Hard magnetic driving vascular intervention guide wire robot system and method

Abstract

The invention provides a hard magnetic driving vascular intervention guide wire robot system and method, and the system comprises a hard magnetic flexible guide wire catheter which sleeves a hard magnetic flexible guide wire, and the tip of the hard magnetic flexible guide wire has hard magnetism and completes the stretching or retracting action in the catheter; the guide wire catheter driving device is used for clamping the hard magnetic flexible guide wire catheter and providing forward and backward driving force for the hard magnetic flexible guide wire catheter; the seventh-degree-of-freedom magnetic navigation system is arranged above the guide wire catheter driving device and comprises a mechanical arm, a servo rotating motor and a cylindrical permanent magnet, and the movement direction of the hard magnetic flexible guide wire is controlled by changing the relative position of the cylindrical permanent magnet and the sharp end of the hard magnetic flexible guide wire; the visual identification module is arranged between the guide wire catheter driving device and the seventh-degree-of-freedom magnetic navigation system and comprises a support and an industrial camera, and the position of the hard magnetic flexible guide wire in the blood vessel is returned through the industrial camera and used for follow-up path analysis. The method has great application potential in minimally invasive vascular interventional operations.

Description

Technical Field

[0001] The present invention relates to the technical field of vascular interventional treatment, and in particular to a hard magnetic driven vascular intervention guidewire robot system and method. Background Art

[0002] At present, the morbidity and mortality of cardiovascular and cerebrovascular diseases remain high, which has caused a great burden on people's physical health and the world's medical system. For most cardiovascular and cerebrovascular diseases, vascular interventional surgery (VIS) is an important treatment method. It is usually assisted by medical imaging equipment, using guidewires, catheters and other instruments to reach the lesions through blood vessels for diagnosis and treatment. However, traditional medical guidewire catheters cannot achieve autonomous propulsion and path selection functions, but rely on the doctor's remote manual operation outside the body during surgery. Therefore, VIS has high requirements for the doctor's relevant experience and technology. On the other hand, the operation must be performed under X-rays. The longer the operation time, the greater the damage to the health of the doctors and patients involved in the operation.

[0003] In response to the above problems, researchers began to consider combining microrobots and their control technology to complete the directional delivery of guidewires, stents or drugs in the human body. However, most of the existing vascular interventional robots use master-slave control, and doctors still need to manually control the catheter or guidewire to achieve simpler guidewire catheter advancement and rotation movements, and cannot achieve autonomous steering. They still lack flexibility and controllability at complex vascular bifurcations. In the medical field, especially in in vivo interventional treatment, the application potential of cordless microrobots is enormous. Among them, magnetic drive has the characteristics of remote drive, precise control, miniaturization, biocompatibility and multimodal manipulation, and has become one of the preferred drive methods for medical robots. At present, scholars have developed magnetically driven microrobots that can be used for intestinal targeted drug delivery, vascular targeted treatment and minimally invasive surgery, and verified their feasibility.

[0004] The development of a vascular interventional robot system that can propel and steer autonomously can assist doctors in remote control, effectively reduce their radiation exposure, and eliminate physiological tremors caused by manual operation and fatigue, thereby improving surgical accuracy. Summary of the invention

[0005] In view of the defects in the prior art, the object of the present invention is to provide a hard magnetic driven vascular intervention guidewire robot system and method.

[0006] A hard magnetic driven vascular intervention guidewire robot system provided by the present invention comprises:

[0007] Hard magnetic flexible guidewire catheter: The flexible catheter is sheathed on the outside of the hard magnetic flexible guidewire. The tip of the hard magnetic flexible guidewire has hard magnetism and can be extended or retracted in the catheter.

[0008] Guidewire catheter driving device: clamps the hard magnetic flexible guidewire catheter and provides driving force for the hard magnetic flexible guidewire catheter to move forward and backward;

[0009] The seventh degree of freedom magnetic navigation system is arranged above the guidewire catheter drive device, including a mechanical arm, a servo rotary motor and a cylindrical permanent magnet. The movement direction of the hard magnetic flexible guidewire is controlled by changing the relative position of the cylindrical permanent magnet and the tip of the hard magnetic flexible guidewire;

[0010] Visual recognition module: It is set between the guidewire catheter drive device and the seventh degree of freedom magnetic navigation system, including a bracket and an industrial camera. The industrial camera returns the position of the hard magnetic flexible guidewire in the blood vessel for subsequent path analysis.

