A wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels

Abstract

The application discloses a wearable intelligent ultrasonic system for dynamically monitoring deep blood vessels, and relates to the technical field of intelligent wearable devices. The system acquires an initial image in an initial monitoring mode through an ultrasonic sensing module; a central processing module automatically identifies and locates the deep blood vessels in the image, analyzes the dynamic motion trajectory of the deep blood vessels with the breathing and heart cycle, switches a secondary monitoring mode according to the dynamic motion trajectory, activates a matched ultrasonic array element to perform directional signal acquisition, synchronously acquires the structure and hemodynamic echo signals of the blood vessels, generates target data containing quantitative parameters and a dynamic tracking image fused with the motion trajectory, and transmits the data and the image to a data analysis module after encryption and compression. The data analysis module extracts features and compares them with medical reference thresholds to realize abnormality detection and early warning. The system realizes long-term dynamic monitoring of deep blood vessels in an automatic, accurate and low-power consumption manner.

Description

Technical Field

[0001] This invention relates to the field of intelligent wearable device technology, specifically a wearable intelligent ultrasound system for dynamically monitoring deep blood vessels. Background Technology

[0002] In the field of medical imaging, with the rapid development of wearable smart devices, artificial intelligence algorithms, wireless data transmission and signal processing technologies, research based on wearable ultrasound systems has gradually emerged, becoming an important direction for long-term dynamic monitoring of internal tissues and blood vessels, and providing a new technical path for clinical diagnosis and long-term home health monitoring.

[0003] Current ultrasound detection technologies and equipment for long-term dynamic monitoring of deep blood vessels in the human body still have many significant shortcomings, making it difficult to meet the clinical needs for accurate, continuous, and low-cost monitoring. Deep blood vessels are located deep within the body's tissue layers, and traditional ultrasound detection methods are easily affected by tissue attenuation, scattering, and stray signals when identifying and locating them, resulting in poor image quality and signal-to-noise ratio. This makes it difficult to accurately capture the location and real-time dynamic changes of target blood vessels, requiring frequent manual intervention and resulting in low detection efficiency. Monitoring equipment often uses a full-array continuous scanning mode, lacking intelligent local scanning control based on vascular motion patterns. This leads to high overall system power consumption and significant energy waste, making it impossible to achieve low-energy operation while maintaining monitoring accuracy, thus failing to meet the requirements for long-term continuous monitoring. Furthermore, the lack of intelligent adaptive control capabilities for different ultrasound detection modalities prevents flexible switching of appropriate detection modes based on the actual motion state of deep blood vessels, hindering the full utilization of the technical advantages of each ultrasound modality. This not only increases the risk of monitoring errors but also significantly reduces the system's ability to achieve stable and continuous dynamic monitoring of deep blood vessels during complex physiological cycles such as cardiac and respiratory cycles. Summary of the Invention

[0004] The purpose of this invention is to provide a wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels, in order to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels, the system comprising an ultrasound sensing module, a central processing module, a data transmission module, and a data analysis module;

[0006] The ultrasonic sensing module is used to emit wide-beam pulse signals to scan the monitoring area and obtain an initial image;

[0007] Furthermore, the process of scanning the monitoring area by transmitting a wide-beam pulse signal to obtain an initial image includes:

[0008] The ultrasonic sensing module generates and transmits wide-beam pulse signals to the monitoring area according to the initial monitoring mode, and collects the echo simulation signals formed in the monitoring area.

[0009] The monitoring area refers to the area on the human body surface that the wearable device is attached to, and the tissue space beneath it that needs to be explored. Its core objective is to cover the deep blood vessels and surrounding tissue structures of the target. The monitoring area is a relatively broad anatomical range because the system does not yet know the exact location of the target blood vessel. Therefore, the monitoring area represents the physical space that the probe can cover and requires a preliminary panoramic scan, aiming to provide complete initial image data for subsequent intelligent identification and precise positioning.

[0010] Wide-beam pulse signal is an ultrasonic beam emitted by the ultrasonic sensing module during the initial monitoring phase. The frequency is usually in the megahertz range and it propagates in the medium in the form of pulses. Its "wide-beam" characteristic means that the ultrasonic beam has a large spatial coverage angle and range, which can explore a large area of ​​the monitoring area in a single or a few transmissions and quickly obtain the overall echo information of the tissue in the area. This signal mode sacrifices some axial resolution in exchange for scanning efficiency, and is suitable for rapid and comprehensive preliminary imaging of unknown target areas in the early stage of system operation.

[0011] The initial monitoring modes include B-ultrasound monitoring and color Doppler ultrasound monitoring;

[0012] Ultrasound mode can provide two-dimensional grayscale images of the monitored area, clearly showing the structure, boundaries and relative positions of different tissues, which is the basis for identifying the morphology and location of blood vessels;

[0013] Color Doppler ultrasound mode overlays color-coded information on blood flow direction and velocity onto the B-mode structural image, which can intuitively display the blood flow situation in the blood vessel; the combination of the two can obtain dual key information about the structural morphology and blood flow of the target blood vessel in the initial scan at one time.

[0014] The analog echo signal is converted into a digital echo signal via analog-to-digital conversion; the digital echo signal is then reconstructed to obtain an initial image of the monitoring area.

