A humanoid physical intelligent general blood vessel intervention surgery robot system and method

Abstract

The application provides a humanoid embodied intelligent general vascular interventional surgery robot system and a control method thereof. The system is constructed based on a perception-cognition-decision-execution-feedback embodied intelligent closed loop architecture, comprising: an intelligent assessment module for disease assessment and vascular interventional surgery indication determination of a patient; a surgery planning module for reconstructing a three-dimensional vascular model of a target vascular segment and generating a personalized interventional surgery planning scheme; a humanoid vascular interventional execution module comprising a humanoid robot body, a guide wire control module and a contrast guide module, for delivering an interventional instrument to a target vascular segment through a vascular approach and executing a vascular interventional operation based on intraoperative contrast guide; an auxiliary rehabilitation module for generating a personalized postoperative rehabilitation scheme. The system realizes the intelligentization, autonomy and full-process closed loop management of general vascular interventional surgery.

Description

Technical Field

[0001] This application relates to the field of medical device technology, and more specifically, to a humanoid intelligent panvascular interventional surgical robot system and method. Background Technology

[0002] Interventional vascular surgery is a minimally invasive surgical procedure that involves inserting catheters, guidewires, balloons, stents, and other interventional devices into the human vascular system via a peripheral vascular approach to diagnose and treat lesions in the cardiovascular, cerebrovascular, and peripheral vascular systems. It is widely used in clinical fields such as percutaneous coronary intervention (PCI), cerebrovascular intervention, endovascular aortic repair, and peripheral artery intervention.

[0003] However, existing vascular interventional surgery techniques face many pressing technical challenges. Current surgical robot systems fail to integrate preoperative assessment and planning, intraoperative autonomous execution, emergency control, and postoperative rehabilitation management. They lack embodied intelligent autonomous decision-making capabilities based on real-time intravascular sensing information and cannot be designed to replace human surgeons in their physical positions within the radiation area. Therefore, a novel surgical robot system that deeply integrates human-like embodied intelligence technology with pan-vascular interventional surgery is urgently needed. Summary of the Invention

[0004] To address the aforementioned problems, the first aspect of this application provides a humanoid embodied intelligent panvascular interventional surgical robot system. The system is constructed based on an embodied intelligent closed-loop architecture of perception-cognition-decision-execution-feedback, comprising: The intelligent assessment module is used to conduct targeted disease assessments and determine the indications for vascular interventional surgery based on the patient's medical data and vital signs data, generate personalized auxiliary assessment results for the patient, and display them. The surgical planning module is used to reconstruct a three-dimensional vascular model of the target vascular segment based on the patient's angiography image data, perform targeted interventional surgical planning for the patient based on the three-dimensional vascular model, generate a personalized interventional surgical planning scheme, and display it. The human-shaped vascular interventional execution module is used to deliver interventional instruments to the target vascular segment via a vascular approach according to the interventional surgery plan, and to perform vascular interventional operations based on intraoperative angiography guidance; The auxiliary rehabilitation module is used to generate and display personalized postoperative rehabilitation plans for patients based on their medical data and postoperative assessment results.

[0005] The second aspect of this application provides a control method for a humanoid intelligent pan-vascular interventional surgical robot system, applied to the aforementioned humanoid intelligent pan-vascular interventional surgical robot system, comprising: Based on the patient's medical data and physical signs, targeted disease assessments and determination of indications for vascular interventional surgery are conducted, generating personalized auxiliary assessment results for the patient and displaying them. Based on the patient's angiography data, a three-dimensional vascular model of the target vascular segment is reconstructed. Based on the three-dimensional vascular model, a targeted interventional surgery plan is developed for the patient, generating a personalized interventional surgery plan, which is then displayed. According to the interventional surgery plan, the interventional device is delivered to the target vascular segment via a vascular approach, and the vascular interventional operation is performed based on intraoperative angiography guidance; Based on the patient's medical data and postoperative assessment results, a personalized postoperative rehabilitation plan is generated and displayed.

[0006] A third aspect of this application provides an electronic device comprising: a memory and a processor; the memory for storing a program; and the processor, coupled to the memory, for executing the program to implement the aforementioned humanoid intelligent panvascular interventional surgical robot system control method.

[0007] The fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the control method of the humanoid intelligent panvascular interventional surgical robot system described above.

[0008] The humanoid intelligent panvascular interventional surgical robot system provided in this application has the following beneficial effects: By adopting an embodied intelligent closed-loop architecture of perception-cognition-decision-execution-feedback, the panvascular interventional surgical robot is upgraded from a passive execution tool to an intelligent surgical system with autonomous decision-making capabilities. It can autonomously adjust delivery strategies and interventional operation plans according to real-time changes in the intravascular environment, reducing the dependence on the surgeon's personal operation skills.

[0009] By reconstructing three-dimensional vascular models, determining vascular access and interventional pathways, selecting interventional devices, and iteratively optimizing contrast agent usage protocols in interventional surgery planning, personalized and precise interventional surgery design has been achieved. In particular, by iteratively optimizing the contrast agent usage protocol, the amount of contrast agent used has been minimized, reducing the risk of contrast agent-induced nephropathy.

[0010] The standardized mechanical interface design of the humanoid robot body enables it to adapt to different types of surgical execution modules, realizing modular expansion of surgical execution capabilities; the precision drive mechanism of the guidewire control module enables precise delivery of catheters and guidewires; and the real-time fusion of the preoperative three-dimensional vascular model and intraoperative fluoroscopic images of the angiography guidance module enables enhanced three-dimensional navigation guidance.

[0011] The emergency control module, with its control cabin, main control handle, force feedback module, and image display module, enables operators to quickly take over intervention and control when risks escalate outside the radiation protection area, ensuring the safety of the surgery.

[0012] By using a radiation-resistant humanoid torso with guidewire and catheter control arms and equipped with dexterous hands, the robot simulates the operation mode of interventional doctors delivering medication with both hands, allowing the humanoid robot to replace the physical position of human doctors in X-ray irradiation environments.

[0013] The collaborative module enables spatial linkage between the humanoid robot and the intraoperative image acquisition equipment, automatically coordinating the humanoid robot's posture to avoid the image acquisition space and ensuring that the angiography is not obstructed or interfered with by the robot.

[0014] By autonomously performing cognitive understanding and decision-making reasoning on multimodal intravascular perception information through the embodied intelligent decision-making module, delivery strategies and interventional operation strategies are generated and autonomously iterated and corrected based on execution feedback, thus realizing intelligent autonomous decision-making for vascular interventional surgery. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below.

[0016] Figure 1 This is a schematic diagram of the overall architecture of the humanoid intelligent pan-vascular interventional surgical robot system provided in the embodiments of this application; Figure 2 This is a schematic diagram of the humanoid blood vessel intervention execution module provided in an embodiment of this application; Figure 3 A schematic diagram of a humanoid robot with a standardized mechanical interface provided in an embodiment of this application; Figure 4 This is a schematic diagram of the emergency control module architecture provided in an embodiment of this application; Figure 5 A schematic diagram of the collaborative module provided in the embodiments of this application; Figure 6 A schematic diagram of the architecture of the embodied intelligent decision-making module provided in the embodiments of this application; Figure 7 A flowchart illustrating the control method for a humanoid intelligent pan-vascular interventional surgical robot system provided in this application embodiment;

[0017] Figure 8 This is a schematic diagram of an electronic device architecture provided in an embodiment of this application. Detailed Implementation

[0018] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, specific embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0019] Although exemplary embodiments of this application are shown in the accompanying drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0020] Rather, these embodiments are provided so that this application can be understood more thoroughly and its scope can be fully conveyed to those skilled in the art.

[0021] It should be noted that, unless otherwise stated, the technical or scientific terms used in this application shall have the ordinary meaning as understood by one of ordinary skill in the art to which this application pertains.

[0022] like Figure 1 As shown, this application provides a humanoid embodied intelligent pan-vascular interventional surgical robot system. The system is built upon an embodied intelligent closed-loop architecture of perception-cognition-decision-execution-feedback, and includes: The intelligent assessment module is used to conduct targeted disease assessments and determine the indications for vascular interventional surgery based on the patient's medical data and vital signs data, generate personalized auxiliary assessment results for the patient, and display them. The surgical planning module is used to reconstruct a three-dimensional vascular model of the target vascular segment based on the patient's angiography image data, perform targeted interventional surgical planning for the patient based on the three-dimensional vascular model, generate a personalized interventional surgical planning scheme, and display it. The human-shaped vascular interventional execution module is used to deliver interventional instruments to the target vascular segment via a vascular approach according to the interventional surgery plan, and to perform vascular interventional operations based on intraoperative angiography guidance; The auxiliary rehabilitation module is used to generate and display personalized postoperative rehabilitation plans for patients based on their medical data and postoperative assessment results.

