CN116236278B - Bone tunnel establishment system - Google Patents

Abstract

The invention discloses a system for establishing a bone tunnel, comprising: a tracker placed on a patient's affected part; a marker installed at a fixed position relative to the tracker; a target placed at a target position of the patient's affected part; a mechanical arm device, Equipped with a terminal tracker; optical tracking equipment to identify the patient's affected area and the tracker on the robotic arm equipment; a host computer to obtain the three-dimensional image of the patient's affected area, identify the position information of the target and the marker in the three-dimensional image, and carry out the operation according to the position information of the target. Bone tunnel planning: The upper computer obtains the transformation relationship between the marker and the tracker, and combines the tracker with the optical tracking device to identify the affected part of the patient to calculate the position information of the planned bone tunnel under the optical tracking device, and move the robotic arm accordingly Planning and execution is in place for bone tunneling. The present invention can accurately establish a bone tunnel, and avoids affecting postoperative rehabilitation effect due to position deviation of the start and end points of the bone tunnel.

Description

A bone tunnel establishment system

technical field

The invention relates to the technical field of image processing, and in particular designs a system for establishing a bone tunnel.

Background technique

Anterior cruciate ligament (ACL) rupture is a common and serious sports injury. Improper treatment will lead to instability of the knee joint, causing a series of sequelae and seriously affecting the function of the knee joint. Therefore, its clinical research and treatment It has always been an important topic in the field of orthopedics and sports trauma, and has been valued by experts and scholars. The treatment of ACL rupture often adopts the method of ligament reconstruction. Knee ligament reconstruction usually involves diagnostic arthroscopy, excision and preparation of ligament grafts, creation of bone tunnels with access to the anatomical attachment points of the ligaments, implantation and fixation of the grafts. key points and difficulties.

Taking the most common ACL reconstruction surgery as an example, theoretically, a straight line can be determined based on the two attachment points of the ACL on the femur and tibia, so it is extremely important to determine the position of the ligament attachment points on the femur and tibia , plays a vital role in the postoperative effect of the operation.

In the current traditional surgery, in the scene where the bone tunnel needs to be established, especially in the reconstruction of the anterior cruciate ligament, combined with the use of arthroscopy, the ligament attachment points of the tibia and femur are determined, and then the tunnel is established using traditional tunnel positioning tools. Established, this method is difficult for inexperienced operators, and the position is often inaccurate, which leads to the unsatisfactory position of the tunnel, which affects the subsequent surgical effect.

Contents of the invention

Purpose of the invention: In view of the above-mentioned deficiencies, the present invention proposes a bone tunnel establishment system. By intuitively placing the target at the position used to plan the bone tunnel, and then registering through the markers, the identification is more accurate and the bone tunnel can be accurately identified. The starting and ending points of the bone tunnel can be precisely established.

Technical solution: a bone tunnel establishment system, comprising:

A tracker, placed on the affected part of the patient, used to characterize the location of the affected part;

a marker mounted in a fixed position relative to the tracker;

a target placed on the affected part of the patient at a position for establishing a bone tunnel;

A robotic arm device with an end tracker on it;

an optical tracking device for identifying a tracker on the affected part of the patient and an end tracker on the robotic arm device;

The upper computer obtains the 3D image of the patient including the patient's affected part and markers, recognizes the position information of the target and the marker in the 3D image, and performs bone tunnel planning according to the position information of the target;

The upper computer obtains the transformation relationship between the marker and the tracker according to the position parameters between the marker and the tracker, and combines the tracker with the optical tracking device to identify the patient's affected area to obtain the coordinate system corresponding to the optical tracking device and the image coordinate system. Transform the relationship, and then calculate the position information of the planned bone tunnel in the corresponding coordinate system of the optical tracking device, and based on this, perform motion planning for the mechanical arm device and execute it in place to establish the bone tunnel.

The marker is a target ball mounted on a bracket having a material density higher than that of bone tissue.

The number of the target balls is at least 3.

