Computer-assisted positioning method for dead center of medial and lateral collateral ligaments

Abstract

The invention discloses a computer-aided positioning method for dead center of medial and lateral collateral ligaments. The method comprises the following steps: firstly, acquiring a CT (Computed Tomography) image and an MRI (Magnetic Resonance Imaging) image, preprocessing the CT image to segment a bone region, segmenting the CT image into independent bone tissue regions by maximum flow and minimum cutting, marking and training the MRI image, and acquiring bone tissue and ligament parts; and performing three-dimensional reconstruction and voxel registration on the CT image segmentation result and the MRI image segmentation result to obtain a bone tissue three-dimensional model. According to medical anatomy knowledge, pose adjustment of the thighbone model and the tibia model is achieved through computer assistance, and the dead center of the medial collateral ligament and the dead center of the lateral collateral ligament are positioned according to medical anatomy knowledge. A knee joint motion trail is simulated based on rigid body dynamics and forward kinematics and is applied to a three-dimensional model, an isometric test is carried out, and the accuracy and rationality of the dead center are verified.

Description

Technical Field

[0001] The present invention relates to the field of medical image technology, and in particular to a computer-assisted method for locating the centers of the insertion points of the medial and lateral collateral ligaments. Background Art

[0002] The collateral ligaments are important stabilizing structures of the knee joint, primarily composed of the medial and lateral collateral ligaments. The medial collateral ligament is the primary resistance to knee valgus, while the lateral collateral ligament effectively resists varus stress. However, there is currently no effective computer-assisted method for locating the insertion center of the collateral ligament. This method relies primarily on the surgeon observing the structure of the knee joint through an arthroscopic lens during surgery to determine the insertion center of the collateral ligament. However, due to the limited field of view of the arthroscopic lens, the relatively small surgical operating space, and the extremely high level of experience and technical skills required of the surgeon, damage to the knee cartilage may occur, increasing surgical risks and complications, and in severe cases, even leading to disability. Summary of the Invention

[0003] In response to the problems existing in the prior art, the present invention discloses a computer-assisted method for locating the insertion center of the medial and lateral collateral ligaments. The method is based on authoritative medical literature and focuses on computer-assisted positioning of the insertion center. The method specifically includes the following steps:

[0004] Obtain CT and MRI images of the medial and lateral collateral ligaments of the human knee joint, pre-process the CT images to segment the bone region; perform deep learning training on the MRI images to segment the medial and lateral collateral ligament regions;

[0005] After threshold segmentation of the CT image, a preliminary bone tissue region was obtained. The maximum flow minimum cut algorithm was used to segment the adhesions between the bone tissues. The CT image was then segmented into the patella, femur, tibia, and fibula regions. VTK-based surface rendering technology was used for reconstruction to obtain a 3D mesh model of the femur, tibia, patella, and fibula of the knee joint.

[0006] The bone and ligament areas in the segmented MRI images were reconstructed using VTK-based surface rendering technology to obtain a three-dimensional mesh model of the coronal bone tissue, medial collateral ligament, and lateral collateral ligament.

[0007] The tilt angle is calculated based on the anterior femoral cortex, the point set information of the anterior femoral cortex is obtained, and the least squares method is used to fit a straight line to calculate the tilt angle and make adjustments;

[0008] Based on the femoral condyle angle, the overlap ratio of the medial and lateral femoral condyles is calculated to maximize overlap and achieve the clinically appropriate lateral position. The tibia is similarly adjusted using the tibial plateau as described above.

[0009] A 10*10 grid is drawn based on the surface of the femoral condyle, and the center of the femoral medial collateral ligament insertion point is located according to anatomical knowledge; similarly, a 10*10 grid is drawn based on the side of the tibia, and the center of the tibial medial collateral ligament insertion point is located according to anatomical knowledge.

[0010] The bone tissue reconstructed based on MRI and CT was registered using the Ransac algorithm and the ICP algorithm. The rotation matrix was applied to the lateral collateral ligament. The intersection area between the lateral collateral ligament and the bone tissue was obtained and regarded as the footprint area. The center point of the footprint area was selected, and the nearest point on the bone model was found according to the KdTree method as the center of the insertion point. The center of the lateral collateral ligament on the fibula was rotated with the tibia to draw a 10*10 grid, and the center of the insertion point was located according to anatomical knowledge.