[0023] Specifically, the embodied intelligent closed-loop architecture is the core architecture that upgrades the pan-vascular interventional surgical robot from a tool that passively executes the doctor's control instructions to an intelligent agent with the ability to perceive the environment, understand the cognition, make autonomous decisions, execute precisely, and provide feedback on the results.

[0024] The design philosophy of this architecture is based on the theory of embodied intelligence, which states that intelligent agents must acquire and utilize knowledge through direct interaction with the physical environment.

[0025] In the context of panvascular interventional surgery, the physical environment is the complex and tortuous human vascular system. The intelligent agent interacts directly with the vascular environment through catheters and guidewires, sensing angiographic images, guidewire tactile force, and instrument position information within the blood vessels. After cognitive reasoning and decision planning, it controls the guidewire manipulation module to perform corresponding advance, retraction, and rotation operations, and continuously corrects and optimizes the delivery strategy through a feedback mechanism.

[0026] The effect of this closed-loop architecture is that it enables the pan-vascular interventional surgical robot to autonomously navigate, deliver, and respond to emergencies in complex vascular environments.

[0027] Specifically, the intelligent assessment module is the system's preoperative assessment entry point, responsible for comprehensively analyzing the patient's multi-dimensional medical data to determine whether the patient is suitable for vascular interventional surgery.

[0028] The medical data includes, but is not limited to, the patient's medical history, laboratory test results (complete blood count, coagulation function, liver and kidney function, cardiac markers, etc.), imaging data (CTA, MRA, ultrasound, etc.), and previous vascular interventional surgery records.

[0029] The vital signs data include the patient's vital signs, cardiac function classification, renal function classification (eGFR value), and vascular access status assessment results.

[0030] The intelligent assessment module uses a multimodal basic model to fuse and analyze the above-mentioned multi-source data and perform cognitive reasoning to generate personalized auxiliary assessment results for vascular interventional surgery. These assessment results include not only disease diagnosis and severity grading, but also feasibility assessment of access vascular pathways, contrast agent tolerance assessment, and anticoagulation status assessment.

[0031] The personalization of the personalized auxiliary assessment results is essentially generated by the multimodal basic model based on the patient's individual medical history, examination data and anatomical features to generate assessment conclusions autonomously, rather than using a preset general assessment template.

[0032] The effect of this feature is that it provides a scientific and comprehensive preoperative assessment basis for subsequent surgical planning, avoiding intraoperative accidents caused by insufficient assessment.

[0033] Specifically, the surgical planning module is responsible for constructing a three-dimensional vascular model of the target vascular segment based on the patient's angiography image data, and formulating a detailed interventional surgical plan based on this model.

[0034] This module first uses the patient's CTA or MRA image data to construct a complete three-dimensional vascular model of the target vascular segment through a three-dimensional reconstruction algorithm. This model accurately reflects the course of the blood vessel, its bifurcation structure, diameter changes, degree of tortuosity, and lesion distribution.

[0035] Then, on the three-dimensional vascular model, the vascular access and interventional path are determined, the catheter and guidewire passability is assessed, the interventional device is selected, the implantation location and operation parameters are determined, and the contrast agent usage plan is planned.

[0036] The personalization of the personalized interventional surgery planning scheme is essentially generated by a multimodal basic model that autonomously infers interventional operation parameters based on the patient's individual vascular anatomy and lesion characteristics, rather than using a preset general parameter template.

[0037] The effect of this feature is that it provides detailed path guidance and operational instructions for the precise delivery of catheters and guidewires during the procedure, significantly improving the predictability and safety of interventional surgery.

[0038] Specifically, the human-shaped vascular interventional execution module is responsible for delivering interventional devices to the target vascular segment via a vascular access route and performing vascular interventional procedures.

[0039] This module includes a humanoid robot body and a guide wire control module and an angiography guidance module mounted on it.

[0040] The humanoid robot can autonomously move to the surgical position and dock with the operating table. Its end effector is equipped with a standardized mechanical interface to adapt to different types of surgical execution modules.

[0041] The guidewire control module uses a precision drive mechanism to advance, retract, and rotate the catheter and guidewire, accurately delivering the interventional device to the target blood vessel segment.

[0042] The angiography-guided module acquires real-time imaging information of the vascular lumen through X-ray fluoroscopy and digital subtraction angiography, and integrates the preoperative three-dimensional vascular model with the intraoperative fluoroscopic images in real time to assist navigation.

[0043] The effect of this feature is that it enables precise delivery of interventional devices within complex vascular systems and real-time image-guided operation.

[0044] Specifically, the auxiliary rehabilitation module is responsible for generating personalized postoperative rehabilitation plans based on the patient's medical data and postoperative assessment results.

[0045] This module comprehensively considers factors such as interventional surgical methods, vascular lesion treatment outcomes, stent or balloon implantation status, and individual patient differences. It uses a multimodal basic large model for reasoning and analysis to generate personalized rehabilitation plans that include antiplatelet therapy, anticoagulation management, cardiovascular rehabilitation exercise prescriptions, risk factor management, contrast agent nephropathy prevention, and follow-up plans.

[0046] The effect of this feature is that it enables closed-loop management of the entire process from preoperative assessment to postoperative rehabilitation, avoiding information gaps between postoperative rehabilitation and intraoperative treatment.

[0047] The beneficial effects of the aforementioned humanoid intelligent panvascular interventional surgical robot system are as follows: through the embodied intelligent closed-loop architecture of perception-cognition-decision-execution-feedback, the panvascular interventional surgical robot possesses autonomous decision-making and adaptive adjustment capabilities; through the intelligent assessment module, it achieves comprehensive preoperative disease assessment and indication determination; through the surgical planning module, it achieves personalized interventional surgical planning based on a three-dimensional vascular model; through the humanoid vascular interventional execution module, it achieves precise delivery of interventional instruments within the blood vessel and real-time angiography guidance; and through the auxiliary rehabilitation module, it achieves personalized postoperative rehabilitation management.

[0048] In a preferred embodiment, the interventional procedure planning is configured to: reconstruct a three-dimensional vascular model of the target vascular segment based on preoperative imaging data, and mark vascular bifurcation, stenosis and lesion sites; determine the vascular access and interventional path on the three-dimensional vascular model, and assess the passability of the catheter and guidewire; select interventional devices according to the lesion type and vascular morphology, and determine the implantation location and operating parameters; iterate the contrast agent usage plan until the contrast agent dosage is minimized.

[0049] Specifically, the three-dimensional vascular model reconstruction is the basis for interventional surgery planning.

[0050] The process begins by segmenting the patient's CTA or MRA image data into blood vessels, and then using a deep learning blood vessel segmentation network (such as a 3D U-Net architecture) to automatically extract the lumen contour and wall boundary of the target blood vessel segment.

[0051] Then, the segmentation results are converted into a three-dimensional blood vessel model through a three-dimensional surface reconstruction algorithm. This model accurately reflects the diameter distribution, bending angle, bifurcation angle, and lesion morphology of the blood vessels.

[0052] The system automatically marks the bifurcation points of blood vessels (marked with bifurcation nodes), stenosis sites (marked with stenosis percentage), and lesion sites (marked with lesion length and type) on the 3D model, providing an intuitive 3D visualization basis for subsequent path planning and device selection.

[0053] The effect of this feature is that it upgrades traditional two-dimensional image analysis to three-dimensional visualization analysis, making the planning of interventional pathways and device selection more intuitive and accurate.

[0054] Specifically, the determination of the vascular access and interventional path is based on optimal path planning using a three-dimensional vascular model.

[0055] The system first determines the optimal vascular approach (such as via the femoral artery or the radial artery) based on the location of the target vascular segment and the assessment results of the access vessel. Then, it plans the optimal interventional path from the access point to the target lesion segment on a three-dimensional vascular model, evaluates the impact of the diameter, curvature, and bifurcation angle of each vascular segment on the catheter and guidewire passage, and recommends matching catheter and guidewire models and specifications.

[0056] For complex multi-vessel lesions, the system also plans the intervention sequence and path switching strategy for multiple target vessels.