[0011] Preferably, in the hard magnetic flexible guidewire catheter:

[0012] The hard magnetic flexible guide wire is composed of a titanium alloy inner core and a braided structure, and is coated with polytetrafluoroethylene material on the outside; the guide wire tip sheath is filled with axially magnetized neodymium iron boron magnetic particles;

[0013] The flexible catheter is made of thermoplastic resin material and is sheathed on the outside of the hard magnetic flexible guide wire.

[0014] Preferably, in the guidewire catheter driving device:

[0015] There are 2 groups of drive modules, which respectively clamp the hard magnetic flexible guide wire and the flexible catheter, and drive the hard magnetic flexible guide wire catheter to move forward and backward in the blood vessel; each drive module is composed of a pair of friction wheels, a servo motor, a worm gear, a power supply and a controller; the end of the servo motor shaft is connected to a worm gear meshing with the worm gear, and the worm gear is coaxially connected to the driving wheel of the friction wheel. The flexible catheter is clamped between the driving wheel and the driven wheel of the friction wheel, and the servo motor drives the worm gear to rotate, thereby driving the driving wheel and the friction wheel to rotate. The flexible guide wire is clamped in the wire feeding mechanism, and the servo motor drives the worm gear to rotate, thereby driving the driving wheel and the friction wheel to rotate. The code is transmitted through the control board, and then the hard magnetic flexible guide wire or the flexible catheter is controlled to move forward and backward respectively;

[0016] Positioning shell: Two sets of drive modules are embedded in the bottom of the positioning shell, and the controller and power supply are fixed to the shell around the positioning shell.

[0017] Preferably, in the seventh degree of freedom magnetic navigation system:

[0018] 6-DOF robotic arm;

[0019] The servo motor is arranged at the end of the 6-DOF robotic arm;

[0020] The upper connecting seat is arranged between the 6-DOF mechanical arm and the servo motor;

[0021] The lower connecting seat is arranged at the lower part of the upper connecting seat;

[0022] The special-shaped central axis is arranged in the central hole of the lower connecting seat;

[0023] The top cover is arranged outside the special-shaped central axis;

[0024] The connecting plate is arranged at the end of the special-shaped central axis;

[0025] The cylindrical box is arranged below the connecting plate;

[0026] The cylindrical permanent magnet is arranged inside the cylindrical box.

[0027] Preferably, in the visual recognition module:

[0028] Includes: base;

[0029] The lower column is arranged on the base;

[0030] The upper column is arranged on the lower column;

[0031] The positioning clamp is arranged outside the upper column or the lower column;

[0032] The industrial camera is arranged in a positioning fixture.

[0033] According to a hard magnetic driven vascular intervention guidewire robot method provided by the present invention, using any one of the hard magnetic driven vascular intervention guidewire robot systems, the execution includes:

[0034] Step S1: color marking the tip of the guidewire to obtain the original two-dimensional image required for vascular intervention;

[0035] Step S2: convert the collected original image into the HSV color space, generate a binary mask of the image, extract the target color area contour, and extract the center coordinates of the minimum circumscribed rectangle;

[0036] Step S3: Predicting the position of the obscured guide wire tip based on a convolutional neural network and constructing a model;

[0037] Step S4: Collect images containing the features of the hard magnetic flexible guide wire tip and train them according to the model.

[0038] Preferably, in step S2:

[0039] The collected original image is converted from BGR color space to HSV color space, and a specific HSV threshold is set to separate different color ranges; the color mark set at the tip of the guide wire is used as the target, and a binary mask of the image is generated according to the set domain value. The white part of the mask represents pixels that meet the conditions, and the black part represents pixels that do not meet the conditions; the mask is expanded into the same shape as the original image, applied to the original image channel by channel, the outline of the target color area is extracted, the minimum enclosing rectangle is fitted to the outline, and the center coordinates of its minimum enclosing rectangle are extracted.