[0011] The knee joint motion trajectory was simulated based on rigid body dynamics and forward kinematics, and applied to a three-dimensional model for isometric testing to verify the accuracy and rationality of the insertion center.

[0012] Furthermore, when segmenting the knee joint CT image: first, the image is exported as a JPG format image, the image is segmented into the left and right legs and a threshold segmentation operation is performed to obtain the preliminary segmentation results of the patella, femur, tibia and fibula.

[0013] Furthermore, the maximum flow minimum cut algorithm was applied to the adhesion locations of the femur and patella, fibula and tibia in the preliminary segmentation image, and they were cut to obtain independent segmentation areas of the patella, femur, tibia and fibula. The 3D mesh models of the femur, tibia, patella and fibula of the knee joint were reconstructed using VTK's surface rendering technology.

[0014] Furthermore, after labeling the bone tissue and medial and lateral collateral ligaments in the MRI images using the annotation software Labelme, deep learning training using U-Net was performed to obtain the bone tissue and medial and lateral collateral ligament regions in the MRI images. Using VTK's surface rendering technology, a 3D mesh model of the femur, tibia, patella, and fibula of the knee joint was reconstructed.

[0015] Furthermore, the current position information of the femoral model reconstructed from the CT image is exported into a two-dimensional image, and the anterior cortical point set of the femur is collected and fitted into a straight line using the least squares method. The offset angle k1 can be obtained and applied to the femoral model to obtain a vertical femur. Then, a greedy algorithm is used to rotate the angles in different directions according to the current overlap of the two femoral condyles, and the overlap rate is calculated, continuously approaching 100%, ultimately making the two femoral condyles overlap, and finally achieving a femur that meets the position requirements.

[0016] Furthermore, the current position information of the tibial model reconstructed from the CT image is exported into a two-dimensional image, and the tibial plateau point set is collected and fitted into a straight line using the least squares method. The offset angle k2 can be obtained and applied to the tibia and fibula models to obtain a horizontal tibia. Then, a greedy algorithm is used to rotate the angles in different directions according to the current overlap of the two tibial plateaus, and the overlap rate is calculated, continuously approaching 100%, and finally the two tibial plateaus are overlapped, and finally the tibia and fibula that meet the position requirements are achieved.

[0017] Next, the FPFH algorithm is used to calculate feature descriptors for the point clouds, providing a basis for subsequent matching. A RANSAC algorithm is then used for coarse registration to calculate the transformation matrix, finding the optimal transformation matrix to align the two knee joint models during the coarse registration phase. The ICP algorithm is then used for fine registration, using the iterative closest point method to gradually optimize the transformation matrix, ensuring precise alignment of the two point clouds in detail.

[0018] Furthermore, the method for obtaining the center position of the medial collateral ligament of the femur is as follows: after adjusting the position of the femur according to the above method, the anterior cortex of the femur to the lowest position below the femoral condyle is used as the Y-axis, and the length is recorded. The length of the left side of the femoral condyle is added as the X-axis, and a 10*10 grid is drawn. According to the literature, the center position of the medial collateral ligament of the femur is located at 61% of the X-axis and 42% of the Y-axis of the grid.

[0019] Furthermore, the method for obtaining the center position of the medial collateral ligament insertion point of the tibia is as follows: after adjusting the position of the tibia according to the above method, the position from the left to the right side of the tibia is used as the X-axis, and the length is recorded. The tibial plateau plus the length is used as the Y-axis, and a 10*10 grid is drawn. According to the literature, the center position of the medial collateral ligament insertion point of the tibia is located at 43.1% of the X-axis and 92.9% of the Y-axis of the grid.

[0020] Furthermore, the method for obtaining the center position of the insertion point of the lateral collateral ligament of the femur is as follows: the lateral collateral ligament is rotated to the position of the CT bone model after applying a rotation matrix, the intersection with the surface of the CT bone model is obtained, all the points of the intersection are traversed, the center point position is calculated, and the nearest point on the surface of the bone model is calculated using kdTree as the insertion point center.

[0021] Furthermore, the method for obtaining the center position of the insertion point of the lateral collateral ligament of the fibula is as follows: after adjusting the position of the fibula following the tibia according to the above method, the distance from the left to the right side of the fibula head is taken as the X-axis, and the length is recorded. The length of the upper side of the fibula head plus the length is taken as the Y-axis, and a 10*10 grid is drawn. According to the literature, the center position of the insertion point of the lateral collateral ligament of the fibula is located at 52.4% of the X-axis and 37.2% of the Y-axis of the grid.