[0015] The echo digital signal is amplified and filtered to reduce noise. The processed echo digital signal is arranged and mapped according to its corresponding spatial position (determined by beamforming technology) and signal intensity. The signal intensity is usually converted into grayscale values ​​(B-mode ultrasound) or into color codes based on Doppler frequency shift information (color Doppler ultrasound). Through a specific image reconstruction algorithm (such as a time-delay overlay algorithm), the points on all scan lines are combined into a complete two-dimensional cross-sectional image, which is the initial image of the monitoring area. The initial image contains both the structural information of the tissue (B-mode ultrasound component) and the blood flow information in the blood vessels (color Doppler ultrasound component).

[0016] The central processing module is used to locate the monitored object in the initial image and analyze the dynamic motion trajectory of the monitored object; and to control the ultrasonic sensing module according to the dynamic motion trajectory to obtain the target image and target data of the monitored area.

[0017] Furthermore, the process of locating the monitoring object in the initial image includes:

[0018] The central processing module invokes a pre-trained artificial intelligence algorithm to extract vascular features and match targets from the pre-processed initial image, automatically identifying and locating deep blood vessels as monitoring targets. The pre-processing includes operations such as noise reduction, contrast enhancement, and standardization to improve image quality.

[0019] The artificial intelligence algorithm refers to a machine learning model based on deep learning convolutional neural networks that is trained and optimized using a large amount of labeled medical ultrasound image data (including information on the structure, location, and morphology of various deep blood vessels).

[0020] Vascular feature extraction and target matching refers to the algorithm automatically analyzing the initial image to find unique visual patterns that can represent blood vessels, such as specific tubular structures, boundary echo characteristics, internal echo intensity distribution, and blood flow color signals in color ultrasound mode; and comparing and verifying these extracted features with the deep blood vessel model in the algorithm to distinguish vascular tissue from other non-target tissues (such as muscle, bone, and fat), and confirming that the discovered feature set conforms to the typical definition of deep blood vessels;

[0021] After completing feature extraction and target matching, it not only determines the presence of target blood vessels in the image, but also marks the range, contour, center line, or key points (such as the inner and outer boundaries of the blood vessel wall) of the target blood vessels in the image coordinate system.

[0022] Furthermore, the process of analyzing the dynamic motion trajectory of the monitored object includes:

[0023] The ultrasound sensing module is controlled to continuously acquire continuous image data of the monitored object within the rhythm cycle. The continuous image data is image information of the position, shape and relative position of the surrounding tissues of the monitored object within the rhythm cycle. The rhythm cycle includes the respiratory cycle and the cardiac cycle.

[0024] The respiratory cycle is the time cycle in which the lungs complete one inhalation and exhalation, which causes regular displacement of the trunk and internal organs (including deep blood vessels); the cardiac cycle is the time cycle in which the heart completes one systole and diastole, which directly causes the pulsatile movement of the blood vessel walls and periodic changes in hemodynamic parameters; these two rhythmic cycles are the most important and fundamental physiological rhythm sources affecting the spatial position and morphology of deep blood vessels.

[0025] Continuous image data refers to a series of B-mode and color Doppler ultrasound images continuously acquired by an ultrasound sensing module at high temporal resolution (high frame rate) within at least one complete rhythm cycle. These images are closely linked in time, forming an image sequence. The location refers to the coordinates of the geometric center or specific landmark of the target blood vessel in the image coordinate system; the morphology refers to the structural features of the target blood vessel, such as the diameter of the vessel lumen and the thickness of the vessel wall; and the relative position of the surrounding tissues refers to the spatial relationship between the target blood vessel and the surrounding relatively fixed anatomical structures (such as bones and specific muscle interfaces).

[0026] Temporal features are extracted from continuous image data. These features include the position, vascular diameter, morphological changes, and displacement of the monitored object at different time points (the change in position of the blood vessel between adjacent time points or within a specific period). A motion mathematical model of the monitored object changing with the human circadian rhythm is constructed based on these temporal features. The motion mathematical model refers to an expression or computational framework that describes the quantitative relationship between the movement of the monitored object (deep blood vessels) and the human circadian rhythm by fitting and abstracting the aforementioned temporal features using mathematical functions or algorithms. For example, the positional shift of the monitored object is a periodic function of the respiratory phase and the cardiac phase.

[0027] The dynamic motion law of the monitored object is obtained based on the motion mathematical model. The dynamic motion law is the positional shift of the monitored object with the rhythm cycle (the range and trend of the spatial position change of blood vessels within the rhythm cycle), morphological expansion and contraction (the amplitude and pattern of the change of blood vessel diameter or area within the rhythm cycle), motion speed and motion amplitude (the instantaneous speed, average speed, and peak amplitude of blood vessels at different points on the motion path).

[0028] Based on the dynamic motion law, the dynamic motion trajectory of the monitored object within the monitoring area is fitted. The dynamic motion trajectory is the spatial path curve of the monitored object moving with the rhythmic cycle within the monitoring area and the set of positions corresponding to each time node. Specifically, it refers to the continuous spatial path depicted by the monitored object within the monitoring area as the rhythmic cycle changes. The dynamic motion trajectory is a (or a set of) spatial curves, and each point on the trajectory is associated with a specific time node and its corresponding precise position coordinate set, which is the direct spatial basis for predicting the future position of blood vessels.