[0057] The effect of this feature is that, through three-dimensional path planning and accessibility assessment, it reduces the risk of intraoperative catheter and guidewire delivery failure or inappropriate instrument selection.

[0058] Specifically, the selection of interventional devices is precisely matched based on the lesion type (stenosis, occlusion, dissection, etc.) and vascular morphology (diameter, curvature, bifurcation angle, etc.).

[0059] For stenotic lesions, the system recommends appropriate stent and balloon models based on lesion length, reference vessel diameter, and degree of stenosis; for bifurcation lesions, the system recommends appropriate bifurcation stent technology and stent models based on bifurcation angle and branch vessel diameter.

[0060] The implantation location is determined by precise positioning on a three-dimensional vascular model, and the operating parameters (such as balloon dilation pressure, stent release pressure, etc.) are set according to the degree of lesion calcification and vascular wall characteristics.

[0061] The effect of this feature is that it enables precise device matching and parameter setting based on the individual vascular characteristics of the patient.

[0062] Specifically, the system comprehensively considers the patient's renal function status (eGFR value), the complexity of the planned interventional procedure, and the expected number of angiography sessions, and formulates a contrast agent usage plan through an iterative optimization algorithm.

[0063] The specific iterative process is as follows: First, an initial contrast agent usage plan is generated, and the total contrast agent usage under this plan is calculated; then, the necessity of each contrast acquisition is evaluated, and contrast acquisitions that can be replaced by 3D fusion navigation are eliminated; then, the injection parameters for each contrast acquisition are optimized (reducing the injection volume and injection rate) to minimize the single contrast agent usage while ensuring image quality; the above process is iterated repeatedly until the total contrast agent usage reaches the minimum value.

[0064] The system also sets a maximum limit for the total amount of contrast agent (usually the smaller of the patient's weight multiplied by 5ml or 3.7 multiplied by eGFR), accumulates the amount of contrast agent used in real time during the procedure, and issues an alert when the limit is approached.

[0065] The effect of this feature is that, through intelligent iterative optimization, the risk of contrast-induced nephropathy can be minimized while ensuring image quality.

[0066] The beneficial effects of the above-mentioned interventional surgery planning are as follows: the traditional two-dimensional planning is upgraded to three-dimensional precise planning through three-dimensional vascular model reconstruction; the optimal path selection and catheter / guidewire passability assessment are achieved through the determination of vascular access and interventional path; precise device matching based on individual vascular characteristics is achieved through interventional device selection; and the contrast agent dosage is minimized and the risk of contrast agent nephropathy is proactively prevented through iterative contrast agent usage protocols.

[0067] like Figure 2 As shown, in a preferred embodiment, the human-shaped blood vessel intervention execution module includes: The humanoid robot body is used to autonomously move to the surgical position and dock with the operating table. The end of the humanoid robot body is equipped with a standardized mechanical interface to adapt to different types of surgical execution modules. The guidewire control module is installed on the humanoid robot body through the standardized mechanical interface. It is used to control the advancement, retraction and rotation of the catheter and guidewire through a precision drive mechanism to deliver the interventional device to the target blood vessel segment. The angiography-guided module is used to acquire real-time imaging information of the vascular lumen through X-ray fluoroscopy and digital subtraction angiography, and to fuse the preoperative three-dimensional vascular model with the intraoperative fluoroscopic image in real time to assist navigation.

[0068] like Figure 3 As shown, it is a schematic diagram of a humanoid robot with a standardized mechanical interface.

[0069] Specifically, the humanoid robot body adopts a humanoid robot structure design, including an autonomous mobile chassis, a torso, and a surgical execution module mounted on the end effector.

[0070] The autonomous mobile chassis integrates omnidirectional moving wheels and a lidar navigation system, enabling the humanoid robot to navigate and move autonomously within the catheterization lab to reach the target parking position next to the operating table.

[0071] The docking mechanism is located at the connection between the chassis and the operating table. It adopts a combination design of precision guide pins and locking buckles. After the robot reaches the target position, the guide pin of the docking mechanism is precisely inserted into and locked with the pre-set docking hole on the side of the operating table, realizing a rigid connection between the humanoid robot body and the operating table.

[0072] The effect of this feature is that it enables the humanoid robot to autonomously reach the surgical site and precisely dock with the operating table, providing a stable physical support platform for subsequent interventional procedures.

[0073] Specifically, the standardized mechanical interface is located at the end of the humanoid robot body (arm end or torso end), and adopts a unified mechanical connection standard and electrical communication protocol, so that different types of surgical execution modules (such as guidewire control module, balloon dilation module, stent release module, etc.) can be quickly installed and disassembled through the same interface.

[0074] The standardized mechanical interface includes a mechanical locking part and an electrical connection part: the mechanical locking part adopts a quick-release bayonet design with a positioning pin. The surgical execution module is aligned with the bayonet and then rotated to lock, which can complete the loading and unloading operation within 10 seconds; the electrical connection part includes a power supply interface, a control signal interface and a sensor data interface, which automatically establishes an electrical connection while mechanically locking.

[0075] The effect of this feature is that it makes the humanoid robot body a universal surgical execution platform. By changing the end effector module, it can be adapted to different types of vascular interventional operations, which greatly improves the system's flexibility and scalability.

[0076] The guide wire control module is installed on the humanoid robot body through a standardized mechanical interface.

[0077] The guidewire control module uses a precision drive mechanism to control the advancement, retraction, and rotation of the catheter and guidewire with sub-millimeter precision.

[0078] The precision drive mechanism includes a high-precision servo motor, a planetary reducer, and a precision roller clamping system. The servo motor drives the precision rollers to apply controllable frictional force to the outer surface of the catheter or guidewire, thereby achieving the advance and retraction movements. The axial rotation of the catheter or guidewire is achieved through the rotary drive module.

[0079] During delivery, the motion parameters of the drive mechanism (propulsion speed, rotational angular velocity, clamping force, etc.) are adjusted in real time by the control module based on the angiography guidance information to ensure the accuracy and safety of the delivery action.

[0080] The effect of this feature is that it enables precise delivery and manipulation of catheters and guidewires within complex blood vessels.

[0081] Specifically, the angiography-guided module provides intravascular imaging information through two imaging methods: X-ray fluoroscopy and digital subtraction angiography (DSA).

[0082] X-ray fluoroscopy uses a low-dose pulse mode to display the position and movement of the catheter and guidewire within the blood vessel lumen in real time.

[0083] DSA obtains high-resolution images of vascular lesions by injecting contrast agents into the blood vessels and performing digital subtraction.

[0084] The angiography guidance module integrates the real-time fusion function of preoperative three-dimensional vascular model and intraoperative fluoroscopic image: X-ray projection parameters are obtained through the geometric calibration of the C-arm, the preoperative three-dimensional vascular model is virtually projected according to the current C-arm angle, and precise alignment is achieved with the intraoperative fluoroscopic image through feature point registration and deformation field optimization. The three-dimensional vascular anatomy is displayed on the two-dimensional fluoroscopic image in a semi-transparent color overlay mode, providing enhanced spatial navigation guidance for interventional operations.

[0085] The effect of this feature is that it effectively compensates for the lack of depth information and three-dimensional spatial sense in traditional two-dimensional perspective images, and significantly reduces the difficulty of spatial judgment in intervention operations.

[0086] The beneficial effects of the aforementioned humanoid vascular interventional execution module are as follows: rapid preoperative deployment is achieved through the autonomous movement and docking of the humanoid robot body; modular replacement of the surgical execution module and expansion of system capabilities are achieved through standardized mechanical interfaces; sub-millimeter precision delivery of interventional instruments is achieved through the precision drive mechanism of the guidewire control module; and enhanced intraoperative spatial navigation guidance is achieved through the three-dimensional fusion navigation of the angiography guidance module.

[0087] like Figure 4 As shown, in a preferred embodiment, the system further includes an emergency control module, which comprises: a control cabin, located outside the radiation protection area of ​​the operating room, for taking over interventional control actions when risks escalate or the operator actively triggers the intervention; a main control handle, for mapping the operator's hand movements to the advance, retraction, and rotation movements of the guidewire control module; a force feedback module, for transmitting the contact force between the distal end of the guidewire and the blood vessel wall to the operator's hand in the form of tactile feedback; and an image display module, for displaying intravascular image information acquired by the angiography guidance module in real time.

