous tissue pressure signal to be submerged by powerful motion artifacts, and the original waveform loses its periodic structure. At this time, the processor does not process the data in real time through the conventional 0.5-20Hz bandpass filter, but instead directly saves the original sampling data of all sensing units with 16-bit precision, so as to preserve complete and distortion-free information for subsequent analysis.

[0044] The processor then evaluates the raw sampled data in real time. The evaluation algorithm analyzes the periodic information in the motion data from the motion state monitoring module 6, the skin impedance value from the auxiliary physiological sensing module 5, and the changing blood pressure waveform in parallel, and calculates the signal quality index (SQI). When the SQI is below 0.4, the processor marks this data as low reliability. Then, the processor evaluates the range of blood pressure output, such as 100-130 mmHg, without outputting a specific blood pressure value. Without providing a potentially distorted accurate reading, the processor provides the user with macroscopic and valuable trend information on the animal's physiological state.

[0045] Finally, when the exercise intensity decreases, the triaxial accelerometer captures the periodic acceleration changes in the forward and backward directions. The vector amplitude stabilizes within the range of 0.5-1.5g. The processor determines that it is in walking mode, exits the high-intensity exercise mode, and resumes regular pulsed blood pressure monitoring and specific blood pressure output, ensuring the continuity of monitoring and being able to fully capture the complete physiological dynamic changes from high-intensity exercise to full recovery.

[0046] Please see Figure 1 , Figure 2 , Figure 7 and Figure 8 The present invention provides an embodiment of a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion. The main control module 1 has a connection port 3 at the top of its outer wall. The connection port 3 is connected to a battery via a connection cable for charging the battery. The main control module 1 has a display screen 7 and an indicator light 8 on its front side. The display screen 7 and the indicator light 8 are connected to a processor via a connection cable. The collar strap 2 is connected to an auxiliary physiological sensing module 5 via a first connection slot 15 and to a motion state monitoring module 6 via a second connection slot 16. The collar strap 2 has a buckle 9 and a slot 17 at its two tail ends, respectively. A button 10 is provided at the top of the tail end of the collar strap 2 with the slot 17.

[0047] Furthermore, the operator monitors the cat's blood pressure. Since cats' skin is generally thinner and more sensitive than dogs', the length of the microneedle 19 is set to 0.5mm. The operator wraps the collar 2 around the cat's neck and adjusts its position so that the inner minimally invasive blood pressure sensing module 4 is directly facing the cat's carotid artery epidermis. Then, the tightness of the collar 2 is adjusted so that the gap between the minimally invasive blood pressure sensing module 4 and the cat's skin can accommodate a finger without significant pressure. Then, the device is initialized and the contact quality is assessed. After the main control module 1 is started, it first drives the skin impedance sensor of the auxiliary physiological sensing module 5 to work. The skin impedance sensor monitors the amplitude of the impedance signal in real time and transmits the signal to the processor. If the reading is stable within the preset reliability threshold range, the processor determines that the microneedle 19 is in good contact with the cat's skin tissue surface, and the indicator light 8 of the main control module 1 shows a green ready state.

[0048] After the processor confirms good contact, the minimally invasive blood pressure sensing module 4 begins to continuously collect pressure fluctuation signals of subcutaneous tissue fluid at a sampling rate of 1kHz. At the same time, the motion state monitoring module 6 is activated. The inertial measurement unit built into the motion state monitoring module 6 captures the posture changes of the cat's head and neck in real time. At this time, the processor calls the internally stored feline motion feature library, which contains feline-specific motion pattern parameters, such as light jumping, curling up and lying prone, and rapid head rotation. The processor compares and matches the real-time collected motion data with the feature library to identify the current motion type and generate a corresponding motion interference model. The motion interference model can improve the device's recognition accuracy of motion artifacts caused by feline-specific behaviors, such as sudden jumping followed by instantaneous stillness, laying the foundation for the next step of accurate compensation.