[0029] Furthermore, the process of controlling the ultrasonic sensing module based on the dynamic motion trajectory includes:

[0030] The central processing module generates control commands for the ultrasonic sensing module based on the dynamic motion trajectory of the monitored object. The control commands include monitoring mode switching commands and ultrasonic array element activation commands.

[0031] The monitoring mode switching command is to switch the initial monitoring mode to a secondary monitoring mode, which includes A-mode ultrasound monitoring and Doppler monitoring.

[0032] The purpose of the initial monitoring mode (B-mode ultrasound / color Doppler ultrasound) is imaging, identification and localization. B-mode ultrasound provides two-dimensional anatomical images, and color Doppler ultrasound overlays blood flow color coding. The combination of the two provides rich spatial and blood flow presence information for locating targets in unknown areas. Its output is a two-dimensional image containing rich spatial relationships.

[0033] The purpose of the secondary monitoring mode (A-mode ultrasound / Doppler) is to accurately measure and continuously track the target point. A-mode ultrasound provides one-dimensional accurate measurement of the depth of the target point, while Doppler provides the velocity-time spectrum of blood flow at that point. Its output is a high-precision quantitative data stream that changes over time (such as diameter value and velocity value).

[0034] The ultrasonic array element activation command activates the ultrasonic array element that matches the dynamic motion trajectory, while the remaining ultrasonic array elements remain in a dormant state.

[0035] The ultrasound array element is a basic functional unit in a wearable intelligent ultrasound system that converts electroacoustic signals based on the piezoelectric effect. As an independently controllable transmission and reception channel, the ultrasound array element is integrated in an array and controlled by a precise beamforming algorithm to jointly realize the generation, transmission, scanning focusing, and echo reception of ultrasound waves. In the system, the ultrasound array element is further configured as a dynamically selectable and activated unit that responds to the instructions of the central processing module to achieve precise, directional, and low-power tracking and monitoring of moving target blood vessels.

[0036] The ultrasound sensing module responds to the control command and the activated ultrasound array element performs directional pulse transmission along the dynamic motion trajectory of the monitored object, simultaneously acquiring the A-mode ultrasound echo signal and Doppler blood flow echo signal of the monitored object.

[0037] Furthermore, the process of obtaining the target image and target data of the monitoring area includes:

[0038] The structural features of the monitored object are obtained based on the A-scan echo signal, including vessel wall thickness, vessel diameter, vessel elasticity, and the relative position of surrounding tissues. The hemodynamic features of the monitored object are obtained based on the Doppler blood flow echo signal, including blood flow velocity, blood flow direction, blood flow volume, blood flow pulsatility index, and blood flow spectrum morphology. The structural feature signals and hemodynamic feature signals are integrated to form the target data of the monitored object.

[0039] Specifically, the time series of structural feature signals and hemodynamic feature signals are time-aligned, and then the structural feature signals and hemodynamic feature signals are spatially aligned; thus achieving spatiotemporal alignment between structural feature signals and hemodynamic feature signals.

[0040] The spatiotemporally aligned structural feature signals and hemodynamic feature signals are standardized and encapsulated according to a predetermined data structure and format; the generated target data is a structured data frame or data packet, which contains the structural features and hemodynamic features under each spatiotemporal label (time stamp, spatial coordinate) in an orderly manner, indexed by time.

[0041] The structural feature signals are reconstructed to generate target images of the monitored object as it changes with its dynamic motion trajectory. These target images are structural cross-sectional images of the monitored object at various time points and dynamic tracking images superimposed on the dynamic motion trajectory.

[0042] The central processing module receives and calls the extracted structural feature signals of the monitored object, and performs coordinate mapping and contour generation processing on the structural feature signals:

[0043] Based on the transmission timing and reception delay of the A-mode echo signal, the spatial depth coordinates of the reflection points of the anterior and posterior walls of the blood vessel relative to the ultrasound sensing module at each time point are accurately calculated. Combined with the lateral position coordinates predicted by the dynamic motion trajectory of the monitored object, each structural feature data point is assigned its three-dimensional spatial coordinates within the monitoring area. All spatial coordinate points of the blood vessel wall at the same time are connected by a curve fitting algorithm (such as spline interpolation) to generate smooth and continuous contour lines of the inner and outer walls of the blood vessel, thereby constructing a two-dimensional structural cross-sectional image of the monitored object at a single time point. This image can clearly show the cross-sectional shape, wall thickness, and lumen diameter of the blood vessel.

[0044] Then, in chronological order, the cross-sectional images of vascular structures or their key contour lines at each time point are extracted and arranged. These serialized morphological information are fused and superimposed with the pre-generated dynamic motion trajectory spatial path curves. Through specific visualization algorithms (such as spatiotemporal fusion rendering), the contours of all time points are drawn along their motion trajectories, and color gradients or transparency changes are used to characterize the evolution over time, forming a composite image that can comprehensively display the morphological changes of the monitored object and its spatial motion path, i.e., a dynamic tracking image.