[0022] Furthermore, the method for obtaining simulated motion trajectory points of the collateral ligament insertion center and performing isometric testing involves calculating the simulated motion trajectory points based on rigid body dynamics and forward kinematics. Once the simulated motion trajectory points are obtained, they are applied to a three-dimensional bone model, which is then moved along with the model. The lengths of the insertion center at different angles can be determined, and isometric tests can be performed to verify the accuracy of the positioning.

[0023] Due to the adoption of the above technical solution, the present invention provides a computer-assisted method for locating the centers of the insertion points of the medial and lateral collateral ligaments. During the implementation of this method, the user only needs to provide CT and MRI images of the knee joint to obtain the required centers of the insertion points of the medial and lateral collateral ligaments, thereby reducing the operation of the physician manually searching for the centers of the insertion points during clinical surgery, and at the same time reducing some errors caused by human errors, thereby improving accuracy and efficiency. In addition, this method is easy to run, has a low program memory occupation, and is simple to operate, providing convenience for physicians. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 Flowchart for realizing the method of the present invention

[0026] Figure 2 CT image segmentation effect diagram in the method of the present invention

[0027] Figure 3 This is the effect diagram of the bone model reconstructed from CT images in the method of the present invention

[0028] Figure 4 This is the effect diagram of the bone model reconstructed from MRI images in the method of the present invention

[0029] Figure 5 This is the effect diagram of the femoral anterior cortex point set in the method of the present invention

[0030] Figure 6 The effect diagram of the femur after the posture adjustment in the method of the present invention

[0031] Figure 7 The tibial model effect diagram in the method of the present invention

[0032] Figure 8 This is the tibial plateau point collection effect diagram in the method of the present invention

[0033] Figure 9The effect diagram of adjusting the tibial inclination angle in the method of the present invention

[0034] Figure 10 The effect diagram of the tibia after adjusting the posture in the method of the present invention

[0035] Figure 11 This is the bone model registration effect diagram in the method of the present invention

[0036] Figure 12 The effect diagram of the insertion center of the medial collateral ligament of the femur in the method of the present invention

[0037] Figure 13 The effect diagram of the insertion center of the medial collateral ligament of the tibia in the method of the present invention

[0038] Figure 14 This is the effect diagram of the insertion center of the lateral collateral ligament of the femur in the method of the present invention

[0039] Figure 15 This is the effect diagram of the center of the lateral collateral ligament of the fibula in the method of the present invention

[0040] Figure 16 This is the isometric simulation effect diagram in the method of the present invention

[0041] Specific implementation

[0042] To make the technical solutions and advantages of the present invention more clear, the technical solutions in the embodiments of the present invention are clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention:

[0043] like Figure 1 The computer-aided method for locating the insertion center of the medial and lateral collateral ligaments is shown. This method is based on relevant medical literature and clinical requirements as theoretical support. By adjusting the knee joint bone model to the medical observation position, a 10*10 grid is drawn according to the paper to locate the insertion center. The insertion center position is located according to the intersection of the ligament and the knee joint bone model. In the specific implementation process, the medical observation position is achieved by adjusting the posture of the femur and tibia, such as Figure 5-10 As shown. Draw a 10*10 femoral grid based on the width of the femoral condyle and anterior cortex to determine the position of the medial collateral ligament on the femur, as shown Figure 12 As shown. Draw a 10*10 tibial grid based on the width of the tibia and determine the position of the medial collateral ligament on the tibia, as shown Figure 13 As shown. The intersection of the lateral collateral ligament and the femoral surface is used as the footprint area, and its center point is located. The kdTree is used to locate the nearest point on the bone surface as the center of the lateral collateral ligament insertion point of the femur, as shown in Figure 14 As shown. Draw a 10*10 fibular grid based on the width of the fibular head and determine the position of the lateral collateral ligament on the fibula, as shown Figure 15According to rigid body dynamics and forward calculation, the human motion trajectory is simulated, the trajectory points are applied to the three-dimensional bone model, and the isometric properties of the insertion center are calculated, as shown in Figure 16 shown.