The host computer recognizes and obtains the position information of the target ball in the three-dimensional image, specifically:

(1) Identify the brackets of the markers in the 3D image, perform binary morphological opening processing to obtain the connected domains, and calculate the corresponding physical volume and spatial centroid position of each connected domain;

(2) Calculate the connected domain whose physical volume meets the set conditions compared with the design parameters of the marker as the target ball group area, and replace the voxel value of the bracket area in the target ball group area with the volume of the material whose density is smaller than the bracket prime value;

(3) Taking the cross-section of the 3D image as the reference plane and the horizontal direction as the reference direction, select the planes that pass through the center of the reference plane and are perpendicular to the reference plane at different angles to the reference direction as a projection plane, and select the simultaneous vertical The other side of the reference plane and the projection plane is used as another projection plane, and then multiple sets of orthogonal projection planes are obtained, and the projection image of the target ball group area projected onto the corresponding orthogonal projection plane is calculated;

(4) Identify the connected domains whose radii meet the set conditions in the orthogonal projection images of multiple groups of target ball groups as candidate areas, and calculate the center coordinates of each candidate area, that is, each candidate target ball in each group of positive The coordinates of the center of the projection area on the cross-projection image;

(5) According to step (4) and the normal vector calculation of the corresponding projected image, the corresponding two sets of space straight line equations are obtained, and the space intersection point of the two sets of space line equations is calculated to be the space center of the corresponding candidate target ball, and then multiple sets of The space center of the candidate target ball corresponding to the orthogonal projection plane;

(6) Judging whether the number of space sphere centers corresponding to each group obtained in step (5) is consistent with the number of target balls in the designed marker, if so, use the space sphere center data corresponding to step (5) as candidate data; otherwise select The spatial sphere center data that can be extracted in all sets of orthogonal projection surfaces are used as candidate data;

(7) According to the candidate data obtained in step (6), it is judged whether the number M of space centers of the extracted target balls is consistent with the number N of target balls in the designed target ball group;

If so, each target ball in the three-dimensional image is marked according to the topological structure between each target ball obtained by the design parameters of the marker, and the position information of each target ball is obtained according to the space center of each target ball;

If not, arrange and combine the candidate data , according to the results of permutations and combinations, screening is carried out according to the topological structure of each target ball, and the least square method is used to calculate the error between the screened combination and the actual target ball group, and the group with the smallest error is selected as the final target ball combination, and The position information of each target ball is obtained according to the space center of each target ball.

The errors are specifically:

Carry out the least squares method between the actual coordinate system corresponding to the screened combination and the theoretical coordinate system corresponding to the actual marker to calculate the transformation relationship between the two coordinate systems, and calculate the paired target ball point pair between any one of them after transformation and the other. The spatial distance between them, and the root mean square of each paired target ball point pair is taken as the error.

The marker is located at the edge of the acquired three-dimensional image, then in the step (2), the connected domain whose physical volume is compared with the design parameters of the marker and meets the set conditions is used as the target ball group area. Specifically: first select the physical volume phase The connected domain whose design parameters of the markers meet the set conditions is used as the candidate connected domain, and then the space center of mass in the candidate connected domain is selected as the target ball group area which is closest to the edge of the 3D image.

On the basis of the target ball group area obtained in step (2), extend it to the three coordinate axes of the image coordinate system by a set distance to obtain the final target ball group area.

During the expansion process, if the target ball group area exceeds the scope of the image space, the maximum intersection with the image space is taken as the final target ball group area.

The calculated physical volume compared with the connected domain whose design parameters of the marker meet the set conditions is used as the target ball group area specifically: the volume V occupied by the marker is calculated according to the design parameters of the marker, and its physical volume is calculated to satisfy [k ₁ V, k ₂ V] as the target ball group area.

k ₁ =0.8, k ₂ =1.2.

The different angles are respectively 0°, 30°, and 60°.

The connected domain whose radius satisfies the set condition in the orthogonal projection image of multiple groups of target ball group regions obtained from the recognition is specifically as the candidate region:

The obtained orthogonal projection image of the target ball group area is subjected to adaptive threshold segmentation processing to obtain each connected domain, and each connected domain is searched by the Hough circle finding algorithm to obtain the radius satisfying [k ₃ R, k ₄ R] The connected domain is used as a candidate area; where, R is the design radius of the target ball.

k ₃ =0.8, k ₄ =1.2.

The patient's affected part is the patient's affected limb, and the target is placed at the ligament attachment points of the tibia and femur after the arthroscopic probe penetrates into the cruciate ligament of the patient's affected limb through the instrument.