[0045] The data transmission module is used to transmit the target image and target data to the data analysis module;

[0046] Furthermore, the process by which the data transmission module transmits the target image and target data to the data analysis module includes:

[0047] The data transmission module performs standardized format conversion (converting images and data from different sources and formats into a unified standard data format agreed upon by the system), data compression, and encryption processing on the target image and target data to form a transmission data frame;

[0048] The data transmission module wirelessly transmits data frames to the data analysis module via a wireless communication protocol and establishes a data transmission feedback link. The data transmission feedback link refers to a reverse communication channel established between the data transmission module and the data analysis module at the same time as wirelessly transmitting data frames, which is used to confirm the reception status. This link enables the data transmission module to know whether the data analysis module has successfully received the data.

[0049] The wireless communication protocol includes short-range wireless protocol and long-range wireless protocol;

[0050] Short-range wireless protocols such as Bluetooth (BLE) and Zigbee are characterized by low power consumption and fast pairing, making them suitable for data transmission between devices and nearby terminals (such as mobile phones and tablets).

[0051] Long-range wireless protocols such as Wi-Fi and cellular mobile networks (4G / 5G) have the characteristics of high bandwidth and wide coverage, making them suitable for directly uploading data to cloud servers or remote medical centers.

[0052] The data transmission module receives a reception confirmation signal from the data analysis module through a data transmission feedback link. If the reception confirmation signal is not received, the wireless transmission is repeated until the data transmission module receives the reception confirmation signal.

[0053] The data analysis module is used to obtain analysis results based on the target image and target data;

[0054] Furthermore, the process by which the data analysis module obtains analysis results based on the target image and target data includes:

[0055] The data analysis module extracts the dynamic changes in the structural morphology of the monitored object at various time points from the target image as the image analysis result. The dynamic changes in structural morphology refer to the changes in the structural morphology of the monitored object with the dynamic motion trajectory and rhythmic cycle. Specifically, it refers to the regular information about the changes in the morphological parameters of the monitored object with time and space, extracted from a series of target images through image processing algorithms. This is manifested as the coordinated and periodic changes in the structural morphology of blood vessels (such as contour, diameter, and cross-sectional area) with their own dynamic motion trajectory (spatial path) and rhythmic cycle (respiration and cardiac cycle). For example, blood vessels exhibit regular expansion at specific phases of movement along the trajectory, or their morphology exhibits periodic pulsation with the heartbeat.

[0056] By performing time-series statistical analysis on the structural feature signals, the dynamic range and fluctuation patterns of the structural feature signals can be obtained. For example, the maximum, minimum, average and standard deviation of the blood vessel diameter in a complete respiratory cycle (dynamic range), as well as its trend and periodicity over time (fluctuation pattern).

[0057] Spectral and time series analyses are performed on hemodynamic characteristic signals to obtain the variation characteristics of the hemodynamic characteristic signals and their correlation with the rhythm cycle. For example, the specific values ​​(variation characteristics) of parameters such as peak systolic velocity, end-diastolic velocity, mean velocity, and pulsatility index, as well as the time delay and correlation strength between these parameters and the ECG R wave or respiratory phase (correlation).

[0058] The dynamic range and fluctuation pattern of structural feature signals, the variation characteristics of hemodynamic feature signals and their correlation with the rhythm cycle are used as the target data analysis results.

[0059] A preset deep vascular medical reference threshold range is defined, which refers to the range of normal and abnormal vascular parameters established based on a large number of clinical studies and medical consensus and pre-stored in the database of the data analysis module. These thresholds include the normal range of vascular diameter, normal spectrum of blood flow velocity, critical value of pulsatility index, etc., for different age groups, genders, and body positions, and are objective benchmarks for judging whether the state of the monitored object is abnormal.

[0060] The image analysis results and target data analysis results are compared item by item with the deep vascular medical reference threshold range; if both the image analysis results and target data analysis results are within the deep vascular medical reference threshold range, no warning is issued; otherwise, a warning is issued.

[0061] The image analysis results (such as whether the morphological changes exceed the normal movement pattern) and the target data analysis results (such as whether the diameter is within the normal range, whether the blood flow velocity is below the critical value, and whether the pulsatility index is abnormally increased) are compared with the threshold item by item; when any one or more analysis results continuously and significantly exceed the reference threshold range, the system determines that there is a potential abnormality in the vascular structure and blood flow status of the monitored object, and generates a corresponding abnormality detection report or warning signal.

[0062] The output of the central processing module is connected to the input of the data transmission module;

[0063] The output of the data transmission module is connected to the input of the data analysis module;

[0064] The ultrasonic sensing module and the central processing module have bidirectional data communication.

[0065] Compared with the prior art, the beneficial effects of the present invention are:

[0066] 1. This invention utilizes a pre-trained artificial intelligence algorithm to automatically identify and accurately locate deep blood vessels in the initial scan image, and analyzes their movement patterns with respiration and cardiac cycles. Based on these patterns, the system can predict the future location of the blood vessels and control the ultrasound beam for directional tracking scanning. This effectively overcomes the problems of poor image quality and inaccurate positioning caused by tissue attenuation and scattering, significantly improving the accuracy and efficiency of continuous, stable, and automated monitoring of moving targets under complex physiological conditions.

[0067] 2. This invention analyzes the dynamic movement trajectory of blood vessels and intelligently switches the initial wide-beam B-mode and color Doppler ultrasound scanning to targeted A-mode and Doppler monitoring, activating only a small number of array elements that correspond to the movement path. This precise tracking and power consumption control strategy based on the target movement pattern abandons the high-energy-consuming method of traditional full-array continuous scanning. While ensuring high-frequency and continuous acquisition of blood vessel morphology and blood flow signals, it significantly reduces overall energy consumption, providing a crucial energy-saving foundation for wearable devices to achieve long-term dynamic monitoring.

