Intelligent blood vessel puncture point hemostasis device and method based on AI image navigation and multi-mode feedback

Abstract

The invention discloses an intelligent blood vessel puncture point hemostasis device and method based on AI image navigation and multi-mode feedback, and belongs to the technical field of medical instruments. The device comprises a main control unit, an AI image positioning module, a movable compression executing mechanism, a multi-mode sensing module, a pressure control module and a biological activity compression assembly. A DSA image is processed through AI deep learning in the operation, and high-precision positioning of a puncture point and rapid repositioning after body position change are achieved; after an operation, the compression executing mechanism is automatically aligned with a puncture point and applies initial pressure, and the multi-mode sensing module collects pressure, blood flow, temperature and bleeding signals in real time and feeds back the signals to the main control unit; the main control unit is based on a closed-loop feedback control algorithm, aims to effectively stop bleeding, maintain far-end blood flow and reduce tissue damage, and dynamically adjusts compression pressure through the pressure control module. The problems that a traditional device is insufficient in positioning precision and single in pressure adjustment are solved, and the hemostasis safety and the intelligent level are improved.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, specifically to an intelligent pressure device and its control method for hemostasis at the puncture site after percutaneous vascular interventional procedures (such as coronary angiography, coronary stent implantation, lower extremity arterial intervention, and PICC placement), which is particularly suitable for clinical scenarios requiring high-precision positioning, dynamic pressure adjustment, and active wound care. Background Technology

[0002] After interventional procedures via radial artery, femoral artery, or peripheral vascular puncture, effective compression of the puncture site is necessary to prevent complications such as bleeding, hematoma, pseudoaneurysm, vascular occlusion, thrombosis, and infection at the puncture site. The first 24 hours post-procedure are the peak period for complications, making precise hemostasis and dynamic monitoring crucial. Currently, manual compression, mechanical compression devices (such as TR Band™), vascular closure devices, or simple pressure dressings are commonly used in clinical practice, but the following prominent issues remain: (1) Inaccurate compression position: Traditional methods rely on the visual and tactile positioning of medical staff, which can easily cause the compression point to deviate from the actual puncture point due to changes in the patient's position during the operation, postoperative limb movement, or initial positioning deviation; existing image positioning devices often have problems such as insufficient recognition accuracy and poor adaptability, and cannot be quickly repositioned after changes in position, and the radiation exposure and contrast agent usage are too high, increasing the burden on doctors and patients.

[0003] (2) The pressure is static and fixed and the adjustment is singular: the pressure preset depends entirely on the clinical experience of medical staff and cannot be dynamically adjusted according to the patient's bleeding status, blood pressure fluctuations, limb activity or tissue tolerance; insufficient pressure can easily lead to delayed bleeding, while excessive pressure can cause pain, skin damage, vascular occlusion or thrombosis. Moreover, existing devices mostly use a single pressure application method, which cannot be adapted to different puncture sites and individual patient differences.

[0004] (3) Limited monitoring dimensions and delayed early warning: The hemostasis process relies on intermittent manual observation. Existing monitoring devices can only monitor a single physiological parameter (such as pressure or blood flow), lacking real-time monitoring of multiple parameters such as bleeding signals, tissue temperature, and pressure distribution uniformity. Potential complications such as hematoma and infection cannot be detected in time, and risk warning is delayed.

[0005] (4) Lack of proactive wound care: Existing hemostatic devices can only achieve mechanical compression hemostasis, which cannot promote the healing of the puncture site wound. Some devices are made of highly sensitizing and poorly breathable materials, which can easily lead to skin irritation or infection. Furthermore, the postoperative immobilization time is relatively long, which affects the patient's recovery experience.

[0006] (5) Insufficient operational efficiency and intelligence: Some pressurization devices require multiple medical staff to operate together, and cannot achieve real-time data synchronization and remote monitoring. The workload of medical staff is large, which is not conducive to rapid hemostasis in emergency situations.