[0040] Preferably, in step S3:

[0041] The pose prediction of the occluded guide wire tip is performed based on the convolutional neural network. The model is constructed by including multiple convolutional layers, maximum pooling layers, activation layers and fully connected layers. The image features are extracted through the convolutional layer, and the convolutional layer output is nonlinearly transformed using the ReLU activation function. The data dimension is reduced through the maximum pooling layer, and the image data is batch normalized. The flattening layer is connected to the fully connected layer, and the high-level feature map is mapped to the output space. The final result is generated through the output layer.

[0042] Preferably, in step S4:

[0043] Model training: collect pictures containing hard magnetic flexible guide wire tip features, normalize and data enhance them to obtain a learning data set, identify the tip position of each picture and obtain its tip center coordinates as feature labels; extract 30% of the pictures in the data set, perform local color occlusion on the guide wire tip, and train the position recognition ability under occlusion; divide the data set into a training set and a validation set, use the mean square error as the loss function, and optimize the model parameters through the Adam optimizer. The MSE loss function calculation formula is as follows:

[0044]

[0045] Among them, y i are the actual coordinates, are the prediction coordinates and N is the number of samples.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] 1. The present invention utilizes a cordless hard magnetic drive method to achieve directional movement of the guidewire catheter in the blood vessel, solves the problem that the movement direction of the traditional guidewire catheter in the blood vessel is difficult to control, and realizes automatic control of the feedback closed loop.

[0048] 2. The present invention has a simple manufacturing process, is easy to operate, and has great application potential in minimally invasive vascular interventional surgery. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0050] Figure 1 A top view of the hard magnetic driven vascular intervention guidewire robot system model of the present invention;

[0051] Figure 2 A three-dimensional view of the hard magnetic driven vascular intervention guidewire robot system model of the present invention;

[0052] Figure 3 It is a schematic diagram of the structure of the magnetic navigation device in the system of the present invention;

[0053] Among them: 1 is a medical catheter; 2 is a medical guide wire; 3 is a tube feeding motor; 4 is a driving wheel; 5 is a driven wheel; 6 is a wire guide wheel; 7 is a wire limiting mechanism; 8 is a wire feeding and tube feeding mechanism control board; 9 is a driving component housing; 10 is a worm gear; 11 is a wire feeding motor; 12 is a camera fixing device; 13 is a base; 14 is an industrial camera; 15 is a blood vessel model; 16 is a magnetic navigation drive motor; 17 is a flange device; 18 is a robotic arm; 19 is a cable; 20 is a cylindrical permanent magnet; 21 is an upper connecting seat; 22 is a lower connecting seat; 23 is a special-shaped central axis; 24 is a top cover; 25 is a connecting plate; 26 is a lower column; 27 is an upper column; 28 is a positioning clamp; 29 is a cylindrical box. DETAILED DESCRIPTION

[0054] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0055] Embodiment 1:

[0056] The present invention provides a hard magnetic driven vascular intervention guidewire robot system. It includes: a hard magnetic flexible guidewire catheter for establishing a vascular access in vascular intervention surgery, two sets of guidewire catheter driving mechanisms for driving the guidewire catheter, a magnetic navigation mechanism for guiding the movement direction of the guidewire catheter, a vascular model for simulating the vascular environment, and a visual perception module for image recognition and feedback. The intervention catheter is sleeved on the outside of the hard magnetic flexible guidewire. The guidewire is first pushed forward by the guidewire catheter driving mechanism. At the same time, the visual perception module returns the camera shooting picture in real time, and controls the magnetic navigation movement to the corresponding position and posture according to the preset path and the actual position of the guidewire, so that the guidewire enters the designated vascular bend or fork. If the guidewire is detected to have entered the wrong section during the process, the system movement is immediately stopped, the guidewire is retracted, and the movement path is corrected until the guidewire reaches the designated end position.