[0088] Specifically, the control cabin is the physical carrier of the emergency control module, located outside the radiation protection area of ​​the operating room, so that when the operator takes over the intervention operation performed by the humanoid robot in an emergency, he / she can complete all control actions without entering the radiation area.

[0089] The control cabin integrates a main control handle, a force feedback module, an image display module, and necessary communication equipment.

[0090] The emergency takeover is triggered in two ways: first, risk escalation triggering, which means that when the system detects an abnormal situation during the interventional procedure (such as abnormally increased guidewire pushing resistance, or the contact force of the blood vessel wall exceeding the safety threshold), the system automatically issues a risk escalation signal, prompting the operator to take over control through the control cabin; second, operator-initiated takeover, which means that the operator, based on their own judgment, can take over the interventional procedure at any time by pressing the takeover button on the control cabin.

[0091] The effect of this feature is that it provides a safe channel for human intervention in the interventional operations performed autonomously by humanoid robots, ensuring that the operator can still take over the surgical control in real time in a radiation-safe environment.

[0092] Specifically, the main control handle is an interactive device for manual operation by the user.

[0093] The main control handle adopts a multi-degree-of-freedom force feedback handle design, which can accurately capture the three-dimensional spatial movement of the operator's hand (pushing and pulling forward and backward, turning left and right, axial rotation, etc.), and convert the hand movement into the advance, retraction and rotation movement of the guidewire control module in real time through motion mapping algorithm.

[0094] The motion mapping algorithm supports adjustable motion scaling ratios. Operators can choose between coarse adjustment mode (1:1 mapping) or fine adjustment mode (1:5 to 1:10 mapping) according to their current operational accuracy requirements. In fine adjustment mode, larger movements of the operator's hand are mapped to smaller movements of the guidewire control module, enabling precise control of the guidewire.

[0095] The effect of this feature is that it enables the operator to precisely control the distal guidewire control module with natural, ergonomic hand movements.

[0096] Specifically, the force feedback module is a key component in the emergency control module for achieving a sense of tactile presence.

[0097] The force feedback module converts the real-time contact force signal between the distal end of the guidewire and the blood vessel wall into tactile feedback that can be perceived by the operator's hand through a high-fidelity force feedback actuator.

[0098] The force feedback process is as follows: the force sensor in the guidewire control module collects the axial pushing force and torque change data of the proximal end of the guidewire in real time. The original force signal is filtered, amplified and frequency compensated by the force signal processing algorithm. Then the processed force signal is transmitted to the force feedback actuator in the control cabin. The force feedback actuator generates a reaction force on the main control handle through motor drive, which corresponds to the contact force at the distal end, so that the operator can intuitively feel the contact state between the distal end of the guidewire and the blood vessel wall.

[0099] The effect of this feature is that it allows operators to obtain a tactile sense of presence similar to that when directly manipulating the guidewire in a remote control environment, significantly improving the accuracy and safety of emergency control.

[0100] Specifically, the image display module is used to display all intravascular image information acquired by the angiography guidance module in real time within the control cabin.

[0101] The image display module includes a high-resolution medical display screen and an image transmission unit. The real-time displayed content includes: X-ray fluoroscopy images (real-time position and travel status of the catheter and guidewire), DSA angiography images (clear images of vascular lesions), three-dimensional vascular fusion navigation view (fusion image of preoperative three-dimensional model and intraoperative fluoroscopy), and key intraoperative parameters (guidewire pushing force, cumulative contrast agent usage, etc.).

[0102] All images are displayed simultaneously in a multi-window layout to ensure that the operator can have a full grasp of the intraoperative status when taking over the operation.

[0103] The effect of this feature is that it provides the operator with complete intraoperative imaging information, enabling emergency takeover control to have the same information acquisition capabilities as the main intraoperative operation.

[0104] The beneficial effects of the aforementioned emergency control module are as follows: by setting up the radiation protection area outside the control cabin, the operator is kept away from radiation hazards; the timeliness of emergency control is ensured through two takeover mechanisms: risk escalation triggering and operator-initiated takeover; precise remote control of the guidewire control module is achieved through motion mapping of the main control handle; tactile presence is achieved through the force feedback module; and the image display module ensures that the operator has a comprehensive grasp of the intraoperative status.

[0105] In a preferred embodiment, the humanoid vascular interventional execution module includes a control module and a radiation-resistant humanoid torso. A guidewire control arm and a catheter control arm are mounted on the radiation-resistant humanoid torso. The guidewire control arm is equipped with a dexterous hand for controlling the advancement, retraction, and rotation of the guidewire. The catheter control arm is equipped with a dexterous hand for controlling the advancement, retraction, and rotation of the catheter, simulating the coordinated hand-to-hand delivery operation mode of an interventional physician. The control module generates control commands based on intraoperative angiography images to control the guidewire control arm and catheter control arm to perform interventional operations.

[0106] Specifically, the radiation-resistant humanoid torso is constructed from radiation-resistant materials and integrates a radiation shielding structure, used to replace the physical position of human interventional physicians in a continuous X-ray irradiation environment.

[0107] The radiation-resistant humanoid torso is designed with ergonomics in mind, and its height and shoulder width are comparable to those of an adult human doctor, allowing it to directly replace interventional doctors in standing below the C-arm to perform catheter and guidewire operations.

[0108] The torso uses carbon fiber composite material as the main structural material. This material has excellent radiation resistance and mechanical strength, and will not degrade or degrade in performance under long-term X-ray irradiation.

[0109] Key electronic components employ a radiation-hardening design, and an integrated radiation shielding layer provides focused protection for precision sensors and control circuits.

[0110] The effect of this feature is that it enables humanoid robots to work stably for extended periods in high-radiation environments, fundamentally solving the radiation exposure problem for interventional physicians.

[0111] Specifically, the guidewire control arm and the catheter control arm are respectively located on the left and right sides of the radiation-resistant humanoid torso, simulating the division of labor between the left and right hands of a human interventional physician.

[0112] The guidewire control arm is responsible for controlling the advancement, retraction, and rotation of the guidewire, corresponding to the fine manipulation function of the interventional physician's right hand (primary hand); the catheter control arm is responsible for controlling the advancement, retraction, and rotation of the catheter, corresponding to the support and guidance function of the interventional physician's left hand (secondary hand).

[0113] Both arms employ a multi-joint humanoid arm structure, including the shoulder joint (3 degrees of freedom), elbow joint (1 degree of freedom), and wrist joint (3 degrees of freedom), for a total of 7 degrees of freedom per arm, providing movement flexibility comparable to that of a human arm.

[0114] The dual-arm collaborative delivery mode simulates the operating habits of interventional surgeons in actual surgery: the guidewire arm advances the guidewire first to explore the path, and the catheter arm follows to push the catheter along the guidewire path; when vascular superselection is required, the catheter arm fixes the catheter position, and the guidewire arm independently controls the guidewire to enter the target branch vessel.

[0115] The effect of this feature is that, through the humanoid dual-arm collaborative operation mode, the robot's interventional actions are consistent with the operating habits of human doctors, reducing the complexity of the control algorithm and the operational risks.

[0116] Specifically, the dexterous hand is a key actuator at the end of the guidewire control arm and the catheter control arm, respectively used for precise manipulation of the guidewire and the catheter.

[0117] The design concept of the dexterous hand originates from bionics, simulating the human finger's ability to precisely grasp and manipulate slender, flexible instruments.

[0118] The guidewire control arm's dexterous hand comprises two sets of symmetrical finger-like gripping mechanisms and an axial rotation mechanism: the finger-like gripping mechanism achieves controllable gripping of the guidewire's outer surface through flexible hinges and micro piezoelectric ceramic actuators, with the gripping force precisely adjustable within the range of 0.1N to 5N, ensuring reliable gripping of the guidewire without damaging the guidewire coating; the axial rotation mechanism employs a hollow rotary motor design, allowing the guidewire to rotate axially at any angle while gripped.

[0119] The dexterous hand structure of the catheter control arm is similar, but the opening diameter and clamping force range of the clamping mechanism have been adapted to accommodate the larger outer diameter of the catheter.

[0120] The dexterous hand also integrates a force sensor to sense changes in the contact force between the guidewire or catheter and the blood vessel wall in real time, providing sensing data for force-sensing closed-loop control.

[0121] The effect of this feature is that it enables the end effector of the humanoid robot to have a fine control capability comparable to that of a human doctor's finger, and to perform sub-millimeter precision advance, retraction and rotation control of guidewires and catheters.