Beneficial effects: the scheme provided by the present invention provides a more intuitive and accurate method for tunnel establishment of femur and tibia in the scene where bone tunnel establishment is required, especially in anterior cruciate ligament reconstruction surgery. By intuitively placing the target at the position used to plan the bone tunnel, and then registering through the markers, the starting and ending points of the bone tunnel can be accurately identified, and the bone tunnel can be accurately established to avoid affecting postoperative rehabilitation due to the position deviation of the starting and ending points of the bone tunnel. Effect. In addition, the present invention obtains the approximate position of the tracer in the image through rough calculation, and then obtains the precise position of the marker in the image through fine calculation, avoiding repeated iterations and faster search for markers in the image more accurate. The small size of the tracer prevents the tracer from being out of the imaging range of the imaging device.

Description of drawings

Fig. 1 is the architecture diagram of the navigation system of the present invention;

FIG. 2 is a schematic diagram of constructing an orthogonal projection surface on a cross-section of a three-dimensional image.

In the figure, 1. perspective device, 2. optical tracking device, 3. mechanical arm device, 4. tibia, 5. femur, 6. tibia tracker, 7. femur tracker, 8. target, 9. tracking ball, 10 .Tracker tooling, 11. Target ball, 12. Sleeve.

Implementation

The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments.

Referring to Fig. 1, the bone tunnel establishment system according to the specific embodiment of the present invention includes an operating bed, a fluoroscopy device 1, an optical tracking device 2, a mechanical arm device 3, a host computer, and a tibia installed on the tibia 4 and the femur 5 of the patient's affected limb respectively. The tracker 6 and the femur tracker 7 and the target 8 placed on the ligament attachment points of the tibia 4 and the femur 5 respectively; wherein, the patient's affected limb is fixed on the operating table; the fluoroscopy device 1 is used to scan the knee joint of the patient's affected limb to obtain fluoroscopy Image; in the present invention, the fluoroscopy device 1 is a CT device; an end tracker is provided on the mechanical arm device 3, and a sleeve 12 is installed at the end of the mechanical arm device 3, and the sleeve 12 is used to perforate the bone tunnel and punch holes Kirschner wire or drill bit; optical tracking device 2 is placed on the position information that can identify tibial tracker 6 and femoral tracker 7, used to identify tibial tracker 6 and femoral tracker 7 on tibia 4 and femur 5, and optical tracking device 2 The tracking device 2 identifies the position information of the end tracker on the mechanical arm device 3; the target 8 can be inserted into the cruciate ligament of the patient's affected limb through the arthroscope at the ligament attachment points of the tibia 4 and femur 5 respectively; The target 8 is made of a material with good imaging effect and no artifacts under the fluoroscopy of the fluoroscopy device 1 , especially a material with good imaging and easy placement on the bone surface.

Wherein, the tibial tracker 6 and the femoral tracker 7 both include several coplanar and non-colinear tracking balls 9, and tracker tooling 10 is installed on the tibial tracker 6 and the femoral tracker 7 as markers, specifically , the tracer tooling 10 includes a bracket and a number of target balls 11 installed on the bracket. The target balls 11 need to have a good imaging effect under CT without artifacts and have a density greater than that of bone tissue. The preferred material is ceramic balls. The density of the target ball is much lower than that of the bracket; the material density of the bracket needs to be higher than that of bone tissue, and the metal material is preferred in the present invention.

In the present invention, the number of target balls 11 is at least 3 and preferably 3, and they are installed on a metal bracket, which is then installed on the tibia tracker 6 or femur tracker 7 .

In the present invention, when the fluoroscopy device 1 scans the patient's knee joint, it needs to cover the tracer tooling 10 installed on the tibial tracker 6 and the femoral tracker 7, and it is necessary to ensure that the tracer tooling 10 is located in the scanned fluoroscopic image. The range of the 3D image obtained by 3D reconstruction is generally at the edge of the 3D image due to the limitation of the imaging range. Further, in the 3D image obtained through 3D reconstruction, the patient's bone tissue is located in the central area of the visual field, and the tracer tooling 10 is located at the edge of the 3D image. In the present invention, the source of the 3D image is not limited to scanning by the fluoroscopy device 1 and then obtained by 3D reconstruction, and can also be obtained by other methods.