[0007] Furthermore, while existing biological hemostatic dressings possess rapid hemostasis and promote healing, they lack precise positioning and dynamic pressure adjustment capabilities, making them unsuitable for the precise compression requirements following vascular interventional procedures. Simple pressure devices, though inexpensive, lack positioning accuracy and monitoring capabilities, resulting in a higher risk of complications. Therefore, there is an urgent need for a highly intelligent hemostatic device that achieves precise AI navigation and positioning, multimodal real-time monitoring, adaptive dynamic pressure adjustment, and active wound care, while also considering clinical safety, practicality, and cost-effectiveness. Summary of the Invention

[0008] Technical problems to be solved

[0009] Overcoming the shortcomings of existing technologies such as pressure position deviation, single pressure adjustment, limited monitoring dimensions, delayed early warning, lack of proactive wound care capabilities, and insufficient intelligence, this invention provides an intelligent vascular puncture point hemostasis device and method based on AI image navigation, multimodal physiological feedback, and bioactive wound care. It achieves the integration of precise positioning, dynamic pressure adjustment, multi-dimensional monitoring, proactive wound care, and intelligent early warning, adapting to the clinical needs of multiple scenarios, improving hemostasis effect and patient recovery experience, and reducing the burden on medical staff.

[0010] Technical solution

[0011] A smart vascular puncture point hemostasis device based on AI image navigation and multimodal feedback includes:

[0012] Main control unit: includes an embedded microprocessor, a memory and an alarm unit. The memory stores an improved feedback control algorithm, an AI image recognition model and patient physiological parameter thresholds. The alarm unit includes an audible and visual alarm and a wireless early warning module, which can trigger multi-level early warnings when parameters are abnormal. The main control unit supports multi-module collaborative control and data integration processing.

[0013] AI Image Recognition and Positioning Module: Acquires intraoperative DSA images via DICOM interface, integrates self-developed AngiSight-like AI deep learning navigation software, and adopts an improved U-Net neural network and real-time image fusion algorithm to automatically segment and identify puncture instruments and vascular anatomy, accurately locate the projection position of the puncture point on the body surface, and generate coordinate instructions; supports the completion of new image fusion and repositioning within 3-5 seconds after a change in body position, is compatible with various brands of DSA models, requires no equipment hardware upgrades, and can reduce radiation exposure by 40% and contrast agent usage by 60%.

[0014] The movable compression actuator includes a high-precision stepper motor slide that can move in the XYZ three-dimensional plane, a compression head mounted on the slide, and a device fixing assembly. The compression head can be adjusted 360° to adapt to different puncture sites such as the radial artery and femoral artery. The compression head is equipped with a pressure application unit, which can be selected as an inflatable and deflated silicone airbag, a linear motor pressure rod, or a deformable memory alloy component, supporting switching between uniform pressure and pulsed pressure modes. The device fixing assembly adopts a structure combining low-allergenic breathable adhesive tape and adjustable Velcro, which is compact and does not affect the patient's clothing or body position, while maximizing the protection of the sterile environment of the puncture area.

[0015] Multimodal physiological sensing module: integrated into the area near the compression head or puncture point, including a thin-film pressure sensor array (monitoring tissue pressure and pressure distribution uniformity), a miniature ultrasonic Doppler probe (monitoring blood flow velocity and blood flow status at the distal end of the puncture point), a bleeding sensor (monitoring active bleeding signals at the puncture point), and a temperature sensor (monitoring the temperature of the tissue around the puncture point, providing early warning of infection and hematoma); the sensors use wireless transmission to avoid cable entanglement affecting patient movement.

[0016] Pressure control module: Responding to the main control unit's instructions, it drives the movable pressure actuator to move and applies controllable pressure output from the unit. It supports custom initial pressure settings, and the pressure adjustment accuracy can reach 1 mmHg. It can realize step-by-step pressure increase, slow pressure reduction and pulse pressure adjustment to adapt to the tissue tolerance and hemostasis needs of different patients.