[0057] A hard magnetic driven vascular intervention guidewire robot method, comprising:

[0058] Step S1: color marking the tip of the guidewire to obtain the original two-dimensional image required for vascular intervention;

[0059] Step S2: convert the collected original image into the HSV color space, generate a binary mask of the image, extract the target color area contour, and extract the center coordinates of the minimum circumscribed rectangle;

[0060] Specifically, in step S2:

[0061] The collected original image is converted from BGR color space to HSV color space, and a specific HSV threshold is set to separate different color ranges; the color mark set at the tip of the guide wire is used as the target, and a binary mask of the image is generated according to the set domain value. The white part of the mask represents pixels that meet the conditions, and the black part represents pixels that do not meet the conditions; the mask is expanded into the same shape as the original image, applied to the original image channel by channel, the outline of the target color area is extracted, the minimum enclosing rectangle is fitted to the outline, and the center coordinates of its minimum enclosing rectangle are extracted.

[0062] Step S3: Predicting the position of the obscured guide wire tip based on a convolutional neural network and constructing a model;

[0063] Specifically, in step S3:

[0064] The pose prediction of the occluded guide wire tip is performed based on the convolutional neural network. The model is constructed by including multiple convolutional layers, maximum pooling layers, activation layers and fully connected layers. The image features are extracted through the convolutional layer, and the convolutional layer output is nonlinearly transformed using the ReLU activation function. The data dimension is reduced through the maximum pooling layer, and the image data is batch normalized. The flattening layer is connected to the fully connected layer, and the high-level feature map is mapped to the output space. The final result is generated through the output layer.

[0065] Step S4: Collect images containing the features of the hard magnetic flexible guide wire tip and train them according to the model.

[0066] Specifically, in step S4:

[0067] Model training: collect pictures containing hard magnetic flexible guide wire tip features, normalize and data enhance them to obtain a learning data set, identify the tip position of each picture and obtain its tip center coordinates as feature labels; extract 30% of the pictures in the data set, perform local color occlusion on the guide wire tip, and train the position recognition ability under occlusion; divide the data set into a training set and a validation set, use the mean square error as the loss function, and optimize the model parameters through the Adam optimizer. The MSE loss function calculation formula is as follows:

[0068]

[0069] Among them, y i are the actual coordinates, are the prediction coordinates and N is the number of samples.

[0070] Embodiment 2:

[0071] Embodiment 2 is a preferred example of Embodiment 1, and is used to illustrate the present invention in more detail.

[0072] The present invention also provides a hard magnetic driven vascular intervention guidewire robot system, which can be realized by executing the process steps of the hard magnetic driven vascular intervention guidewire robot method, that is, those skilled in the art can understand the hard magnetic driven vascular intervention guidewire robot method as a preferred embodiment of the hard magnetic driven vascular intervention guidewire robot system.

[0073] According to the present invention, a hard magnetic driven vascular intervention guidewire robot system is provided. Figure 3 As shown, including:

[0074] Hard magnetic flexible guidewire catheter: the flexible catheter 1 is sleeved on the outside of the hard magnetic flexible guidewire 2, the tip of the hard magnetic flexible guidewire 2 has hard magnetism, and can be extended or retracted in the catheter;

[0075] Preferably, in the hard magnetic flexible guidewire catheter:

[0076] The hard magnetic flexible guide wire 2 is composed of a titanium alloy inner core and a braided structure, and is coated with polytetrafluoroethylene material on the outside; the guide wire tip sheath is filled with axially magnetized neodymium iron boron magnetic particles;

[0077] The flexible catheter 1 is made of thermoplastic resin material and is sheathed on the outside of the hard magnetic flexible guide wire.