[0068] 3. This invention leverages the advantages of multiple monitoring modes, including B-mode ultrasound, color Doppler ultrasound, A-mode ultrasound, and Doppler ultrasound, through a phased and coordinated approach. The system first performs rapid imaging and identification in unknown areas, then switches to high-precision quantitative measurement after target locking, and performs spatiotemporal alignment and fusion analysis on the acquired structural and hemodynamic signals. Finally, the extracted morphological change patterns and blood flow parameters are intelligently compared with preset medical reference thresholds, achieving multi-dimensional and quantitative anomaly detection of vascular conditions, significantly improving the clinical information value and reliability of dynamic monitoring. Attached Figure Description

[0069] Figure 1 This is a schematic diagram of a wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels according to the present invention. Detailed Implementation

[0070] 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.

[0071] Example: Figure 1 As shown, the present invention provides a wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels. The system includes an ultrasound sensing module, a central processing module, a data transmission module, and a data analysis module.

[0072] This embodiment takes the scenario of dynamic monitoring of deep blood vessels in the femoral vein of the lower extremity as an example:

[0073] The ultrasonic sensing module is used to emit wide-beam pulse signals to scan the monitoring area and obtain an initial image;

[0074] In this embodiment, the structure of the ultrasonic sensing module includes a probe, array element parameters, and initial scan configuration:

[0075] The probe is a flexible, bendable ultrasound array probe with a medical-grade silicone base, measuring 12cm×5cm×0.8cm. It conforms to the inner thigh curve and is fixed with Velcro straps. The wearing pressure is ≤3kPa.

[0076] The array parameters include 32 independent ultrasound elements, a center frequency of 10MHz, an element spacing of 0.3mm, and precise calibration to ensure beam consistency; it supports four working modes: B-mode ultrasound, color Doppler ultrasound, A-mode ultrasound, and Doppler ultrasound, with a maximum transmit power of 50mW per element and a receive sensitivity of ≤-100dBm.

[0077] The initial scan configuration was as follows: wide beam pulse signal parameters were beam angle 60°, pulse width 80ns, and repetition frequency 5kHz, ensuring that a single scan covered an 8cm×5cm monitoring area and quickly acquired two-dimensional structural and blood flow distribution images of the femoral vein and surrounding tissues.

[0078] After the patient returned to the ward after surgery, the medical staff placed the flexible probe on the femoral vein projection area on the inner thigh and adjusted the tightness of the strap to ensure seamless contact between the probe and the skin without any air bubbles. After the device was started, the system automatically completed self-tests (array element calibration and signal path detection).

[0079] The process of scanning the monitoring area by transmitting a wide-beam pulse signal to obtain an initial image includes:

[0080] The ultrasonic sensing module generates and transmits wide-beam pulse signals to the monitoring area according to the initial monitoring mode, and collects the echo simulation signals formed in the monitoring area; the initial monitoring mode includes B-mode ultrasound monitoring and color Doppler ultrasound monitoring.

[0081] The analog echo signal is converted into a digital echo signal through analog-to-digital conversion; the digital echo signal is then reconstructed to obtain an initial image of the monitoring area.

[0082] In this embodiment, the ultrasound sensing module emits a wide-beam pulse signal in the initial monitoring mode, scans continuously for 10 seconds, and acquires 20 frames of two-dimensional structural images and blood flow distribution images. The echo analog signal is converted into a 16-bit digital signal by an ADC.

[0083] The central processing module is used to locate the monitored object in the initial image and analyze the dynamic motion trajectory of the monitored object; and to control the ultrasonic sensing module according to the dynamic motion trajectory to obtain the target image and target data of the monitored area.

[0084] In this embodiment, the central processing module includes a core chip, functional units, a control and scheduling unit, and a storage unit.

[0085] The core chip uses the STM32H743ZI2 embedded processor with a main frequency of 480MHz, 1MB SRAM and 16MB Flash, and supports hardware floating-point operations to meet the requirements of real-time signal processing.

[0086] The functional unit integrates a 16-bit ADC converter (sampling rate 100MHz) and an 8th-order low-pass filter (cutoff frequency 20MHz) to achieve noise reduction, amplification and standardization of echo signals.

[0087] It also features a deep learning model that integrates ResNet50 and U-Net, with pre-training data including a large number of clinical lower extremity ultrasound images (including normal blood vessels, pre-thrombotic, and thrombotic scenarios), achieving a blood vessel recognition accuracy of ≥95%.

[0088] The control and scheduling unit has a built-in rhythm cycle synchronization module, which accurately captures the cardiac cycle (0.8-1.2s) and respiratory cycle (3-5s) by linking with ECG and respiratory sensing signals.

[0089] The storage unit is equipped with 8GB of high-speed solid-state storage, which stores raw data and analysis results by timestamp and supports local caching of 7 days of continuous monitoring data.

[0090] The process of locating the monitoring object in the initial image includes:

[0091] The central processing module calls a pre-trained artificial intelligence algorithm to extract vascular features and match targets from the pre-processed initial image, and automatically identifies and locates deep blood vessels as monitoring objects from the pre-processed initial image.