[0017] Bioactive Compression Component: Detachably installed at the bottom of the compression head, it comes into direct contact with the skin at the puncture site. It uses thrombin-anchored bacterial cellulose (T-BC) dressing as its core material, combining rapid hemostasis, gentle wound care, and active healing promotion. It can initiate rapid hemostasis within 1 minute, while accelerating the healing of the puncture site wound, increasing the healing speed by approximately 40%. The component has good biocompatibility, breathability, and toughness, is non-allergenic, avoids skin damage, and is detachable and replaceable for single use, reducing the risk of infection.

[0018] Wireless communication module: Employing Bluetooth or 5G communication, it synchronizes real-time data collected by the multimodal physiological sensing module to the medical terminal. (Computer and mobile APP) It supports real-time data display, historical data query and abnormal data warning. It can also realize remote parameter adjustment and hemostasis process monitoring, so that medical staff can keep abreast of the patient's hemostasis status and reduce the frequency of manual inspection.

[0019] Improved feedback control algorithm: Stored in the main control unit's memory and executed by the processor, it uses "no active bleeding, distal blood flow not lower than the safety threshold, and tissue temperature around the puncture site within the normal range" as triple optimization objectives. Combining individual patient physiological parameters and clinical hemostasis data, it dynamically adjusts the compression pressure, compression method, and compression time to form a closed-loop adaptive control of "sensing-decision-execution-feedback". The algorithm has self-learning capabilities and can automatically optimize the pressure adjustment parameters according to the hemostasis requirements of different puncture sites and different surgical types.

[0020] The present invention also provides an intelligent hemostasis method based on the above-mentioned device, comprising the following steps:

[0021] S1. Intraoperative positioning: During percutaneous vascular intervention, intraoperative DSA images are acquired through the AI ​​image recognition and positioning module. The AI ​​deep learning algorithm automatically segments and identifies the intersection of the puncture needle and the blood vessel, maps it to the device coordinate system through coordinate system transformation, accurately locates the projection position of the puncture point on the body surface, generates coordinate instructions and stores them in the main control unit; if the patient's position changes during the operation, the module automatically completes the fusion of new images and achieves repositioning of the puncture point within 3-5 seconds to ensure positioning accuracy.

[0022] S2. Device Fixation and Alignment: After the operation, the device is fixed near the patient's puncture site using the fixation component to ensure a secure fixation without compressing blood vessels and nerves; the main control unit calls the coordinate instructions stored in S1 to drive the movable compression actuator to move, automatically aligning the compression head (with the bioactive compression component installed) with the puncture point, adjusting the angle and height of the compression head to ensure that the bioactive compression component fits tightly against the skin at the puncture point, while not affecting the sterile environment of the puncture area.

[0023] S3. Initial Pressure Application and Monitoring Start-up: Medical staff set the initial pressure (default 20 mmHg, customizable) through the medical terminal or device control panel. The pressure control module responds to the command and drives the pressure application unit to apply the initial pressure. At the same time, the multimodal physiological sensing module and wireless communication module are activated to continuously collect multi-dimensional physiological parameters such as pressure, blood flow, bleeding signals, and tissue temperature, and synchronize the data to the medical terminal in real time.

[0024] S4. Closed-loop dynamic pressure regulation: The improved feedback control algorithm analyzes the data collected by the multimodal physiological sensing module in real time and dynamically adjusts the compression pressure and compression mode according to the following logic:

[0025] (1) If an active bleeding signal is detected (triggered by the bleeding sensor), start stepwise pressurization (+1-5 mmHg each time, adjustable) until the bleeding signal disappears, while maintaining distal blood flow at a safe threshold.

[0026] (2) If the distal blood flow velocity is detected to be lower than the safe threshold, start slow decompression (-1-3 mmHg each time, adjustable) until the blood flow velocity is restored to the safe range, while ensuring no active bleeding.

[0027] (3) If the temperature of the tissue around the puncture site is abnormally high (exceeding the normal range), it indicates that there may be a hematoma or infection, triggering an audible and visual alarm and a wireless warning. At the same time, the pressure should be appropriately reduced to avoid further tissue damage, and medical staff should be notified to handle the situation in a timely manner.