[0044] The specific steps of the method disclosed in the present invention are as follows:

[0045] S1: Input knee MRI and CT images and export the corresponding JPG images. Perform threshold segmentation on the CT image to obtain preliminary segmentation results for the patella, femur, tibia, and fibula. Annotate the MRI image and train a U-Net model to identify the bone tissue, medial collateral ligament, and lateral collateral ligament regions in the MRI image.

[0046] S2: Perform the maximum flow minimum cut algorithm on the preliminarily processed CT image. The maximum flow minimum cut formula is:

[0047] max(f)=min (S,T) c(S,T)

[0048] Where c(S,T) represents the capacity of the cut, which can be calculated as

[0049]

[0050] Cut the adhesion locations between the patella and femur, and between the fibula and tibia, and obtain the segmentation results of the independent patella, femur, tibia and fibula, as shown in the figure. Figure 2 shown.

[0051] S3: For the segmentation results of CT images and MRI images, the VTK-based surface rendering technology is used to extract the isosurface using vtkContourFilter. This method is set to 170 for easy observation and analysis. The surface rendering method is fast and suitable for real-time interactive operations. By inputting the same pixel distance and layer thickness as when taking the CT image through the SetDataSpacing function, the three-dimensional mesh model of the femur, tibia, patella and fibula of the knee joint reconstructed from the CT image and the three-dimensional mesh model of the femur, tibia, patella, fibula, medial collateral ligament and lateral collateral ligament of the knee joint reconstructed from the MRI image are obtained, as shown in the figure. Figure 3 shown.

[0052] S4: Adjust the posture of the femur reconstructed by CT, export the current two-dimensional image of the femur and obtain the point set of the anterior cortex of the femur, and use the least squares method to fit a straight line y = mx + b. The least squares formula is

[0053]

[0054] Calculate the tilt angle and adjust the femur, e.g. Figure 5As shown. The overlap rate of the femoral condyle is calculated as the initial position. The posture is adjusted according to the greedy algorithm, and the overlap rate after rotation along the X-axis and Y-axis is calculated at different angles. The point cloud of the lateral femoral condyle is recorded as f L , and the point cloud of the medial femoral condyle is recorded as f M The overlap rate OA is calculated as

[0055]

[0056] Select a path with an overlap rate greater than the overlap rate of the current position and greater than other angles as the current path, and iterate continuously. The path selection formula is:

[0057] P next =max(max(max(R3,R4),R2),R1)

[0058] Until the overlap reaches 100%, as shown in Figure 6 The same is true for the tibia. Replace the anterior cortex of the femur with the tibial plateau to collect the point set, and replace the femoral condyle with the tibial plateau to calculate the overlap rate, as shown in Figure 7-10 shown.

[0059] S5: Locate the position of the medial collateral ligament on the femur based on the anatomical results of the paper, e.g. Figure 12 As shown, the specific method is as follows:

[0060] After adjusting the femur according to S4, the femoral position is medically defined in the lateral position, and a two-dimensional image is exported. The height of the anterior cortex of the femur to the lower side of the femoral condyle is used as the Y-axis, and the length is recorded. The left side of the femoral condyle is used as one side of the X-axis, and the length is added as the other side of the X-axis. The grid is drawn into a 10*10 grid. According to the literature, the position at 61% of the X-axis and 42% of the Y-axis of the grid is the center position of the medial collateral ligament insertion point of the femur.

[0061] S6: Locate the position of the medial collateral ligament on the tibia based on the anatomical results of the paper, e.g. Figure 13 As shown, the specific method is as follows:

[0062] After adjusting the tibia according to S4, the lateral position of the tibia was defined by medical standards, and a two-dimensional image was exported. The leftmost to rightmost position of the tibia was set as the X-axis, and the length was recorded. The tibial plateau position plus the length was the Y-axis, and the grid was drawn into a 10*10 grid. According to the literature, the center position of the medial tibial collateral ligament insertion point was located at 43.1% of the grid X-axis and 92.9% of the Y-axis.