[0049] Then the processor performs dynamic identification and compensation of motion artifacts. The processor of the main control module 1 is equipped with an adaptive filtering algorithm. The adaptive filtering algorithm takes the motion interference model optimized by the species feature library provided by the motion state monitoring module 6 as the reference input and processes the noisy raw blood pressure signal from the minimally invasive blood pressure sensing module 4 in real time. The adaptive filtering algorithm can separate and eliminate artifact components related to the cat's movement posture from the mixed signal and output a blood pressure waveform that truly reflects the changes in arterial pressure. This allows the device to maintain high signal quality even when the cat is engaged in moderate-intensity activities, such as pacing and playing at home, thus overcoming the bottleneck of traditional devices in dynamic monitoring.

[0050] Finally, the main control module 1 performs feature extraction, model calculation, and output on the signal. From the processed blood pressure waveform, the main control module 1 extracts feature parameters such as waveform area and peak interval. The processor inputs the feature parameters into an individualized calibration model adapted for cats in real time. The model can incrementally learn by inputting reference blood pressure values ​​through an external terminal, thereby continuously optimizing the blood pressure calculation accuracy for individual cats. The processor calculates systolic blood pressure, diastolic blood pressure, and mean blood pressure in real time through the individualized calibration model. The individualized calibration model can solve the impact of physiological differences between different animal individuals on the accuracy of blood pressure measurement. Finally, the processor transmits the calculated blood pressure value wirelessly to a smartphone APP via Bluetooth. While ensuring that the feline wearer is comfortable and unrestrained, it provides users with continuous, accurate, and reliable blood pressure monitoring data, greatly facilitating chronic disease management and health tracking. When blood pressure measurement is not required, the operator can press button 10 to release the locking of buckle 9 and remove collar strap 2.

[0051] Please see Figure 1 , Figure 2 , Figure 7 and Figure 8The present invention provides an embodiment of a minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion. The minimally invasive blood pressure sensing module 4 includes a pressure sensor 14, an electrode array patch 20, and a microneedle electrode 12. A connecting rod 13 is provided at the top of the outer wall of the pressure sensor 14, and the other end of the connecting rod 13 is connected to the electrode array patch 20. A base 11 is provided at the top of the outer wall of the electrode array patch 20, and the base 11 is connected to the microneedle electrode 12 through an electrode interface 18. The electrode interface 18 is located at the bottom of the outer wall of the microneedle electrode 12, and a microneedle tip 19 is provided at the top of the outer wall of the microneedle electrode 12. A flexible connecting wire is provided inside the connecting rod 13, one end of which is connected to the electrode array patch 20, and the other end of which is connected to the pressure sensor 14.

[0052] Furthermore, the operator monitors the blood pressure of laboratory mice. First, a miniature collar 2 is selected based on the size characteristics of the mice. Mice typically weigh 20-30g and have a neck circumference of about 3-4cm. The width of collar 2 is set to 5mm to ensure comfortable wear without affecting the normal activity and physiological state of the mice. Before the experiment begins, the operator puts collar 2 on the neck of the mice. When wearing it, it is necessary to ensure that collar 2 is not too tight, so that there is no obvious gap between collar 2 and the neck skin but it can still move slightly. After wearing, the operator starts the device. The device first performs initialization. The skin impedance sensor in the auxiliary physiological sensing module 5 starts working. The skin impedance sensor scans AC signals in the range of 1kHz to 100kHz to detect the impedance value of the mouse skin. If the impedance value is stable and meets the preset range, the processor determines that the contact is good and the initialization is successful. The indicator light 8 emits a steady green light, and the device enters the ready state.

[0053] Once the device is operational, the minimally invasive blood pressure sensing module 4 begins to continuously collect minute pressure fluctuations in the subcutaneous tissue fluid of mice at a sampling rate of 1kHz. Simultaneously, the motion monitoring module 6 is activated to detect the unique high-frequency and small-amplitude motion characteristics of mice, such as rapid darting, continuous biting behavior, and running on the wheel. The motion data is transmitted in real time to the processor of the main control module 1 via a connecting cable. Because the mice are active and vigorous, the motion artifacts generated by the detection can seriously interfere with the original blood pressure signal. Therefore, a motion noise model optimized for rodents is set in the processor. The processor uses an adaptive filtering algorithm to input the acceleration and angular velocity data provided by the motion monitoring module 6 as a reference into the motion noise model. It identifies, separates, and filters out interference components related to mouse movement from the mixed blood pressure signal, significantly improving the signal-to-noise ratio of the blood pressure signal in the dynamic environment and ensuring that valuable blood pressure waveform data can still be obtained even when the mice are not stationary.