The upper computer obtains the image of the knee joint of the patient's affected limb scanned by the fluoroscopy device 1, performs three-dimensional reconstruction on it, and recognizes the position information of the target 8 and the target ball 11, that is, obtains the target 8 and the target ball 11 in the image space coordinate system. According to the position information of the target ball 11, the position information of the target ball 11 in the tooling coordinate system corresponding to the tracker tooling 10 is obtained according to the design parameters of the tracker tooling 10, and then according to the tracer tooling 10 and the tibial tracker 6 or The installation parameters between the femoral trackers 7 are calculated to obtain the transformation relationship between the tooling coordinate system and the corresponding tracker coordinate system, and the position information of the tibial tracker 6 and the femoral tracker 7 is obtained by combining the optical tracking device 2 to calculate the target ball 11 The positional relationship between the optical tracking coordinate systems corresponding to the optical tracking device 2, and then calculate the transformation relationship between the optical tracking coordinate system and the image space coordinate system, so that it can be directly acquired by the optical tracking device 2 and recognized in the image The position information of the target ball 11;

Wherein, the position information of the target 8 in the three-dimensional image is obtained by the upper computer recognition, and the spatial position is extracted and recognized by back-projection calculation after the center of the circle is identified in the 2D image of the orthogonal projection;

The host computer recognizes and obtains the position information of the target ball 11 in the three-dimensional image as follows:

(1) Target ball group area positioning;

In the present invention, the target ball 11 is fixed by a metal bracket. The density of the metal bracket is higher than that of bone tissue, and the density of the material of the target ball 11 is much lower than that of the metal bracket. Groups are differentiated and positioned as follows:

(11) Load the 3D image, and perform threshold segmentation and recognition on it to obtain the metal bracket in the tracer tooling 10, including but not limited to using OTSU (Otsu method, also known as the maximum between-class variance method), threshold iteration and other thresholds Segmentation method; in the present invention, the preferred 2000HU value is used as a threshold;

(12) In view of the adhesion between the metal bracket of the tracer tooling 10 and the target ball 11 or the existence of metal artifacts, etc., the present invention performs binary morphology processing on the three-dimensional image processed in step (11) to remove metal artifacts. shadows or other noise;

(13) Find the connected domains of the binarized image processed in step (12), and calculate the physical volume and spatial centroid position corresponding to each connected domain;

(14) Screen the target connected domain;

The volume V occupied by the tracer tool 10 is calculated according to the design parameters of the tracer tool 10, combined with the physical volume corresponding to each connected domain obtained in step (13), the physical volume satisfies [k ₁ V, k ₂ V ] as the candidate connected domain, and then select the candidate connected domain whose spatial centroid position is closest to the edge of the 3D image as the target connected domain;

In the present invention, k ₁ =0.8, k ₂ =1.2;

(15) Boundary extension;

On the basis of the target connected domain obtained in step (14), extend it to the three coordinate axes of the image space coordinate system for a set distance to obtain the target ball group area. The specific set distance is based on the metal bracket of the tracer tooling 10 It depends on the design size of the target ball 11;

Replace the voxel value of the image area corresponding to the metal bracket in the target ball group area with the voxel value of the minimum density in the effective image area (that is, replace the voxel value corresponding to the metal bracket with the voxel value of the minimum density material, such as voxel value of the air), to obtain the corresponding image, as the image for subsequent use. In the present invention, it is only necessary to replace the voxel value of the image area corresponding to the metal bracket in the area of the target ball group with the voxel value of a material whose density is lower than that of the metal. The voxel value of the image area corresponding to the metal bracket in the ball group area is replaced with the voxel value of the minimum density material.

Among them, if there is beyond the scope of the image space during the expansion process, the maximum intersection with the image space is used as the target ball group area.

(2) Target ball extraction;

(21) Obtain the orthogonal projection image of the target ball group area;

Taking the cross-section of the 3D image as the reference plane and the horizontal direction as the reference direction, select the planes that pass through the center of the reference plane and are perpendicular to the reference plane at different angles to the reference direction as a projection plane, and select the plane perpendicular to the reference plane at the same time. and the other side of the projection surface as another projection surface, and then use the two projection surfaces as orthogonal projection surfaces to obtain multiple sets of orthogonal projection surfaces, and calculate the projection of the target ball group area onto the corresponding orthogonal projection surface Image; as shown in Figure 2, the present invention uses 0°, 30°, and 60° as three angles between the selected and reference directions.