[0063] S7: Locate the position of the lateral collateral ligament on the femur based on the junction area between the lateral collateral ligament and the bone tissue, such as Figure 14 As shown, the specific method is as follows:

[0064] First, the point cloud data of the knee joint bone model reconstructed by MRI and CT are preprocessed, including data cleaning and alignment, to remove noise and outliers and ensure that the data are in the same coordinate system. Then, the feature descriptor of the point cloud is calculated using the FPFH algorithm, where FPFH is calculated as

[0065]

[0066] Among them, k represents the query point p q and adjacent point p k The number of point pairs. k Represents the query point p q adjacent point p k The distance between them. The characteristics of SPFH are The feature calculation model is

[0067] α=v·n t

[0068]

[0069] θ=arctan(ω·n t ,u·n t )

[0070] Then, by setting appropriate RANSAC parameters, the RANSAC algorithm is used for coarse registration to find the best transformation matrix to align the two point clouds in the coarse registration stage. After that, the ICP parameters are set and the ICP algorithm is used for fine registration. The transformation matrix RT is gradually optimized by the iterative closest point method. RT is

[0071]

[0072] Align two point clouds precisely in details, such as Figure 11 shown.

[0073] The rotation matrix RT is applied to the lateral collateral ligament model, resulting in a model rotated to the position of the CT bone model. The intersection region with the CT bone model is then determined. This region is then traversed to obtain the coordinates of each point and the center position. After obtaining the center position, a KdTree is constructed based on the 3D coordinates of the bone model points, forming a tree-like structure. Starting from the root node, a recursive downward search is performed in the direction closest to the target point's coordinates to find the nearest leaf node. Distance comparison is used to determine the final closest point, which is then used as the center of the lateral collateral ligament insertion on the femur.

[0074] S8: Locate the position of the lateral collateral ligament on the fibula based on the anatomical results of the paper, e.g. Figure 15As shown in FIG, the specific method is as follows: the fibula model is adjusted along with the tibia in S4, the leftmost boundary to the rightmost boundary of the adjusted fibula head is used as the X-axis, and the distance is recorded. The Y-axis is drawn from the top of the fibula head plus the distance, and the Y-axis is drawn into a 10*10 grid. According to the literature, the center of the lateral collateral ligament insertion point on the fibula is located at 52.4% of the grid X-axis, and the 37.2% position of the Y-axis is the center of the lateral collateral ligament insertion point of the fibula.

[0075] S9: After obtaining the center position of the medial and lateral collateral ligament insertion points, perform an isometric test on the center position of the insertion point to determine the accuracy and rationality of the selection of the center position of the insertion point, such as Figure 16 As shown in the figure, rigid body dynamics and forward calculation are used to simulate the human motion trajectory. The trajectory points are applied to the three-dimensional bone model, and the end point center is also applied to the bone model. The end point center length is calculated for each angular rotation, and the maximum difference in all end point center lengths is calculated. If the error is within 2mm, the end point center is considered to be correctly positioned.

[0076] The present invention discloses a computer-assisted method for locating the insertion centers of the medial and lateral collateral ligaments. This method automatically segments and locates the insertion centers of the knee joint. This method replaces the previous method of manually searching to obtain the insertion centers of the medial and lateral collateral ligaments. This method only requires the user to input CT and MRI knee joint images, reducing the workload for clinical Chinese medicine practitioners and speeding up the process, thereby reducing errors caused by human error, improving accuracy and efficiency, and providing convenience for doctors.