[0028] (4) If uneven pressure distribution is detected, adjust the angle of the pressure head and the pressure application method to ensure that the pressure is applied evenly to the puncture point and avoid excessive local pressure that could cause skin damage.

[0029] (5) If all physiological parameters are within the normal range and continue for the preset time (customizable, default 10 minutes), then maintain the current compression pressure and compression method, and enter the stable hemostasis stage.

[0030] S5. Hemostasis Completion and Decompression Indication: During the stable hemostasis phase, if the multimodal physiological sensing module detects no active bleeding at the puncture site, stable distal blood flow, and normal tissue temperature, and the duration reaches the preset hemostasis completion threshold (which can be customized according to the surgical type), the main control unit initiates a slow decompression program to gradually reduce the compression pressure while continuously monitoring physiological parameters to avoid secondary bleeding caused by excessively rapid decompression. After decompression is completed, an audible and visual prompt and a medical terminal warning are triggered to indicate that hemostasis is complete. Medical staff can then remove the device and replace it with a bioactive compression component for subsequent wound care.

[0031] S6. Postoperative follow-up: After hemostasis is completed, the bioactive compression component can be left in place for 1-2 days to continue to play a role in wound care and healing; the wireless communication module can continue to synchronize tissue temperature and healing-related data at the puncture site, which is convenient for medical staff to follow up after surgery and detect potential complications in a timely manner.

[0032] Beneficial effects

[0033] Compared with the prior art, the present invention has the following significant advantages:

[0034] (1) More accurate positioning and stronger adaptability: It integrates AI image navigation technology to achieve puncture point positioning at the level of 0.1 mm. It can be quickly repositioned after the body position changes. It is compatible with DSA models of various brands and does not require hardware upgrades. At the same time, it reduces radiation exposure and contrast agent usage, and improves the safety of doctors and patients. The positioning process is automated, eliminating manual positioning deviations and adapting to various puncture sites such as radial artery and femoral artery.

[0035] (2) More intelligent pressure regulation and wider adaptability: The improved feedback control algorithm is adopted to achieve closed-loop adaptive pressure regulation with three objectives. The pressure regulation accuracy is high and the methods are diverse. The compression pressure and compression method can be dynamically optimized according to the individual differences of patients and the hemostasis status, taking into account the hemostasis effect, vascular patency and tissue tolerance, effectively avoiding bleeding caused by insufficient pressure and vascular damage caused by excessive pressure.

[0036] (3) More comprehensive monitoring and more timely early warning: The multimodal physiological sensing module realizes real-time monitoring of pressure, blood flow, bleeding and temperature in multiple dimensions. Combined with the wireless communication module, the data can be synchronized to the medical terminal to realize multi-level early warning of abnormal situations (audio-visual early warning + wireless early warning), and early warning of complications such as bleeding, hematoma, infection, and vascular occlusion. This solves the problem of delayed early warning in traditional devices and improves the safety of hemostasis.

[0037] (4) More efficient wound care and more comfortable experience: The integrated bioactive compression component has the functions of rapid hemostasis, gentle wound care and active healing promotion, which accelerates wound healing. At the same time, it uses low-allergenic and breathable fixation and contact materials, which are small in size and do not affect the patient's activities and clothing, reduce the risk of skin damage and allergy, and improve the patient's postoperative rehabilitation experience.

[0038] (5) More convenient operation and lighter burden on medical staff: The device is easy to fix and can be operated independently by a single medical staff member. It automatically positions, applies pressure and adjusts pressure, reducing manual intervention. The wireless remote monitoring and early warning function reduces the frequency of manual inspection by medical staff, while reducing radiation exposure and contrast agent usage, improving surgical and nursing efficiency, especially suitable for rapid hemostasis in emergency scenarios.

[0039] (6) More versatile and practical: It is compatible with a variety of percutaneous vascular interventional procedures (coronary angiography, coronary stent implantation, lower extremity arterial intervention, PICC placement, etc.), and the parameters can be customized according to different puncture sites and surgical types. The bioactive compression components are detachable and replaceable, achieving single-use, and taking into account versatility, safety and economy.