[0092] In this embodiment, the central processing module preprocesses the digital signal, calls the AI ​​algorithm module to extract vascular features, automatically identifies and locates the central axis of the femoral vein, and marks the vascular contour in the image coordinate system.

[0093] The process of analyzing the dynamic motion trajectory of the monitored object includes:

[0094] The ultrasound sensing module is controlled to continuously acquire continuous image data of the monitored object within the rhythm cycle. The continuous image data is image information of the position, shape and relative position of the surrounding tissues of the monitored object within the rhythm cycle. The rhythm cycle includes the respiratory cycle and the cardiac cycle.

[0095] Temporal features are extracted from continuous image data, including the position, blood vessel diameter, morphological changes, and displacement of the monitored object at different time points; and a motion mathematical model of the monitored object changing with the human body's circadian rhythm is constructed based on the temporal features.

[0096] The dynamic motion law of the monitored object is obtained based on the motion mathematical model. The dynamic motion law is the positional shift, morphological expansion and contraction, motion speed and motion amplitude of the monitored object with the rhythm cycle.

[0097] Based on the dynamic motion law, the dynamic motion trajectory of the monitored object within the monitoring area is fitted. The dynamic motion trajectory is the spatial path curve of the monitored object moving with the rhythmic cycle within the monitoring area and the set of positions corresponding to each time node.

[0098] In this embodiment, the central processing module controls the ultrasound sensing module to continuously acquire continuous image data for 3 complete cardiac cycles (approximately 3 seconds) and 3 complete respiratory cycles (approximately 15 seconds), and extracts temporal features such as the coordinates of the vessel center, the diameter of the vessel lumen, and the displacement.

[0099] Based on temporal features, a motion mathematical model is constructed (position offset = A × sin(2πf1t) + B × sin(2πf2t), where f1 is the heart rate and f2 is the respiratory rate), and the spatial path curve of the femoral vein moving with the rhythmic cycle is fitted to determine the precise set of position coordinates for each time node.

[0100] The process of controlling the ultrasonic sensing module according to the dynamic motion trajectory includes:

[0101] The central processing module generates control commands for the ultrasonic sensing module based on the dynamic motion trajectory of the monitored object. The control commands include monitoring mode switching commands and ultrasonic array element activation commands.

[0102] The monitoring mode switching command is to switch the initial monitoring mode to the secondary monitoring mode, which includes A-mode ultrasound monitoring and Doppler monitoring; the ultrasound array element activation command is to activate the ultrasound array element that matches the dynamic motion trajectory, while the other ultrasound array elements remain in a dormant state.

[0103] The ultrasound sensing module responds to the control command and the activated ultrasound array element performs directional pulse transmission along the dynamic motion trajectory of the monitored object, simultaneously acquiring the A-mode ultrasound echo signal and Doppler blood flow echo signal of the monitored object.

[0104] In this embodiment, the central processing module generates control commands to switch the initial monitoring mode to the secondary monitoring mode, and activates 8 core ultrasonic array elements that match the motion trajectory (the remaining 24 array elements remain dormant). The activated area is dynamically updated with the trajectory (update frequency 100Hz).

[0105] The ultrasound sensing module responds to the command by activating the array element to emit a directional pulse along the predicted trajectory (A-mode pulse width 20ns, Doppler sampling rate 2MHz), and simultaneously acquires A-mode echo signals and Doppler blood flow echo signals.

[0106] The process of obtaining the target image and target data of the monitored area includes:

[0107] The structural feature signals of the monitored object are obtained based on the A-scan echo signal, including vascular wall thickness, vascular lumen diameter, vascular elasticity, and the relative position of surrounding tissues; the hemodynamic feature signals of the monitored object are obtained based on the Doppler blood flow echo signal, including blood flow velocity, blood flow direction, blood flow volume, blood flow pulsation index, and blood flow spectrum morphology; the structural feature signals and the hemodynamic feature signals are integrated to form the target data of the monitored object;

[0108] The structural feature signals are reconstructed to generate target images of the monitored object as it changes with the dynamic motion trajectory. The target images are structural cross-sectional images of the monitored object at each time point and dynamic tracking images superimposed on the dynamic motion trajectory.

[0109] In this embodiment, the central processing module extracts structural feature signals such as vessel wall thickness (accuracy 0.1 mm), vessel lumen diameter (dynamic resolution 0.05 mm), vessel elasticity, and relative position of surrounding tissues from the A-mode ultrasound echo signal; and extracts hemodynamic feature signals such as blood flow velocity (accuracy 0.01 m / s), blood flow direction, blood flow rate, blood flow pulsation index, and blood flow spectrum morphology from the Doppler blood flow echo signal.

[0110] The time series of structural feature signals and hemodynamic feature signals are calibrated by timestamp (time error ≤ 1ms), and then the spatial position matching of the two types of signals is completed based on the spatial coordinates of the vascular dynamic motion trajectory (spatial calibration error ≤ 0.2mm).

[0111] The spatiotemporally aligned structural feature signals and hemodynamic feature signals are standardized and encapsulated, and integrated into structured target data indexed by time. The target data contains all vascular structure and hemodynamic feature parameters under each spatiotemporal label (time stamp and spatial coordinates).