[0078] The guidewire catheter driving device 9 is used to clamp the hard magnetic flexible guidewire catheter and provide driving force for the hard magnetic flexible guidewire catheter to move forward and backward;

[0079] Preferably, in the guidewire catheter driving device 9:

[0080] There are 2 groups of drive modules, which respectively clamp the hard magnetic flexible guide wire 2 and the flexible catheter 1, and drive the hard magnetic flexible guide wire catheter to move forward and backward in the blood vessel; each drive module is composed of a pair of friction wheels, a servo motor 3, a worm gear 10, a power supply and a controller 8; the end of the servo motor 3 shaft is connected to a worm meshing with the worm wheel, and the worm wheel is coaxially connected to the driving wheel of the friction wheel. The flexible catheter 1 is clamped between the driving wheel 4 and the driven wheel 5 of the friction wheel, and the servo motor drives the worm gear 10 to rotate, thereby driving the driving wheel 4 and the friction wheel 5 to rotate. The flexible guide wire 2 is clamped in the wire feeding mechanism 6, and the servo motor drives the worm gear 10 to rotate, thereby driving the driving wheel 4 and the friction wheel 5 to rotate. The code is transmitted through the control board, and then the hard magnetic flexible guide wire 2 or the flexible catheter 1 is controlled to move forward and backward respectively;

[0081] Positioning housing 9: Two sets of driving modules are embedded in the bottom of the positioning housing, and the controller and power supply are fixed to the shell around the positioning housing 9.

[0082] The seventh degree of freedom magnetic navigation system 17 is arranged above the guidewire catheter driving device 9, and includes a mechanical arm 18, a servo rotary motor 16 and a cylindrical permanent magnet 20. The moving direction of the hard magnetic flexible guidewire 2 is controlled by changing the relative position between the cylindrical permanent magnet 20 and the tip of the hard magnetic flexible guidewire 2;

[0083] Preferably, in the seventh degree of freedom magnetic navigation system 17:

[0084] 6-DOF robotic arm 18;

[0085] The servo motor 16 is disposed at the end of the 6-DOF robotic arm 18;

[0086] The cable 19 is arranged at the end of the servo motor 16;

[0087] The upper connecting seat 21 is disposed between the 6-DOF mechanical arm 18 and the servo motor 16;

[0088] The lower connecting seat 22 is arranged at the lower part of the upper connecting seat 21;

[0089] The special-shaped central axis 23 is arranged in the central hole of the lower connecting seat 22;

[0090] The top cover 24 is arranged outside the special-shaped central axis 23;

[0091] The connecting plate 25 is arranged at the end of the special-shaped central axis 23;

[0092] The cylindrical box 29 is disposed below the connecting plate 25;

[0093] The cylindrical permanent magnet 20 is disposed inside the cylindrical box 29 .

[0094] Visual recognition module: It is arranged between the guidewire catheter driving device 9 and the seventh degree of freedom magnetic navigation system 17, including a bracket 12 and an industrial camera 14, and returns the position of the hard magnetic flexible guidewire 2 in the blood vessel through the industrial camera for subsequent path analysis.

[0095] Preferably, in the visual recognition module:

[0096] The invention comprises: a base 13;

[0097] The lower column 26 is disposed on the base 13;

[0098] The upper column 27 is arranged on the lower column 26;

[0099] The positioning clamp 28 is arranged outside the upper column 27 or the lower column 26;

[0100] The industrial camera 14 is disposed in a positioning fixture 28 .

[0101] Embodiment 3:

[0102] Embodiment 3 is a preferred example of Embodiment 1, and is used to illustrate the present invention in more detail.

[0103] The present invention provides a hard magnetic driven vascular intervention guidewire robot system. The system controls the travel direction of the hard magnetic guidewire through a permanent magnet, observes and checks in real time whether the guidewire is located in the expected path and adjusts its posture, thereby realizing automatic control of the guidewire catheter. The purpose of the present invention is to realize the controllable movement of the guidewire catheter in the blood vessel, and based on the visual recognition results, realize the coordinated motion control of the robot arm and the guidewire robot, and complete the closed-loop control of the system.

[0104] The hard magnetic driven vascular intervention guidewire robot system includes: a guidewire catheter driving device; a hard magnetic flexible guidewire catheter clamped in the guidewire catheter driving device; a seventh degree of freedom magnetic navigation system arranged above the guidewire catheter driving device; a visual recognition module arranged between the guidewire catheter driving device and the seventh degree of freedom magnetic navigation system; and a blood vessel model connected to the guidewire catheter driving device.