[0054] The processor extracts feature parameters such as waveform area and peak interval from the blood pressure waveform after motion artifact compensation. The processor inputs the feature parameters into the internal individualized calibration model. The individualized calibration model can perform incremental learning and dynamic optimization using reference values ​​and synchronously acquired feature parameters, so that the output blood pressure value is increasingly consistent with the real physiological state of the mice. The individualized calibration model finally calculates and outputs accurate estimates of the systolic blood pressure, diastolic blood pressure and mean arterial pressure of the mice in real time, realizing personalized and adaptive monitoring and effectively overcoming the influence of physiological differences between different mice on measurement accuracy.

[0055] The processed data is wirelessly transmitted to the laboratory's central computer or the smart terminal equipped by researchers via Bluetooth communication protocol. The connection port 3 set on the top of the outer wall of the main control module 1 charges the built-in battery, enabling researchers to conduct long-term continuous monitoring of mice for several days or even weeks, fully recording the diurnal rhythm of blood pressure and the response to different experimental stimuli. Under the premise of minimizing interference with experimental animals, it provides continuous and accurate blood pressure physiological datasets for cutting-edge scientific research such as cardiovascular pharmacology and pathophysiology.

[0056] Working principle: First, the operator puts the collar 2 on the animal's neck and adjusts the tightness by using the buckle 9 and the slot 17, so that the microneedle electrode 12 of the minimally invasive blood pressure sensing module 4 penetrates the epidermis and contacts the subcutaneous tissue. After the device is started, the skin impedance sensor of the auxiliary physiological sensing module 5 first verifies the contact quality. By applying a safe AC signal, the impedance value is detected. If it is stable within the preset range, such as 20kΩ to 50kΩ, the processor determines that the contact is reliable, and the indicator light 8 shows a green ready state.

[0057] Then, the minimally invasive blood pressure sensing module 4 continuously collects the pressure fluctuation signal of the subcutaneous tissue fluid at a sampling rate of 1kHz. The physical pressure signal is converted into an electrical signal through the microneedle electrode 12 and the pressure sensor 14. At the same time, the motion state monitoring module 6 captures the animal's motion data in real time through the triaxial accelerometer and gyroscope, and assists the physiological sensing module 5 in monitoring skin impedance and local temperature. The processor of the main control module 1 performs synchronous preprocessing on multiple signals, filters out high-frequency noise through the bandpass filter, and uses the motion characteristics as reference input through the internal preset adaptive filtering algorithm to dynamically separate motion artifacts and retain the true blood pressure waveform. At the same time, the local temperature sensor provides real-time temperature compensation to correct the reading of the pressure sensor 14 and eliminate the influence of environmental thermal drift.

[0058] The processed blood pressure waveform is identified by the feature extraction algorithm inside the processor to identify the systolic peak and diastolic trough characteristic parameters. These characteristic parameters are then input into an individualized calibration model to calculate systolic blood pressure, diastolic blood pressure, and mean arterial pressure in real time. The individualized calibration model can incrementally learn from reference values ​​input from an external terminal and dynamically optimize to adapt to individual differences in different animals. Finally, the blood pressure data is displayed locally on display screen 7 and wirelessly transmitted to a smart terminal via Bluetooth, enabling remote monitoring and long-term health tracking.

[0059] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion, comprising a collar (2), a multimodal sensing unit, and a main control module (1), characterized in that: The collar band (2) is made of medical-grade silicone. A main control module (1) is provided on the outer side of the collar band (2). The main control module (1) has a cuboid structure and a thin metal shell. The main control module (1) is connected to a multimodal sensing unit through a flexible connecting line. The multimodal sensing unit includes a minimally invasive blood pressure sensing module (4), an auxiliary physiological sensing module (5), and a motion state monitoring module (6). The minimally invasive blood pressure sensing module (4) is located on the collar band (2). On the inner wall side, the main control module (1) is connected to the pressure sensor (14) of the minimally invasive blood pressure sensing module (4) through a flexible connecting line embedded in the collar strap (2). The pressure sensor (14) is connected to the electrode array patch (20) through a connecting rod (13). The electrode array patch (20) is connected to the microneedle electrode (12) through a base (11). The main control module (1) is connected to the auxiliary physiological sensing module (5) and the motion state monitoring module (6) through a flexible connecting line.