[0040] (7) Higher level of intelligence: The algorithm has self-learning ability and can continuously optimize the pressure regulation strategy by combining clinical data. At the same time, it supports real-time data synchronization, historical query and remote control, realizes intelligent management of the entire hemostasis process, and promotes the development of hemostasis nursing after vascular intervention towards precision, intelligence and remoteness. Attached Figure Description

[0041] Figure 1This is a simplified system block diagram of the intelligent vascular puncture point hemostasis device of the present invention, showing the overall closed-loop control architecture of the device. The main control unit, as the core control module, is connected to the AI ​​image recognition and positioning module, the multimodal physiological sensing module, the pressure control module, and the wireless communication module. Physiological parameters collected by the multimodal physiological sensing module are fed back to the main control unit, which then drives the movable compression actuator to perform hemostasis operations through the pressure control module, forming a closed-loop control chain of "perception-decision-execution-feedback".

[0042] Figure 2 is a block diagram of the intelligent hemostasis device system, showing the specific composition of each module. For example, the main control unit includes a processor and a memory storing the core algorithm; the compression actuator consists of a movable platform and a compression head with an airbag; and the physiological sensing module integrates both pressure and blood flow sensors. This block diagram fully embodies the closed-loop control concept of "perception-decision-execution," ensuring the system's intelligence and reliability.

[0043] Figure 3 is a schematic diagram of the compression actuator and compression head structure, showcasing the most crucial actuating components of the invention—the compression actuator and compression head. It consists of a high-precision two-dimensional moving platform and a multifunctional compression head. The left side shows the mechanical structure; the two-dimensional slider platform is driven by a stepper motor, ensuring the precision of the compression head's movement. The right side shows the internal structure of the compression head, which not only includes a silicone airbag for applying pressure but also integrates a pressure sensor array to monitor pressure uniformity and a miniature Doppler probe to monitor blood flow in distal blood vessels in real time. This integrated design enables precise compression and real-time monitoring.

[0044] Figure 4 is a flowchart of the intelligent hemostasis method, illustrating its workflow. The entire process begins with intraoperative DSA image localization, and postoperatively, the device automatically aligns and applies initial pressure. Subsequently, the system enters a continuous closed-loop adjustment phase: dynamically increasing or decreasing pressure based on real-time monitored bleeding and blood flow data. This cycle continues until the preset hemostasis completion conditions are met. Finally, the system will prompt and gradually reduce pressure. The entire process is safe, intelligent, and requires no manual intervention.

Claims

1. A smart vascular puncture point hemostasis device, characterized in that, include: Main control unit; The image recognition and positioning module is used to acquire and process digital subtraction angiography images during surgery, automatically identify the projection position of the vascular puncture point on the body surface, and generate position coordinate instructions; the movable compression actuator, controlled by the main control unit, includes a translation platform that can move in a two-dimensional plane and a compression head mounted thereon, the compression head including a pressure application unit; the physiological sensing module is set at or near the compression head and is used to collect physiological parameters related to the hemostasis status.

2. The main control unit is configured to dynamically control the pressure applied by the pressure application unit to the puncture point based on the real-time data collected by the physiological sensing module.

3. The apparatus according to claim 1, characterized in that, The physiological sensing module includes: a pressure sensor for monitoring the pressure applied to the tissue by the compression head, and / or a blood flow sensor for monitoring the blood flow status of the blood vessels distal to the puncture point.

4. The apparatus according to claim 2, characterized in that, The main control unit is configured to execute a feedback control algorithm, which dynamically adjusts the applied pressure with the control objectives of simultaneously satisfying "effective hemostasis" and "maintaining minimum patency of distal blood vessels".

5. The apparatus according to claim 1, characterized in that, The pressure application unit is an inflatable airbag, a linearly telescopic pressure bar, or a deformable shape memory alloy component.

6. The apparatus according to claim 1, characterized in that, The image recognition and positioning module uses a machine learning-based image recognition algorithm to automatically identify puncture instruments and vascular anatomy in angiographic images.

Images