[0105] like Figure 1 as well as Figure 2As shown, the propulsion system of the vascular interventional instrument in this example includes a medical catheter 1, a guide wire 2, a tube feeding motor 3, a driving wheel 4, a driven wheel 5, a guide wire wheel 6, a wire limiting mechanism 7, a wire feeding and tube feeding mechanism control board 8, a housing 9, a worm gear 10 and a wire feeding motor 11. The driving wheel 4 and the driven wheel 5 are respectively clamped on both sides of the medical catheter 1 and the driving wheel 4 is connected to the worm gear 10. The worm gear 10 drives the driving wheel to rotate to propel the medical catheter 1 forward or backward, and the forward and backward movement of the medical guide wire 2 is completed by the guide wire wheel 6 and the wire feeding motor 11. The driving component includes the tube feeding motor 3, the wire feeding and tube feeding mechanism control board 8 and the wire feeding motor 11. The control board 8 can control the medical catheter 1 and the guide wire 2 to perform linear motion respectively. The vascular model component includes a vascular model 15.

[0106] The motion strategy of the hard magnetic driven vascular intervention guidewire robot system provided in this example is as follows: First, the internal parameter calibration of the camera 14 and the external parameter calibration of the camera and the robotic arm coordinate system are performed to obtain the conversion relationship between the camera coordinate system and the robotic arm coordinate system. Subsequently, the camera identifies the position of the guidewire tip and the node position in the path. While the guidewire 2 is moving forward at a constant speed, when the camera identifies that the guidewire is approaching a turning node in the path, the robotic arm is controlled to guide the guidewire tip in the direction so that the guidewire moves forward along the pre-specified correct path. After the guidewire passes through the node, the position of the guidewire tip is identified by the camera to determine whether the guidewire is moving on the correct path. If it is determined that the guidewire is not on the correct path, the guidewire is controlled to retract to the previous node and pass through the node again.

[0107] Specifically, the present invention provides a method for visually identifying and separating a blood vessel path, including: introducing a SAM (Segment Anything with Memory) predictor module and a model registry: wherein the SAM predictor module provides classes and functions for loading, configuring and using the SAM model, in which the user can search and load the corresponding pre-trained model according to the specified model type or name; determining the target marker point, calling the predict method to predict the input image: using the loaded model to segment the input image or other related tasks to obtain the required output result; the model selects the mask with the highest score in the prediction result, and its corresponding score and visualizes it. In order to achieve precise control, the system is provided with a path with a width of only one or several pixels, and a skeletonization function skeletonize is introduced, which can reduce the pixels of each object (such as blood vessels, text or other connected areas) in the original binary image to its center line by gradually thinning the image, while maintaining the topological structure of the object, and the final generated "skeleton" is a single-pixel wide line network.

[0108] Specifically, the present invention also provides a method for visually identifying and predicting the position of the tip of a hard magnetic flexible guidewire in a blood vessel, including: color marking the tip of the guidewire to obtain the original two-dimensional image required for vascular intervention; converting the collected original image from the BGR color space to the HSV color space, extracting the target color area contour, fitting the contour with a minimum circumscribed rectangle, and extracting the center coordinates of its minimum circumscribed rectangle; establishing a machine learning model based on a convolutional neural network to predict the posture of the occluded guidewire tip, and performing model training, including: normalizing and data enhancing images containing the tip features of the hard magnetic flexible guidewire, identifying the tip position to obtain the tip center coordinates as a feature label; training the position recognition capability under occlusion; dividing the data set into a training set (85%) and a validation set (15%), using mean squared error (MSE) as the loss function, and optimizing the model parameters through the Adam optimizer.