[0122] Specifically, the control module is a central control unit with a dual-arm humanoid configuration, which generates control commands based on intraoperative angiography images to control the guidewire manipulator arm and catheter manipulator arm to perform interventional operations.

[0123] The control module receives real-time fluoroscopic images and DSA images provided by the angiography guidance module. It extracts key information such as the current position of the guidewire and catheter, the direction of blood vessel course, the location of blood vessel bifurcation, and the location of the lesion through a vascular image analysis algorithm. Then, it generates coordinated movement control commands for both arms in combination with the interventional surgery planning scheme.

[0124] The control module also receives feedback data from the dexterous hand force sensor to achieve force closed-loop control: when the contact force at the distal end of the guidewire exceeds the preset safety threshold, the advance speed is automatically reduced or a retraction operation is performed; when the guidewire has passed through the curved blood vessel segment, the normal advance speed is automatically restored.

[0125] The effect of this feature is that it enables intelligent delivery control based on both visual and force feedback, making the delivery process of guidewires and catheters safer and more efficient.

[0126] The beneficial effects of the aforementioned dual-arm humanoid configuration are as follows: by replacing the physical position of human doctors in the radiation area with an anti-radiation anthropomorphic torso, the radiation exposure of interventional physicians is fundamentally eliminated; the dual-arm collaborative delivery mode of the guidewire control arm and catheter control arm faithfully simulates the clinical habits of interventional physicians' two-handed operation; the precise gripping and force sensing of the dexterous hand enables sub-millimeter-level precision control of the guidewire and catheter; and the dual closed-loop control of vision and force sensing in the control module enables the execution of intelligent delivery strategies.

[0127] like Figure 5 As shown, in a preferred embodiment, the system further includes a coordination module for realizing spatial linkage between the humanoid robot body and the intraoperative image acquisition device, and automatically coordinating the posture of the humanoid robot body to avoid the image acquisition space.

[0128] Specifically, the collaborative module is responsible for resolving the physical interference between the humanoid robot body and the intraoperative image acquisition equipment (mainly the C-arm angiography system) in the surgical space.

[0129] In vascular interventional surgery, the C-arm needs to rotate around the patient at multiple angles to obtain vascular images from different projection angles, while the humanoid robot body station is located next to the operating table to control the guidewire and catheter. The two may collide or obstruct each other in space.

[0130] The collaborative module calculates the rotation trajectory and occupied space of the C-arm in real time through a spatial linkage algorithm, predicts the time window and spatial area where interference may occur, and then generates posture adjustment instructions for the humanoid robot body in advance. This enables the robot to automatically adjust its torso angle, arm posture and overall position before the C-arm rotates to collect data, leaving enough clearance space for the C-arm to rotate and collect data.

[0131] The effect of this feature is that it ensures that the C-arm can perform angiography at any angle without being physically obstructed or disturbed by the humanoid robot body.

[0132] Specifically, the realization of the spatial linkage relies on the collaborative work of three components: the spatial perception submodule, the trajectory planning submodule, and the motion coordination submodule.

[0133] The spatial perception submodule acquires the three-dimensional spatial pose information of the humanoid robot body and C-arm in real time through depth sensors and markers set on the humanoid robot body and C-arm, and constructs a real-time space occupancy model of the operating room.

[0134] The trajectory planning submodule calculates the complete motion trajectory of the C-arm rotating from its current position to the target angle based on the angiography acquisition angle command given by the doctor, and generates the spatial sweep volume along the trajectory (i.e., all the spatial areas that the C-arm may occupy during the rotation process).

[0135] The motion coordination submodule performs collision detection between the spatial sweep volume and the current spatial occupancy of the humanoid robot body. If an interference area is detected, the minimum adjustment posture of the humanoid robot body is calculated so that the robot can avoid the image acquisition space while maintaining the control capability of the guide wire and guide tube, and automatically recover to the optimal control posture after the C-arm rotation is completed.

[0136] The effect of this feature is that, through the coordinated operation of the three sub-modules, fully automatic and non-interventional spatial linkage and avoidance between the humanoid robot and the C-arm is achieved.

[0137] The beneficial effects of the above-mentioned collaborative module are as follows: the spatial linkage algorithm realizes the automatic coordination between the humanoid robot body and the intraoperative image acquisition equipment, avoiding physical collision and image occlusion problems; the posture pre-adjustment enables the C-arm to freely perform multi-angle angiography acquisition without being restricted by the robot's position; and the automatic recovery of the optimal control posture ensures that the collaborative avoidance does not affect the control accuracy and continuity of the guidewire and catheter.

[0138] like Figure 6As shown, in a preferred embodiment, the system further includes an embodied intelligent decision-making module, which is connected to the intelligent assessment module, surgical planning module, human-shaped vascular intervention execution module, and auxiliary rehabilitation module. This module is used to autonomously perform cognitive understanding and decision-making reasoning of intravascular multimodal perception information based on the vascular interventional surgery task objectives, generate guidewire and catheter delivery strategies and interventional operation strategies and send them to the human-shaped vascular intervention execution module, and autonomously iterate and correct the strategies based on execution feedback until the surgical task objectives are achieved.

[0139] Specifically, the embodied intelligent decision-making module adopts a multi-cognitive intelligent agent collaborative architecture, including multiple cognitive intelligent agents and a surgical situation sharing space. The multiple cognitive intelligent agents include: a perception cognitive intelligent agent, used to access intraoperative multimodal sensors, fuse multi-source perception data into a unified structured scene representation, and write the scene representation into the surgical situation sharing space; a planning intelligent agent, used to read the scene representation and surgical task objectives from the surgical situation sharing space, generate surgical sub-task sequences and operation strategies hierarchically, and write the operation strategies into the surgical situation sharing space; a motion intelligent agent, used to read the operation strategies from the surgical situation sharing space and convert the operation strategies into real-time motion control commands for the surgical operating device; and a safety intelligent agent, independent of the decision-making links of the perception cognitive agent, planning intelligent agent, and motion intelligent agent, possessing an independent perception path, used to perform security audits on the motion control commands, and trigger execution veto when a security risk is detected. The surgical situation sharing space serves as the communication medium between the multiple cognitive intelligent agents, employing an asynchronous read / write mechanism for information exchange and collaboration among the cognitive intelligent agents.

[0140] The perceptual and cognitive intelligent agent is responsible for the acquisition, fusion, and semantic processing of multimodal perceptual data during surgery.

[0141] The intelligent agent independently connects to the intraoperative multimodal sensing system, including angiography fluoroscopy equipment, end force / torque sensors, intraoperative DSA imaging equipment, and vascular path navigation system.

[0142] The perceptual and cognitive intelligent agent is based on a multimodal data fusion algorithm to transform heterogeneous sensor data into a unified structured scene representation.

[0143] The structured scene representation covers the following semantic dimensions: vascular state description (including vascular wall morphology annotation, stenosis estimation, blood perfusion status, lesion grading, etc.), instrument pose information (including the six-degree-of-freedom pose and motion state vector of each surgical instrument in the surgical area coordinate system), anatomical structure mapping (including the location of key anatomical landmarks, the delineation of safety boundaries, and the annotation of vascular wall safety boundaries), and abnormal event marking (including spatiotemporal identifiers of abnormal states such as bleeding detection, vascular perforation identification, and guidewire slippage determination).

[0144] Scene representations are continuously refreshed at a fixed frequency and written into the scene state domain of the surgical posture sharing space.

[0145] The other three cognitive agents obtain a unified understanding of the current surgical situation by reading the scene representation in the shared space, without having to process the raw sensor data independently, thus eliminating the risk of redundant computation and data consistency among multiple agents.

[0146] The hierarchical decomposition of surgical tasks and dynamic generation of operational strategies by the planning and reasoning intelligent agent.

[0147] The agent reads the structured scene representation provided by the perceptual and cognitive agent and the preset surgical task objectives from the surgical situation sharing space, and executes hierarchical surgical planning and reasoning.

[0148] At the task level, the planning and reasoning agent decomposes the overall surgical goal into an ordered sequence of sub-tasks (such as vascular access establishment → guidewire advancement → lesion arrival → interventional operation → effect verification → instrument withdrawal).

[0149] Explicit state transition conditions are defined between each subtask. The agent autonomously determines the completion status of the current subtask and whether the triggering conditions of subsequent subtasks are met based on the real-time changes in the scene representation.