[0112] Each structural feature data point is assigned three-dimensional spatial coordinates, and a single-time-node vascular structure cross-sectional image is generated using a spline interpolation algorithm. The serialized cross-sectional image is then fused with the dynamic motion trajectory to generate a vascular dynamic tracking image, i.e., the target image.

[0113] The data transmission module is used to transmit the target image and target data to the data analysis module;

[0114] In this embodiment, the data transmission module adopts a dual-mode communication scheme of BLE 5.2 and Wi-Fi 802.11n, and automatically switches transmission strategies:

[0115] Low power mode (daily monitoring): BLE5.2 transmission, rate 2Mbps, transmission interval 5s, single transmission data amount ≤512KB, power consumption ≤10mA;

[0116] High-speed mode (abnormal warning): Automatically switches to Wi-Fi at a speed of 150Mbps, transmitting images and complete data frames in real time, with power consumption ≤50mA;

[0117] The process by which the data transmission module transmits the target image and target data to the data analysis module includes:

[0118] The data transmission module performs standardized format conversion, data compression, and encryption on the target image and target data to form a transmission data frame.

[0119] The data transmission module wirelessly transmits data frames to the data analysis module via a wireless communication protocol and establishes a data transmission feedback link; the wireless communication protocol includes short-range wireless protocols and long-range wireless protocols.

[0120] The data transmission module receives a reception confirmation signal from the data analysis module through a data transmission feedback link. If the reception confirmation signal is not received, the wireless transmission is repeated until the data transmission module receives the reception confirmation signal.

[0121] In this embodiment, the data transmission module performs format conversion (DICOM 3.0), data compression, and AES-256 encryption on the target image and target data to form a transmission data frame. The system automatically switches the communication mode according to the amount of data transmitted. During daily monitoring, core target data is transmitted every 5 seconds via BLE 5.2. If abnormal parameters are detected, the system immediately switches to Wi-Fi to transmit the complete target data and target image in real time.

[0122] The data analysis module is used to obtain analysis results based on the target image and target data;

[0123] In this embodiment, the terminal configuration of the data analysis module includes a client and a cloud server; the client provides real-time image display, data trend curves, and anomaly warning pop-ups; the cloud server supports data backup, historical comparison, and multi-terminal sharing;

[0124] The data analysis module is used to obtain analysis results based on the target image and target data, including the following process:

[0125] The data analysis module extracts the dynamic changes in the structural morphology of the monitored object at each time node from the target image as the image analysis result. The dynamic changes in the structural morphology are the changes in the structural morphology of the monitored object with the dynamic motion trajectory and rhythmic cycle.

[0126] Time-series statistical analysis was performed on the structural feature signals to obtain their dynamic range and fluctuation patterns. Spectral and time-series analysis was performed on the hemodynamic feature signals to obtain their variation characteristics and correlation with the rhythm period. The dynamic range and fluctuation patterns of the structural feature signals, the variation characteristics of the hemodynamic feature signals, and their correlation with the rhythm period were used as the results of the target data analysis.

[0127] A preset reference threshold range for deep vascular medicine is established. The image analysis results and target data analysis results are compared with the reference threshold range item by item. When both the image analysis results and the target data analysis results are within the reference threshold range, no warning is issued; otherwise, a warning is issued.

[0128] In this embodiment, after receiving the data, the data analysis module performs time-series statistical analysis and spectrum analysis on the target data, and plots the blood vessel diameter fluctuation curve, blood flow velocity spectrum, etc.; at the same time, it extracts the dynamic change features of blood vessel structure morphology in the target image as the image analysis result.

[0129] The image analysis results and target data analysis results are compared item by item with the preset medical reference threshold for the lower extremity femoral vein. An early warning is triggered when any of the following conditions are met:

[0130] The rate of change in blood vessel diameter over three consecutive cycles is ≥20%;

[0131] Blood flow velocity ≤ 0.05 m / s for three consecutive cycles;

[0132] The pulsatility index is >1.5 or <0.3 for two consecutive cycles;

[0133] Vascular elastic compliance <0.01cm² / kPa.

[0134] The system sends pop-up notifications and sound alarms to client devices (such as mobile apps), and simultaneously sends SMS notifications to medical staff's terminals, along with abnormal parameter data and real-time target images.

[0135] The output of the central processing module is connected to the input of the data transmission module;

[0136] The output of the data transmission module is connected to the input of the data analysis module;

[0137] The ultrasonic sensing module and the central processing module have bidirectional data communication.

[0138] 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 exemplary and non-limiting in all respects, 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 wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels, characterized in that: The system includes an ultrasonic sensing module, a central processing module, a data transmission module, and a data analysis module; The ultrasonic sensing module is used to emit wide-beam pulse signals to scan the monitoring area and obtain an initial image; The central processing module is used to locate the monitored object in the initial image and analyze the dynamic motion trajectory of the monitored object; and to control the ultrasonic sensing module according to the dynamic motion trajectory to obtain the target image and target data of the monitored area. The data transmission module is used to transmit the target image and target data to the data analysis module; The data analysis module is used to obtain analysis results based on the target image and target data; The output of the central processing module is connected to the input of the data transmission module; The output of the data transmission module is connected to the input of the data analysis module; The ultrasonic sensing module and the central processing module have bidirectional data communication.