[0109] Specifically, the present invention also provides a node-based hard magnetic flexible guidewire motion strategy, including: the node is a large turning point of the vascular curvature, and the specific curvature critical value is determined according to the quality of the extracted vascular path. For each valid node, the actions required to be completed by each device are consistent, and a large loop is entered to traverse each valid node. In each loop, the system records the positions of the current valid node, the previous node, and the next node. The vector formed from the current node to the tip position of the guidewire is A, the vector formed from the current node to the next node is B, and the vector formed from the current node to the previous node is C. These vectors can be used to determine whether the guidewire has reached the target node and whether it has entered the wrong path. After confirming the target position of the manipulator and the deflection angle of the magnetic navigation target, the manipulator and the permanent magnet begin to move toward the target position, and the guidewire begins to move forward at the same time. In this process, the camera continuously obtains a photo stream, and when it detects that the length of vector A is less than the minimum reach distance (flexibly set according to demand, generally 1mm), it is considered that the guidewire has reached the target node. At this time, it is judged whether the manipulator has reached the target position. If it has not yet arrived, the guidewire waits in place until the manipulator reaches the corresponding position. After the robot arrives, the guide wire continues to move forward. At this time, the camera continues to acquire a stream of photos. After the length of vector A is greater than the minimum judgment distance (flexibly set according to needs, generally 2mm), if the angle between vector A and vector B is less than the allowable error range (flexibly set according to needs, generally 15°), it is considered that the guide wire has entered the correct path. Otherwise, it is considered that the guide wire has entered the wrong path. At this time, the guide wire needs to be retracted until vector A is less than the minimum arrival distance. After the computer analyzes the positional relationship between the wrong path and the correct path, the permanent magnet is deflected in the direction of the principle wrong path by one rotation unit (generally 1°), and the guide wire is re-controlled to move forward until the guide wire enters the correct path and completes a cycle. The program ends when the guide wire reaches the end point.

[0110] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0111] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A hard magnetic driven vascular intervention guidewire robot system, characterized in that: include: Hard magnetic flexible guidewire catheter: the flexible catheter (1) is sleeved on the outside of the hard magnetic flexible guidewire (2), the tip of the hard magnetic flexible guidewire (2) has hard magnetism and can be extended or retracted in the catheter; The guidewire catheter driving device (9) is used to clamp the hard magnetic flexible guidewire catheter and provide driving force for the hard magnetic flexible guidewire catheter to move forward and backward; A seventh degree of freedom magnetic navigation system (17): arranged above the guidewire catheter drive device (9), comprising a mechanical arm (18), a servo rotary motor (16) and a cylindrical permanent magnet (20), and controlling the movement direction of the hard magnetic flexible guidewire (2) by changing the relative position between the cylindrical permanent magnet (20) and the tip of the hard magnetic flexible guidewire (2); A visual recognition module is arranged between the guidewire catheter driving device (9) and the seventh degree of freedom magnetic navigation system (17), and includes a support (12) and an industrial camera (14). The position of the hard magnetic flexible guidewire (2) in the blood vessel is returned through the industrial camera for subsequent path analysis.

2. The hard magnetic driven vascular intervention guidewire robot system according to claim 1, characterized in that: In the hard magnetic flexible guidewire catheter: The hard magnetic flexible guide wire (2) is composed of a titanium alloy inner core and a braided structure, and is coated with polytetrafluoroethylene material on the outside; the guide wire tip sheath is filled with axially magnetized neodymium iron boron magnetic particles; The flexible catheter (1) is made of a thermoplastic resin material and is sleeved on the outside of the hard magnetic flexible guide wire.

3. The hard magnetic driven vascular intervention guidewire robot system according to claim 1, characterized in that: In the guidewire catheter driving device (9): There are two groups of drive modules, which respectively clamp the hard magnetic flexible guide wire (2) and the flexible catheter (1) to drive the hard magnetic flexible guide wire catheter to move forward and backward in the blood vessel; each drive module is composed of a pair of friction wheels, a servo motor (3), a worm gear (10), a power supply and a controller (8); the end of the servo motor (3) shaft is connected to a worm gear meshing with the worm gear, and the worm gear is coaxially connected to the driving wheel of the friction wheel. The flexible catheter (1) is clamped between the driving wheel (4) and the driven wheel (5) of the friction wheel, and the servo motor drives the worm gear (10) to rotate, thereby driving the driving wheel (4) and the friction wheel (5) to rotate. The flexible guide wire (2) is clamped in the wire feeding mechanism (6), and the servo motor drives the worm gear (10) to rotate, thereby driving the driving wheel (4) and the friction wheel (5) to rotate. The code is transmitted through the control board, thereby controlling the advancement and retreat of the hard magnetic flexible guide wire (2) or the flexible catheter (1). Positioning shell (9): two sets of driving modules are embedded in the bottom of the positioning shell, and the controller and power supply are fixed to the shell around the positioning shell (9).