(22) Carry out adaptive threshold segmentation processing on the orthogonal projection image of the target ball group area obtained in step (21) to obtain each connected domain, and retrieve each connected domain through the Hough circle-finding algorithm to obtain the radius satisfying [k ₃ R, k ₄ R] as the candidate area, that is, the projection area of each candidate target ball on the corresponding orthogonal projection image is obtained, and the center coordinates of each candidate area are calculated separately; where R is the target ball 11 the design radius;

In the present invention, k ₃ =0.8, k ₄ =1.2;

(23) Calculating the spatial center of each candidate target ball corresponding to multiple sets of orthogonal projection surfaces;

According to step (22), the center coordinates of the projection areas of each candidate target ball on each set of orthogonal projection images and the normal vectors of the corresponding projection images are calculated to obtain the corresponding two sets of spatial straight line equations, and the space of the two sets of spatial straight line equations is calculated. The intersection point is the space center of the corresponding candidate target ball;

Specifically:

Take a group of orthogonal projection images as an example, such as a projection surface selected at 30° from the reference direction and another surface perpendicular to the reference surface and the projection surface as another projection surface to obtain the orthogonal projection surface. Taking the cross-projection image as an example, the coordinates of the center of a candidate area on a projected image and the normal vector of the projected image are used to calculate the corresponding spatial straight line equation, and the coordinates of the center of the candidate area corresponding to the candidate area on another projected image and The normal vector of another projected image is calculated to obtain another corresponding space straight line equation, and the space intersection point of the two space straight line equations is calculated, which is the space center of the candidate target ball; further, due to the existence of image processing and calculation errors , may result in no intersection point between the two spatial straight lines, then find a straight line that is perpendicular to and intersects the two spatial straight lines as an auxiliary straight line, and calculate the average of the intersection points between the auxiliary straight line and the two spatial straight lines as the spatial center of the candidate target ball ; Further, the space center of the candidate target ball and the distance between the intersection points between the auxiliary straight line and the two space straight lines are calculated separately, and any one of them is greater than the set threshold (such as 1×10 ^-5 mm) When the space center of the candidate target ball is eliminated, return to step (22);

Then use the orthogonal projection surfaces constructed at different angles to the reference direction to calculate and extract the space centers of the candidate target balls corresponding to each group of orthogonal projection surfaces;

(24) Select the space center;

Theoretically, the space centers calculated by each group in step (23) are consistent with the number of target balls in the designed target ball group. At this time, the space center data of any group calculated in step (23) are used as candidate data, and turn to Go to step (25), more specifically, use the average of the space center data of each candidate target ball corresponding to multiple sets of orthogonal projection surfaces calculated in step (23) as the candidate data; The wrong center of the sphere is generated. Therefore, there is a situation where the calculated space sphere is inconsistent with the number of target balls in the designed target ball group. Then select the space sphere that can be extracted from all sets of orthogonal projection planes. The average of the data is used as the candidate data, and the space center that can be extracted in all groups of orthogonal projection surfaces, that is, the space center of a certain candidate target ball can be extracted in all groups of orthogonal projection surfaces;

(25) Extract the target ball, and obtain the position information of each target ball according to the space center of the target ball;

According to the candidate data obtained in step (24), it is judged whether the number M of space centers of the extracted target balls is consistent with the number N of target balls in the designed target ball group;

If so, mark each target ball in the three-dimensional image according to the topological structure between the target balls obtained by the design parameters of the tracer tooling 10, and obtain the position information of each target ball according to the space center of each target ball; wherein , the topological structure between the target spheres generally adopts the length of the connection line between the two target spheres or the angle between the three target spheres;

If not, arrange and combine the candidate data According to the permutation and combination results, screening is carried out according to the topological structure between the target balls, the least square calculation is performed between the screened combination and the actual tracer tooling 10, and the group with the smallest error is selected as the final target ball combination, And obtain the position information of each target ball according to the space center of each target ball; Specifically, perform the least squares method between the actual tool coordinate system corresponding to the combination obtained by screening and the theoretical tool coordinate system corresponding to the actual tracer tool 10 Calculate the transformation relationship of the two coordinate systems, and calculate the spatial distance between any one of the paired point pairs after transformation and the other, and take the root mean square of each paired target ball point pair as the error, and choose the one with the smallest error Each combination serves as the final target ball combination.