[0150] At the strategy level, the planning and reasoning agent generates specific operation strategies for the currently executed sub-task. These operation strategies include positioning strategies (precise spatial coordinates of the target point and optimal entry angle), entry strategies (path selection, velocity curve planning, obstacle avoidance constraints), operation strategies (operation type, force range, duration parameters), and force control strategies (force threshold setting, compliance control parameters, adaptive gain coefficient).

[0151] The generated candidate policies are written to the policy queue field in the shared space for the motion control agent to consume and execute.

[0152] When the scene representation deviates significantly, the planning and reasoning agent automatically triggers a policy revision mechanism to reassess the applicability of the current policy and generate an alternative policy.

[0153] The motion control agent is responsible for converting and issuing real-time motion commands from the operational strategy to the surgical execution mechanism.

[0154] The agent reads the operation strategy from the strategy queue domain of the surgical situation sharing space and transforms the abstract strategy into a real-time motion control command sequence for the guidewire manipulator and catheter manipulator.

[0155] The motion control agent contains three functional sub-modules: The trajectory planning module maps high-level operation strategies into time-series motion trajectories in joint space and Cartesian space, and generates directly executable motion trajectories based on kinematic constraints, collision detection, and motion smoothness optimization criteria.

[0156] The force-position hybrid control module dynamically switches between position control mode and force control mode according to the operation type—switching to force control mode during guidewire pushing operation to precisely constrain the delivery force, and switching to position control mode during alignment and positioning operation to ensure spatial positioning accuracy—the force-position control parameters are adaptively adjusted based on real-time force feedback.

[0157] The real-time anomaly response module detects abnormal conditions such as sudden changes in contact force, slippage of machinery, and accidental collisions with millisecond-level latency. Without waiting for other cognitive agents to make decisions, it immediately executes local protective actions, including emergency braking, safe retraction, and joint locking, giving the motion control agent independent local safety decision-making capabilities.

[0158] Before the control commands generated by the motion control agent are sent to the surgical execution agency, they must be reviewed and approved by the safety protection agent.

[0159] The execution feedback data (including actual motion trajectory, end contact force, execution state, and trajectory deviation) is written back to the execution feedback domain of the shared space for the perception and cognition agent to read for closed-loop updates of the scene state.

[0160] The intelligent safety guardian is responsible for independent safety monitoring and risk veto throughout the entire surgical process.

[0161] This intelligent agent is architecturally independent of the decision-making chain of the other three cognitive agents, and possesses two key mechanisms: Independent perception pathway: The safety guardian intelligent agent directly accesses the raw sensor data stream, without relying on the scene representation generated by the perception and cognition intelligent agent in the surgical situation sharing space. This design ensures that even if the perception and cognition intelligent agent fails, experiences delays, or has abnormal data, the safety guardian intelligent agent can still independently complete the safety assessment of the surgical situation based on the raw data, thus forming a perception redundancy guarantee at the architecture level. One-vote veto mechanism: The security protection agent continuously monitors three types of information sources in parallel: the content written by each cognitive agent in the shared space, the raw sensor data obtained by its own independent perception pathway, and the consistency deviation between the two.

[0162] A veto is triggered when any of the following risk conditions are detected: the operation strategy violates anatomical safety constraints (such as the instrument trajectory intruding into the safety envelope of critical blood vessels or nerves), the motion command exceeds the force / velocity / acceleration safety threshold, there is a significant consistency deviation between the self-perception results and the shared space scene representation (indicating that the perception link may be malfunctioning), or the system's cumulative risk rating exceeds the preset safety threshold.

[0163] Upon triggering a veto, the safety guardian agent executes a tiered response strategy based on the risk level: Mild risk: Freezes current motion control commands, sends a strategy regeneration request to the planning and reasoning agent, pauses the surgical procedure but maintains the system's ready state, and automatically resumes execution after the new strategy passes safety review. Moderate risk: Downgrades the system control mode to physician-assisted mode, suspends the autonomous decision-making link, and allows the surgeon to determine the subsequent operation plan. Severe risk: Immediately triggers an emergency stop command, resets all motion control commands to zero, automatically retracts the surgical execution mechanism along a preset safe trajectory to a safe position, and the system enters a safety lock state, which can only be restarted after being manually unlocked by the physician.

[0164] The security guardian agent continuously writes real-time risk rating information into the risk rating domain of the shared space, which can be read by all cognitive agents, ensuring that the entire multi-agent system maintains a global consensus on the current security situation.

[0165] The surgical situation sharing space is a unified communication medium among the four cognitive agents (with the independent perception path of the security guardian agent being the only exception).

[0166] Essentially, it is a structured shared memory space with strict access control, internally divided into five functional data domains: Scene State Domain, written to at a fixed frequency by the perception and cognition agent, and read-only accessed by other agents, storing a unified structured representation of the current surgical scene, covering tissue state, instrument pose, anatomical structure mapping, and abnormal event markers; Task Progress Domain, written to and maintained by the planning and reasoning agent, recording the hierarchical execution progress of surgical tasks, including the set of completed sub-tasks, the current execution status of sub-tasks, and an ordered queue of sub-tasks to be executed; Policy Queue Domain, written to the planning and reasoning agent and consumed and read by the motion control agent, managing the sequence of operation policies to be executed in a first-in-first-out queue; Risk Rating Domain, written by the safety protection agent, with read access by all agents, recording the system's comprehensive risk rating index and the itemized assessment of specific risk factors in real time; Execution Feedback Domain, written by the motion control agent and read by the perception and cognition agent for closed-loop updates of the scene state, recording the actual execution results of motion commands, including the actual motion trajectory, end-effector contact force, execution status identifier, and deviation from the planned trajectory.

[0167] The shared space adopts an asynchronous read-write mechanism, where each cognitive agent independently performs read-write operations according to its own operating cycle, without blocking, waiting, or synchronization lock dependencies.

[0168] This mechanism enables the four cognitive agents to operate in true parallel: the perception and cognition agent continuously refreshes the scene state, the planning and reasoning agent continuously generates operation strategies, the motion control agent continuously issues motion commands, and the security protection agent continuously performs security monitoring. The four agents work independently at their optimal operating frequencies, achieving loosely coupled asynchronous information exchange and multi-agent collaboration through a shared space.

[0169] The implementation mechanisms of each cognitive agent are explained in detail below.

[0170] The multimodal fusion encoding of the perceptual and cognitive intelligent agent is implemented as follows: For the pan-vascular interventional surgery scenario, the visual input includes two visual sources: intraoperative angiographic fluoroscopy images and a 3D fused navigation view. Independent visual Transformer backbone networks are configured for each source. The input image is segmented into labeled sequences by image blocks, and spatial feature representations are obtained through Transformer encoding. The two labeled sequences are concatenated into a unified long sequence with learnable viewpoint category labels added. This sequence is then fed into a cross-viewpoint Transformer fusion layer. A multi-head self-attention mechanism is used to enable each viewpoint label to pay attention to each other's spatial information. After fusion, the output vectors of each viewpoint category label position are extracted, concatenated, and compressed into a fixed-dimensional scene visual vector using a multilayer perceptron. The proprioceptive encoder receives the joint angles and angular velocities of the guidewire and catheter manipulator arms, as well as the distal guidewire contact force signal collected by the dexterity hand force sensor. After submodal concatenation, the signal is fed into a multilayer fully connected network to output the proprioceptive embedding. The speech encoder performs surgical speech domain fine-tuning based on the Whisper model encoder part and obtains relevant knowledge fragments from the vascular interventional procedure knowledge base by combining a retrieval enhancement generation mechanism. The final scene representation is composed of visual vectors, ontological perception embeddings, and retrieval context vectors, and is continuously written into the scene state domain of the surgical posture sharing space.

[0171] The planning and reasoning agent is a variant of the vision-language large-scale model architecture. The visual projection layer maps scene representations to the labeled embedding space of the language model, concatenates them with textual system prompts, and then feeds them into the Transformer decoder to autoregressively generate operational strategies. Fine-tuning employs a low-rank adaptation method, training only the adapter weights and the visual projection layer, while freezing the weights of the base language model. For general vascular interventional surgery, a stage-adaptive meta-query mechanism maintains a dedicated stage state machine, including stages such as vascular access establishment, guidewire probing and advancement, vascular bifurcation selection, lesion area arrival, interventional operation execution, and safe instrument withdrawal. Each stage corresponds to a set of learnable query vectors, which are injected as additional key-value pairs into the decoder's cross-attention layer, guiding the model's attention to automatically favor the key information of the current stage—the guidewire probing stage focuses on the direction of vascular course and bifurcation structure, while the interventional operation stage focuses on lesion morphology and vascular wall condition. The planning and reasoning agent also generates a dual-arm coordination strategy to coordinate the collaborative delivery timing of the guidewire manipulator arm and the catheter manipulator arm. During stage switching, the meta-query vectors undergo linear interpolation for smooth transition within several frames to prevent abrupt changes in attention patterns.