4. The hard magnetic driven vascular intervention guidewire robot system according to claim 1, characterized in that: In the seventh degree of freedom magnetic navigation system (17): 6-DOF robotic arm (18); A servo motor (16) is arranged at the end of the 6-DOF mechanical arm (18); The upper connecting seat (21) is arranged between the 6-DOF mechanical arm (18) and the servo motor (16); The lower connecting seat (22) is arranged at the lower part of the upper connecting seat (21); The special-shaped central axis (23) is arranged in the central hole of the lower connecting seat (22); The top cover (24) is arranged outside the special-shaped central axis (23); A connecting plate (25) is arranged at the end of the special-shaped central axis (23); The cylindrical box (29) is arranged below the connecting plate (25); The cylindrical permanent magnet (20) is arranged inside the cylindrical box (29).

5. The hard magnetic driven vascular intervention guidewire robot system according to claim 1, characterized in that: In the visual recognition module: It includes: a base (13); A lower column (26) is arranged on the base (13); The upper column (27) is arranged on the lower column (26); The positioning clamp (28) is arranged outside the upper column (27) or the lower column (26); The industrial camera (14) is arranged in a positioning fixture (28).

6. A hard magnetic driven vascular intervention guidewire robot method, characterized in that: The hard magnetic driven vascular intervention guidewire robot system according to any one of claims 1 to 5 is used to perform the following steps: Step S1: color marking the tip of the guidewire to obtain the original two-dimensional image required for vascular intervention; Step S2: convert the collected original image into the HSV color space, generate a binary mask of the image, extract the target color area contour, and extract the center coordinates of the minimum circumscribed rectangle; Step S3: Predicting the position of the obscured guide wire tip based on a convolutional neural network and constructing a model; Step S4: Collect images containing the features of the hard magnetic flexible guide wire tip and train them according to the model.

7. The hard magnetic driven vascular intervention guidewire robot method according to claim 6, characterized in that: In step S2: The collected original image is converted from BGR color space to HSV color space, and a specific HSV threshold is set to separate different color ranges; the color mark set at the tip of the guide wire is used as the target, and a binary mask of the image is generated according to the set domain value. The white part of the mask represents pixels that meet the conditions, and the black part represents pixels that do not meet the conditions; the mask is expanded into the same shape as the original image, applied to the original image channel by channel, the outline of the target color area is extracted, the minimum enclosing rectangle is fitted to the outline, and the center coordinates of its minimum enclosing rectangle are extracted.

8. The hard magnetic driven vascular intervention guidewire robot method according to claim 6, characterized in that: In step S3: The pose prediction of the occluded guide wire tip is performed based on the convolutional neural network. The model is constructed by including multiple convolutional layers, maximum pooling layers, activation layers and fully connected layers. The image features are extracted through the convolutional layer, and the convolutional layer output is nonlinearly transformed using the ReLU activation function. The data dimension is reduced through the maximum pooling layer, and the image data is batch normalized. The flattening layer is connected to the fully connected layer, and the high-level feature map is mapped to the output space. The final result is generated through the output layer.

9. The hard magnetic driven vascular intervention guidewire robot method according to claim 6, characterized in that: In step S4: Model training: collect pictures containing the tip features of hard magnetic flexible guidewires, normalize and enhance them to obtain a learning data set, identify the tip position of each picture and obtain the tip center coordinates as feature labels; 30% of the images in the dataset were extracted, and the guidewire tip was partially color-blocked to train the position recognition ability under occlusion. The dataset was divided into a training set and a validation set. The mean square error was used as the loss function, and the model parameters were optimized by the Adam optimizer. The MSE loss function calculation formula is as follows: Among them, y i are the actual coordinates, are the prediction coordinates and N is the number of samples.

Images