The host computer performs bone tunnel planning according to the position information of the target 8 in the image, and obtains the transformation relationship between the target 8 and the target ball 11 according to the position of the target ball 11 in the image, and then combined with the above, it can be obtained directly through the optical tracking device 2. The position information of the target 8 identified in the image can then directly track the planned bone tunnel through the optical tracking device 2;

The robotic arm device 3 performs motion planning and execution in place according to the position information of the end tracker on the robotic arm device 3 obtained by directly tracking the planned bone tunnel through the optical tracking device 2 and the optical tracking device 2, and then completes the bone tunnel through the mechanical arm device 3. Tunnel punching operation.

The present invention takes knee joint replacement as an example, that is, the patient’s affected part in the present invention is the patient’s affected limb, then the starting and ending points of the bone tunnel are the starting and ending points of the ligament reconstruction operation, then the target can be placed on the ligament attachment points of the femur and tibia , but the present invention is not limited thereto, and the present invention is applicable to any other scene where bone tunnel planning operation needs to be performed in bone tissue. Can.

In the scene where the bone tunnel needs to be established, especially in the reconstruction of the anterior cruciate ligament, the present invention inserts targets at the ligament attachment points of the tibia and femur, sets tracers and target balls on the tibia and femur, and uses fluoroscopy equipment to The target is identified in the scanned 3D image, and then the tunnel planning can be performed on the 3D image; then the positional relationship between the target and the target ball is obtained through the target ball identified in the 3D image, and then the tracer is identified by the optical tracking device Combined with the positional relationship between the tracer and the target ball, the position of the target can be directly identified through the optical tracking device, and then the planned bone tunnel can be directly identified through the optical tracking device. Then, through the optical tracking device and the mechanical arm device The tracer on the robot can directly control the movement of the robotic arm equipment to complete the bone tunnel drilling operation. The bone tunnel of the present invention is fixed and tracked by an optical tracking system, which not only determines the attachment point of the ligament before operation, ensures the accuracy of the position of the attachment point of the ligament, but also ensures the accuracy of the position of the bone tunnel during the operation, ensuring the accuracy of the operation. Effect.

The preferred embodiments of the present invention have been described in detail above, but the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations (such as quantity, shape, etc.) can be carried out to the technical solutions of the present invention. , position, etc.), these equivalent transformations all belong to the protection scope of the present invention.

Claims (11)