[0172] The motion control agent employs a conditional flow matching framework to train the action generation model, defining a linear interpolation path from random noise to the target action trajectory. A vector field network is trained to learn the velocity field direction from the noise to the actual action. The vector field network uses a one-dimensional temporal UNet architecture, with policy conditions, scene conditions, and safety conditions injected layer by layer through a feature modulation mechanism. During inference, the action trajectory is obtained by progressively solving the learned vector field direction from the random noise using multi-step Euler integration. At each integration step, the latest conditions are read from the shared space to achieve dynamic coupling across agents. For the dual-arm collaboration characteristic of panvascular interventional surgery, the guidewire control arm and catheter control arm share scene conditions but receive independent policy conditions, each generating its own independent action trajectory. The action block size is dynamically adjusted according to the surgical stage: small blocks are used for fine control during guidewire probing and vessel bifurcation selection, large blocks are used for straight-line advancement to ensure motion continuity, and minimum blocks are used for emergency response to achieve frame-by-frame real-time reaction. During training, frequency domain regularization is applied to the generated trajectory, penalizing high-frequency components to suppress guidewire tip jitter. The motion increment output at each step of the integral is processed serially by safety constraint projection: the potential field projection of the blood vessel wall prevents the guidewire from intruding into the blood vessel wall, and force constraint truncation and velocity constraint truncation.

[0173] The safety guardian agent includes an adversarial diffusion module, sharing the same network architecture and conditional flow matching framework as the motion control agent, but with the opposite training objective—learning the action patterns most likely to lead to safety accidents given the current scene conditions. The training dataset consists of a pan-vascular interventional surgery safety accident database, including action trajectories in accident scenarios such as vascular perforation, vascular dissection, guidewire knotting, and thrombus dislodgement. During inference, multiple failed trajectories are generated in parallel from multiple random seeds to form a dangerous trajectory envelope. The optimal trajectory output by the motion control agent is dynamically time-normalized, and the distance between the failed trajectories and the optimal trajectory is measured. When the safe distance falls below a dynamic threshold, a graded safety response is triggered. In the contrastive learning dual-path verification mechanism, the safety guardian agent has a built-in independent lightweight safety perception encoder that directly extracts safety guidance features from the raw sensor data. Cosine similarity is calculated in real-time with the scene representation of the perception and cognition agent. When the similarity is below a threshold for several consecutive frames, it is considered a semantic conflict and arbitration is triggered. During training, negative samples are specifically selected from hidden risk scenarios such as weak areas of vascular wall calcification, occult dissections, and stress concentrations at vascular bends, enabling the safety encoder to learn to generate divergent signals when the perception pathway appears normal but a hidden risk actually exists.

[0174] The forward prediction engine embedded in the surgical situation sharing space adopts a lightweight causal Transformer architecture. It takes the recent frame-by-frame scene state embedding sequence and the current policy encoding as input, and regresses the scene state increments for each future time step through multiple prediction heads. For general vascular interventional surgery scenarios, the trend of vascular wall contact force changes and contrast agent diffusion dynamics are incorporated as additional predictive variables into the prediction model. The online calibration mechanism maintains the prediction bias statistics using exponential moving averages and periodically updates the bias parameters of the prediction network output layer, enabling the world model to continuously adapt to the individual patient's vascular biomechanical characteristics during surgery (such as the difference in guidewire push force prediction bias caused by differences in vascular wall elasticity and calcification levels among different patients). When the prediction bias exceeds the normal range for several consecutive frames, the prediction is marked as failed, the motion control agent switches to a pure reactive control mode, and the safety protection agent switches to a conservative threshold mode.

[0175] The data flow and collaborative relationships between the embodied intelligent decision-making module and other modules in the system are as follows: With the intelligent assessment module, the perceptual and cognitive agent obtains the patient's preoperative assessment results (including disease diagnosis, indication determination, access vessel assessment results, contrast agent tolerance assessment results, etc.) from the intelligent assessment module, encodes them as surgical task constraints, and writes them into the task progress domain of the surgical situation sharing space, providing constraint inputs for the planning and reasoning agent when formulating delivery strategies; With the surgical planning module, the planning and reasoning agent obtains the 3D vascular model, interventional path planning, interventional device selection scheme, and contrast agent usage scheme from the surgical planning module, serving as a reference benchmark for task hierarchy decomposition and operation strategy generation. Simultaneously, when the intraoperative scene representation shows a significant deviation between the actual vascular morphology and the preoperative model, the planning and reasoning agent feeds back the corrected vascular model parameters. The surgical planning module updates the surgical plan; between the motion control agent and the human-shaped vascular intervention execution module, the motion control agent sends real-time motion control commands to the human-shaped vascular intervention execution module (including the guidewire control arm, catheter control arm, and angiography guidance module); the perception and cognition agent receives angiography image information, force sensor data, and instrument position data from the human-shaped vascular intervention execution module as multimodal perception inputs; the safety protection agent directly accesses the raw sensor data stream of the execution module through an independent perception pathway for parallel safety monitoring; between the planning and reasoning agent and the auxiliary rehabilitation module, after the surgery is completed, the planning and reasoning agent transmits the intraoperative operation record (including the actual instruments used, the actual delivery path, the actual amount of contrast agent used, and the record of handling intraoperative abnormal events) to the auxiliary rehabilitation module through the task progress domain of the surgical situation sharing space, providing intraoperative data support for the formulation of personalized rehabilitation plans.

[0176] The beneficial effects of the embodied intelligent decision-making module based on the multi-cognitive agent collaborative architecture are as follows: Through the perceptual cognitive agent, a unified cognitive understanding of multimodal sensory information within blood vessels is achieved, transforming heterogeneous sensor data into standardized scene representations and eliminating the risk of data inconsistency between multiple modules; through the planning and reasoning agent, hierarchical autonomous intelligent decision-making for guidewire and catheter delivery actions is achieved, decomposing complex vascular interventional tasks into controllable sub-task sequences; through the motion control agent, precise conversion from high-level operational strategies to low-level motion commands is achieved, and the conditional flow matching framework ensures the smoothness and safety of the action trajectory; through the independent perceptual pathway and veto mechanism of the safety guardian agent, architectural-level safety redundancy is constructed, ensuring the system maintains a safety baseline even if any single perceptual link fails; through the asynchronous read / write mechanism of the surgical situation sharing space, true parallel operation and loosely coupled collaboration of the four cognitive agents are achieved, avoiding real-time losses caused by synchronous blockage; and through the online calibration mechanism of the forward prediction engine, the system can continuously adapt to the differences in vascular biomechanical characteristics of individual patients.

[0177] The introduction of the embodied intelligent decision-making module upgrades this system from a traditional master-slave teleoperated interventional robot to an embodied intelligent surgical robot system with autonomous decision-making capabilities.

[0178] This application provides a control method for the aforementioned humanoid intelligent pan-vascular interventional surgical robot system. The specific solution of this method is as follows: Figure 7 As shown below, the control method of the humanoid intelligent panvascular interventional surgical robot system will be described in detail.

[0179] Combination Figure 7 As shown, the control method of the humanoid intelligent panvascular interventional surgical robot system includes: S101, based on the patient's medical data and vital signs, performs targeted disease assessments and determines the indications for vascular interventional surgery, generates personalized auxiliary assessment results for the patient, and displays them. S102, Based on the patient's angiography image data, reconstruct a three-dimensional vascular model of the target vascular segment, and based on the three-dimensional vascular model, perform targeted interventional surgery planning for the patient, generate a personalized interventional surgery planning scheme, and display it. S103, according to the interventional surgery planning scheme, the interventional device is delivered to the target vascular segment via the vascular access, and the vascular interventional operation is performed based on intraoperative angiography guidance; S104 generates and displays a personalized postoperative rehabilitation plan for the patient based on their medical data and postoperative assessment results.