2. The minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 1, characterized in that: The minimally invasive blood pressure sensing module (4) includes a pressure sensor (14), an electrode array patch (20), and a microneedle electrode (12). A connecting rod (13) is provided at the top of the outer wall of the pressure sensor (14), and the other end of the connecting rod (13) is connected to the electrode array patch (20). A base (11) is provided at the top of the outer wall of the electrode array patch (20), and the base (11) is connected to the microneedle electrode (12) through an electrode interface (18). The electrode interface (18) is located at the bottom of the outer wall of the microneedle electrode (12), and a microneedle tip (19) is provided at the top of the outer wall of the microneedle electrode (12).

3. The minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 2, characterized in that: The connecting rod (13) is provided with a flexible connecting line inside. One end of the flexible connecting line is connected to the electrode array patch (20), and the other end of the flexible connecting line is connected to the pressure sensor (14).

4. The minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 2, characterized in that: The length of the microneedle (19) is set between 0.5 and 1.5 mm, and it is used to penetrate the stratum corneum and part of the epidermis to detect pressure fluctuations in subcutaneous tissue fluid.

5. The minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 1, characterized in that: The motion monitoring module (6) is set on the outer side of the collar strap (2). The motion monitoring module (6) is connected to the main control module (1) through a flexible connecting line. The motion monitoring module (6) is used to transmit the collected motion signals to the main control module (1). The motion monitoring module (6) includes a three-axis accelerometer and a three-axis gyroscope. The three-axis accelerometer is used to measure the linear acceleration of the animal in three-dimensional space, and the three-axis gyroscope is used to measure the angular velocity of the animal around the three coordinate axes.

6. The minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 5, characterized in that: The main control module (1) is connected to the auxiliary physiological sensing module (5) via a flexible connecting line. The auxiliary physiological sensing module (5) is located on the outer side of the collar band (2). The auxiliary physiological sensing module (5) includes a skin impedance sensor and a local temperature sensor. The skin impedance sensor is used to measure the impedance signal in the animal's skin, and the local temperature sensor is used to measure the local temperature of the animal's skin.

7. A minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 6, characterized in that: The main control module (1) includes a processor and a battery. An interface is provided on the outside of the processor. The processor is equipped with a Bluetooth communication protocol and an individualized calibration model. The interface of the processor is connected to a flexible connection line connected to the main control module (1). The processor is connected to the battery through a power line. The processor is used to synchronize and preprocess the minimally invasive blood pressure signal, motion signal and auxiliary signal. The processor dynamically separates motion deviation from the minimally invasive blood pressure signal. The processor extracts feature parameters from the processed blood pressure waveform to calculate the blood pressure value.

8. The minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 7, characterized in that: The processor inputs the extracted feature parameters into the internally preset individualized calibration model. The individualized calibration model receives the reference blood pressure value input from the external terminal for incremental learning and dynamic optimization. The individualized calibration model is used to adapt to the physiological differences of different animals according to the extracted feature parameters, and calculates and outputs systolic blood pressure, diastolic blood pressure and mean blood pressure in real time. The main control module (1) is connected to the external smart terminal through the Bluetooth communication protocol. The Bluetooth communication protocol is used to wirelessly transmit the processed blood pressure data to the external terminal.

9. A minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 8, characterized in that: The main control module (1) has a connection port (3) at the top of its outer wall. The connection port (3) is connected to the battery via a connection cable for charging the battery. The main control module (1) has a display screen (7) and an indicator light (8) on the front side of its outer wall. The display screen (7) and the indicator light (8) are connected to the processor via a connection cable.

10. A minimally invasive wearable animal blood pressure monitoring device based on multimodal signal fusion according to claim 5, characterized in that: The collar strap (2) has a first connecting groove (15) and a second connecting groove (16) at the top of its outer wall. The collar strap (2) is connected to the auxiliary physiological sensing module (5) through the first connecting groove (15) and to the motion state monitoring module (6) through the second connecting groove (16). The collar strap (2) has a buckle (9) and a slot (17) at the tail ends on both sides respectively. The collar strap (2) with the slot (17) has a button (10) at the top of its tail end.

Images