[0077] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A computer-assisted method for locating the insertion centers of the medial and lateral collateral ligaments, comprising: Obtain CT and MRI images of the medial and lateral collateral ligaments of the human knee joint, and perform deep learning training on the MRI images to segment the bone area and the medial and lateral collateral ligament areas; Threshold segmentation was performed on the CT images to obtain preliminary bone tissue regions. The maximum flow minimum cut algorithm was used to segment the adhesions between the bone tissues. The CT images were then segmented into the patella, femur, tibia, and fibula regions. 3D reconstruction was performed using the VTK surface rendering method to obtain 3D mesh models of the femur, tibia, patella, and fibula of the knee joint. The bone region and medial and lateral collateral ligament regions in the segmented MRI images were reconstructed using a VTK-based surface rendering method to obtain a three-dimensional mesh model of the bone tissue, medial collateral ligament, and lateral collateral ligament in the coronal plane. The inclination angle of the femur is calculated based on the anterior femoral cortex, and the information of the anterior femoral cortex point set is obtained. The optimal function matching of the anterior femoral cortex point set is found by minimizing the sum of squared errors to obtain a fitting line. The y-axis inclination angle of the fitting line is calculated and used to adjust the z-axis angle of the femur. The overlap ratio of the medial and lateral femoral condyles was calculated based on the femoral condyle angle to maximize the overlap of the femoral condyles to meet clinical medical standards. The tibia was treated in the same way. Draw a mesh based on the surface of the femoral condyle and locate the center of the femoral medial collateral ligament insertion according to anatomical knowledge. Similarly, draw a mesh based on the side of the tibia and locate the center of the tibial medial collateral ligament insertion according to anatomical knowledge. The Ransac and ICP algorithms were used to perform spatial point cloud registration on the bone tissue models reconstructed from CT and MRI images, obtaining a rotation and translation matrix. This matrix was then applied to the lateral collateral ligament model to obtain a lateral collateral ligament model that matched the CT bone model. The intersection of the lateral collateral ligament model and the CT bone model was considered the footprint area. The center of the footprint area was selected, and the nearest point on the CT bone model was found as the insertion center using the KdTree method. The insertion center of the lateral collateral ligament on the fibula was rotated along the tibia to draw a grid, and the insertion center was located according to anatomical principles. The knee joint motion trajectory was simulated based on rigid body dynamics and forward kinematics, and was applied to a three-dimensional model for isometric testing to verify the accuracy and rationality of the insertion center.

2. The computer-assisted method for locating the insertion centers of the medial and lateral collateral ligaments according to claim 1, wherein: Threshold segmentation is used to obtain the rough outline of bone tissue in the CT image, and the maximum flow minimum cut is used to segment the separate patella, femur, tibia and fibula. The maximum flow minimum cut algorithm is: max(f)=min (S,T) c(S,T) Where c(S,T) represents the capacity of the cut, expressed as Where c(u,v) represents the capacity of each pixel, expressed as Among them, x v and y v They are used to represent the horizontal and vertical coordinates of v, x s and y s They are used to represent the horizontal and vertical coordinates of s, x t and y t Used to represent the horizontal and vertical coordinates of t respectively; The MRI images were segmented using deep learning methods to obtain the patella, femur, tibia, fibula, medial collateral ligament, and lateral collateral ligament regions in the coronal plane. Complete three-dimensional mesh models of the femur and tibia were reconstructed using the VTK surface rendering method.

3. The computer-assisted method for locating the insertion centers of the medial and lateral collateral ligaments according to claim 1, wherein: Adjust the posture of the femur area, export the current two-dimensional image of the femur and obtain the point set of the anterior femoral cortex. Use the least squares method to fit a straight line y = mx + b. The least squares formula is Calculate the tilt angle, adjust the femur, calculate the overlap rate of the femoral condyle, use the greedy algorithm to adjust the posture, rotate along the X-axis and Y-axis, and calculate the overlap rate after different angle adjustments. The point cloud of the lateral femoral condyle is recorded as f L , the point cloud of the medial femoral condyle is recorded as f M , where the overlap rate OA is calculated as Select a path with an overlap rate greater than the overlap rate of the current position and greater than other angles as the current path, and iterate continuously. The path selection formula is P next =max(max(max(R3,R4),R2),R1) Among them, R i (i=1, 2, 3, 4) is a subpath.

4. The computer-assisted method for locating the insertion centers of the medial and lateral collateral ligaments according to claim 1, wherein: Based on the drawn grid, the insertion centers of the medial collateral ligament of the femur and the medial collateral ligament of the tibia, as well as the insertion center of the lateral collateral ligament of the fibula, are located. The MRI bone model and the CT bone model are aligned to obtain a rotation matrix. The rotation matrix is applied to the lateral collateral ligament tissue model to obtain the intersection area with the CT bone model as the footprint area. The center of the footprint area is used as the source point, and KdTree is used to find the nearest point on the bone model as the position of the insertion center of the lateral collateral ligament on the femur.

5. The computer-assisted method for locating the insertion centers of the medial and lateral collateral ligaments according to claim 4, characterized in that: The insertion center of the medial collateral ligament of the femur was obtained as follows: the distance from the anterior cortex of the femur to the lowest position below the femoral condyle was recorded as the Y-axis, the length was added to the left side of the femoral condyle as the X-axis, and a grid was drawn. The position at 61% of the X-axis and 42% of the Y-axis of the grid was the insertion center of the medial collateral ligament of the femur.

Images