1. A bone tunnel establishment system, characterized in that, comprising: A tracker, placed on the affected part of the patient, used to characterize the location of the affected part; a marker mounted in a fixed position relative to the tracker; The markers are at least 3 target balls mounted on a bracket with a material density higher than that of bone tissue; a target placed on the affected part of the patient at a position for establishing a bone tunnel; A robotic arm device with an end tracker on it; an optical tracking device for identifying a tracker on the affected part of the patient and an end tracker on the robotic arm device; The upper computer obtains the three-dimensional image of the patient including the patient's affected part and the marker, recognizes the position information of the target and the target ball of the marker in the three-dimensional image, and performs bone tunnel planning according to the position information of the target; The position information of the target ball of the marker in the three-dimensional image obtained by the host computer is specifically: (1) Identify the brackets of the markers in the 3D image, perform binary morphological opening processing to obtain the connected domains, and calculate the corresponding physical volume and spatial centroid position of each connected domain; (2) Calculate the connected domain whose physical volume meets the set conditions compared with the design parameters of the marker as the target ball group area, and replace the voxel value of the bracket area in the target ball group area with the volume of the material whose density is smaller than the bracket prime value; (3) Taking the cross-section of the 3D image as the reference plane and the horizontal direction as the reference direction, select the planes that pass through the center of the reference plane and are perpendicular to the reference plane at different angles to the reference direction as a projection plane, and select the simultaneous vertical The other side of the reference plane and the projection plane is used as another projection plane, and then multiple sets of orthogonal projection planes are obtained, and the projection image of the target ball group area projected onto the corresponding orthogonal projection plane is calculated; (4) Identify the connected domains whose radii meet the set conditions in the orthogonal projection images of multiple groups of target ball groups as candidate areas, and calculate the center coordinates of each candidate area, that is, each candidate target ball in each group of positive The coordinates of the center of the projection area on the cross-projection image; (5) According to step (4) and the normal vector calculation of the corresponding projected image, the corresponding two sets of space straight line equations are obtained, and the space intersection point of the two sets of space line equations is calculated to be the space center of the corresponding candidate target ball, and then multiple sets of The space center of the candidate target ball corresponding to the orthogonal projection plane; (6) Judging whether the number of space sphere centers corresponding to each group obtained in step (5) is consistent with the number of target balls in the designed marker, if so, use the space sphere center data corresponding to step (5) as candidate data; otherwise select The spatial sphere center data that can be extracted in all sets of orthogonal projection surfaces are used as candidate data; (7) According to the candidate data obtained in step (6), it is judged whether the number M of space centers of the extracted target balls is consistent with the number N of target balls in the designed target ball group; If so, each target ball in the three-dimensional image is marked according to the topological structure between each target ball obtained by the design parameters of the marker, and the position information of each target ball is obtained according to the space center of each target ball; If not, arrange and combine the candidate data , according to the results of permutations and combinations, screening is carried out according to the topological structure of each target ball, and the least square method is used to calculate the error between the screened combination and the actual target ball group, and the group with the smallest error is selected as the final target ball combination, and Obtain the position information of each target ball according to the space center of each target ball; The upper computer obtains the transformation relationship between the marker and the tracker of the patient’s affected part according to the position parameters between the marker and the tracker of the patient’s affected part, and combines the optical tracking device to identify the tracker of the patient’s affected part to obtain the corresponding coordinate system of the optical tracking device and The transformation relationship between the image coordinate systems is used to calculate the position information of the planned bone tunnel in the corresponding coordinate system of the optical tracking device, and based on this, the motion planning of the robotic arm equipment is carried out and executed in place to establish the bone tunnel. 2. The bone tunnel establishment system according to claim 1, wherein the error is specifically: Carry out the least squares method between the actual coordinate system corresponding to the screened combination and the theoretical coordinate system corresponding to the actual marker to calculate the transformation relationship between the two coordinate systems, and calculate the paired target ball point pair between any one of them after transformation and the other. The spatial distance between them, and the root mean square of each paired target ball point pair is taken as the error. 3. The bone tunnel establishment system according to claim 1, wherein the marker is located at the edge of the acquired three-dimensional image, and the physical volume calculated in step (2) is compared with the design parameters of the marker to meet the set conditions The connected domain as the target ball group area is specifically as follows: first select the connected domain whose physical volume compared with the marker design parameters meets the set conditions as the candidate connected domain, and then select the candidate connected domain whose spatial centroid position is closest to the edge of the 3D image as the target ball group area. 4. The bone tunnel establishment system according to claim 1, characterized in that, on the basis of the target ball group area obtained in step (2), it is extended to the three coordinate axes of the image coordinate system by a set distance to obtain The final target ball group area. 5. The bone tunnel establishment system according to claim 4, wherein in the expansion process, if the region of the target ball group exceeds the scope of the image space, the maximum intersection of it and the image space is used as the final target ball group area. 6. The bone tunnel establishment system according to claim 1, wherein the calculated physical volume is compared with the connected domain whose design parameters of the marker meet the set conditions as the target ball group area, specifically: according to the design parameters of the marker The volume V occupied by the marker is calculated, and the connected domain whose physical volume satisfies [k ₁ V, k ₂ V] is calculated as the area of the target ball group. 7. The bone tunnel establishment system according to claim 6, characterized in that k ₁ =0.8, k ₂ =1.2. 8. The bone tunnel establishment system according to claim 1, wherein the different angles are 0°, 30°, and 60° respectively. 9. The bone tunnel establishment system according to claim 1, characterized in that, the connected domain whose radius satisfies the set condition in the orthogonal projection images of multiple groups of target ball group regions obtained from the identification is specifically as the candidate region: The obtained orthogonal projection image of the target ball group area is subjected to adaptive threshold segmentation processing to obtain each connected domain, and each connected domain is searched by the Hough circle finding algorithm to obtain the radius satisfying [k ₃ R, k ₄ R] The connected domain is used as a candidate area; where, R is the design radius of the target ball. 10. The bone tunnel establishment system according to claim 9, characterized in that k ₃ =0.8, k ₄ =1.2. 11. The bone tunnel establishment system according to claim 1, characterized in that, the patient's affected part is the patient's affected limb, and the target is inserted into the cruciate ligament of the patient's affected limb by means of an arthroscope in the tibia and femur. The ligament attachment points were placed separately.