[0180] The control method of the humanoid intelligent panvascular interventional surgical robot system provided in the above embodiments of this application corresponds to the humanoid intelligent panvascular interventional surgical robot system provided in the embodiments of this application. Therefore, the specific content of the method corresponds to the humanoid intelligent panvascular interventional surgical robot system. The specific content can be referred to the records in the humanoid intelligent panvascular interventional surgical robot system, which will not be repeated in this application.

[0181] The humanoid intelligent panvascular interventional surgical robot system method provided in the above embodiments of this application is based on the same inventive concept as the humanoid intelligent panvascular interventional surgical robot system provided in the embodiments of this application, and has the same beneficial effects as the methods adopted, run or implemented by its stored application.

[0182] Based on the same inventive concept, another embodiment of this application provides an electronic device for implementing the control method of the humanoid intelligent pan-vascular interventional surgical robot system described in the above embodiments. Figure 8 As shown, the electronic device includes a memory 301 and a processor 303.

[0183] Memory 301 can be configured to store a program.

[0184] Additionally, memory 301 can also be configured to store various other data to support operation on the electronic device. Examples of this data include instructions for any application or method used to operate on the electronic device, contact data, phonebook data, messages, pictures, videos, etc.

[0185] Memory 301 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. Processor 303, coupled to memory 301, is used to execute programs in memory 301 for: Based on the patient's medical data and physical signs, targeted disease assessments and determination of indications for vascular interventional surgery are conducted, generating personalized auxiliary assessment results for the patient and displaying them. Based on the patient's angiography data, a three-dimensional vascular model of the target vascular segment is reconstructed. Based on the three-dimensional vascular model, a targeted interventional surgery plan is developed for the patient, generating a personalized interventional surgery plan, which is then displayed. According to the interventional surgery plan, the interventional device is delivered to the target vascular segment via a vascular approach, and the vascular interventional operation is performed based on intraoperative angiography guidance; Based on the patient's medical data and postoperative assessment results, a personalized postoperative rehabilitation plan is generated and displayed.

[0186] In this application, Figure 8 The diagram only shows some components and does not mean that the electronic device includes only these components. Figure 8 The components shown.

[0187] The electronic device provided in this embodiment is based on the same inventive concept as the control method of the humanoid intelligent panvascular interventional surgical robot system provided in this application embodiment, and has the same beneficial effects as the methods adopted, run or implemented by the application stored therein.

[0188] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0189] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0190] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0191] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory. Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0192] This application also provides a computer-readable storage medium corresponding to the control method of the humanoid intelligent panvascular interventional surgical robot system provided in the foregoing embodiments, wherein a computer program (i.e., a program product) is stored thereon, and the computer program, when run by a processor, executes the control method of the humanoid intelligent panvascular interventional surgical robot system provided in any of the foregoing embodiments.

[0193] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, optical disc read-only memory (CDROM), digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0194] The computer-readable storage medium provided in the above embodiments of this application and the control method of the humanoid intelligent panvascular interventional surgical robot system provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application stored therein.

[0195] It should be noted that numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of this application may be practiced without these specific details. In some instances, well-known structures and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0196] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, system, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, system, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, system, article, or apparatus that includes said element.

[0197] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A humanoid intelligent panvascular interventional surgical robot system, characterized in that, The system is built upon an embodied intelligence closed-loop architecture of perception-cognition-decision-execution-feedback, including: The intelligent assessment module is used to conduct targeted disease assessments and determine the indications for vascular interventional surgery based on the patient's medical data and vital signs data, generate personalized auxiliary assessment results for the patient, and display them. The surgical planning module is used to reconstruct a three-dimensional vascular model of the target vascular segment based on the patient's angiography image data, perform targeted interventional surgical planning for the patient based on the three-dimensional vascular model, generate a personalized interventional surgical planning scheme, and display it. The human-shaped vascular interventional execution module is used to deliver interventional instruments to the target vascular segment via a vascular approach according to the interventional surgery plan, and to perform vascular interventional operations based on intraoperative angiography guidance; The auxiliary rehabilitation module is used to generate and display personalized postoperative rehabilitation plans for patients based on their medical data and postoperative assessment results.

2. The humanoid intelligent panvascular interventional surgical robot system according to claim 1, characterized in that, The surgical planning module is configured as follows: A three-dimensional vascular model of the target vessel segment is reconstructed based on preoperative imaging data, and the bifurcation, stenosis and lesion sites are marked; the vascular access and interventional path are determined on the three-dimensional vascular model, and the passability of the catheter and guidewire is evaluated; interventional devices are selected according to the lesion type and vascular morphology, and the implantation location and operation parameters are determined; the contrast agent usage plan is iterated until the contrast agent dosage is minimized.

3. The humanoid intelligent panvascular interventional surgical robot system according to claim 1, characterized in that, The human-shaped blood vessel intervention execution module includes: The humanoid robot body is used to autonomously move to the surgical position and dock with the operating table. The end of the humanoid robot body is equipped with a standardized mechanical interface to adapt to different types of surgical execution modules. The guidewire control module is installed on the humanoid robot body through the standardized mechanical interface. It is used to control the advancement, retraction and rotation of the catheter and guidewire through a precision drive mechanism to deliver the interventional device to the target blood vessel segment. The angiography-guided module is used to acquire real-time imaging information of the vascular lumen through X-ray fluoroscopy and digital subtraction angiography, and to fuse the preoperative three-dimensional vascular model with the intraoperative fluoroscopic image in real time to assist navigation.

4. The humanoid intelligent panvascular interventional surgical robot system according to claim 1, characterized in that, The system also includes an emergency control module, which includes: The control cabin is located outside the radiation protection area of ​​the operating room and is used to take over and intervene in control actions when the risk escalates or is actively triggered by the operator. The main control handle is used to map the operator's hand movements into the advance, retraction, and rotation movements of the guidewire control module; The force feedback module is used to transmit the contact force between the distal end of the guidewire and the blood vessel wall to the operator's hand in the form of tactile feedback. The image display module is used to display intravascular imaging information acquired by the angiography guidance module in real time.

5. The humanoid intelligent panvascular interventional surgical robot system according to claim 1, characterized in that, The humanoid vascular interventional execution module includes a control module and a radiation-resistant humanoid torso. A guidewire control arm and a catheter control arm are mounted on the radiation-resistant humanoid torso. The guidewire control arm is equipped with a dexterous hand for controlling the advancement, retraction, and rotation of the guidewire. The catheter control arm is equipped with a dexterous hand for controlling the advancement, retraction, and rotation of the catheter, simulating the coordinated hand-to-hand delivery operation mode of an interventional physician. The control module generates control commands based on intraoperative angiography images to control the guidewire control arm and catheter control arm to perform interventional procedures.

6. The humanoid intelligent panvascular interventional surgical robot system according to claim 1, characterized in that, The system also includes a coordination module, which enables spatial linkage between the humanoid robot body and the intraoperative image acquisition device, and automatically coordinates the posture of the humanoid robot body to avoid the image acquisition space.

7. The humanoid intelligent panvascular interventional surgical robot system according to claim 1, characterized in that, The system also includes an embodied intelligent decision-making module, which is connected to the intelligent assessment module, surgical planning module, human-shaped vascular intervention execution module, and auxiliary rehabilitation module. It is used to autonomously perform cognitive understanding and decision-making reasoning of intravascular multimodal perception information according to the vascular intervention surgery task objectives, generate guidewire and catheter delivery strategies and intervention operation strategies and send them to the human-shaped vascular intervention execution module, and autonomously iterate and correct the strategies according to the execution feedback until the surgical task objectives are achieved.

8. A control method for a humanoid intelligent panvascular interventional surgical robot system, applied to the system according to any one of claims 1 to 7, characterized in that, include: Based on the patient's medical data and physical signs, targeted disease assessments and determination of indications for vascular interventional surgery are conducted, generating personalized auxiliary assessment results for the patient and displaying them. Based on the patient's angiography data, a three-dimensional vascular model of the target vascular segment is reconstructed. Based on the three-dimensional vascular model, a targeted interventional surgery plan is developed for the patient, generating a personalized interventional surgery plan, which is then displayed. According to the interventional surgery plan, the interventional device is delivered to the target vascular segment via a vascular approach, and the vascular interventional operation is performed based on intraoperative angiography guidance; Based on the patient's medical data and postoperative assessment results, a personalized postoperative rehabilitation plan is generated and displayed.

9. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores computer-executable instructions, and the processor executes the computer-executable instructions to implement the steps of the control method of claim 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the control method of claim 8.

Images