2. The wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels according to claim 1, characterized in that, The process of scanning the monitoring area by transmitting a wide-beam pulse signal to obtain an initial image includes: The ultrasonic sensing module generates and transmits wide-beam pulse signals to the monitoring area according to the initial monitoring mode, and collects the echo simulation signals formed in the monitoring area; the initial monitoring mode includes B-mode ultrasound monitoring and color Doppler ultrasound monitoring. The analog echo signal is converted into a digital echo signal through analog-to-digital conversion; the digital echo signal is then reconstructed to obtain an initial image of the monitoring area.

3. A wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels according to claim 2, characterized in that, The process of controlling the ultrasonic sensing module according to the dynamic motion trajectory includes: The central processing module generates control commands for the ultrasonic sensing module based on the dynamic motion trajectory of the monitored object. The control commands include monitoring mode switching commands and ultrasonic array element activation commands. The monitoring mode switching command is to switch the initial monitoring mode to the secondary monitoring mode, which includes A-mode ultrasound monitoring and Doppler monitoring; the ultrasound array element activation command is to activate the ultrasound array element that matches the dynamic motion trajectory, while the other ultrasound array elements remain in a dormant state. The ultrasound sensing module responds to the control command and the activated ultrasound array element performs directional pulse transmission along the dynamic motion trajectory of the monitored object, simultaneously acquiring the A-mode ultrasound echo signal and Doppler blood flow echo signal of the monitored object.

4. A wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels according to claim 3, characterized in that, The process of obtaining the target image and target data of the monitored area includes: The structural feature signals of the monitored object are obtained based on the A-scan echo signal, including vessel wall thickness, vessel lumen diameter, vessel elasticity, and the relative position of surrounding tissues; the hemodynamic feature signals of the monitored object are obtained based on the Doppler blood flow echo signal, including blood flow velocity, blood flow direction, blood flow volume, blood flow pulsation index, and blood flow spectrum morphology; the structural feature signals and the hemodynamic feature signals are integrated to form the target data of the monitored object; The structural feature signals are reconstructed to generate target images of the monitored object as it changes with the dynamic motion trajectory. The target images are structural cross-sectional images of the monitored object at each time point and dynamic tracking images superimposed on the dynamic motion trajectory.

5. A wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels according to claim 1, characterized in that, The process of locating the monitoring object in the initial image includes: The central processing module calls a pre-trained artificial intelligence algorithm to extract vascular features and match targets from the pre-processed initial image, and automatically identifies and locates deep blood vessels as monitoring objects from the pre-processed initial image.

6. A wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels according to claim 1, characterized in that, The process of analyzing the dynamic motion trajectory of the monitored object includes: The ultrasound sensing module is controlled to continuously acquire continuous image data of the monitored object within the rhythm cycle. The continuous image data is image information of the position, shape and relative position of the surrounding tissues of the monitored object within the rhythm cycle. The rhythm cycle includes the respiratory cycle and the cardiac cycle. Temporal features are extracted from continuous image data, including the position, blood vessel diameter, morphological changes, and displacement of the monitored object at different time points; and a motion mathematical model of the monitored object changing with the human body's circadian rhythm is constructed based on the temporal features. The dynamic motion law of the monitored object is obtained based on the motion mathematical model. The dynamic motion law is the positional shift, morphological expansion and contraction, motion speed and motion amplitude of the monitored object with the rhythm cycle. Based on the dynamic motion law, the dynamic motion trajectory of the monitored object within the monitoring area is fitted. The dynamic motion trajectory is the spatial path curve of the monitored object moving with the rhythmic cycle within the monitoring area and the set of positions corresponding to each time node.

7. A wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels according to claim 1, characterized in that, The process by which the data transmission module transmits the target image and target data to the data analysis module includes: The data transmission module performs standardized format conversion, data compression, and encryption on the target image and target data to form a transmission data frame. The data transmission module wirelessly transmits data frames to the data analysis module via a wireless communication protocol and establishes a data transmission feedback link; the wireless communication protocol includes short-range wireless protocols and long-range wireless protocols. The data transmission module receives a reception confirmation signal from the data analysis module through a data transmission feedback link. If the reception confirmation signal is not received, the wireless transmission is repeated until the data transmission module receives the reception confirmation signal.

8. A wearable intelligent ultrasound system for dynamic monitoring of deep blood vessels according to claim 1, characterized in that, The data analysis module is used to obtain analysis results based on the target image and target data, including the following process: The data analysis module extracts the dynamic changes in the structural morphology of the monitored object at each time node from the target image as the image analysis result. The dynamic changes in the structural morphology are the changes in the structural morphology of the monitored object with the dynamic motion trajectory and rhythmic cycle. Time-series statistical analysis was performed on the structural feature signals to obtain their dynamic range and fluctuation patterns. Spectral and time-series analysis was performed on the hemodynamic feature signals to obtain their variation characteristics and correlation with the rhythm period. The dynamic range and fluctuation patterns of the structural feature signals, the variation characteristics of the hemodynamic feature signals, and their correlation with the rhythm period were used as the results of the target data analysis. A preset reference threshold range for deep vascular medicine is established. The image analysis results and target data analysis results are compared with the reference threshold range item by item. When both the image analysis results and the target data analysis results are within the reference threshold range, no warning is issued; otherwise, a warning is issued.

Images