Bone density evaluation method for implantation position of acetabular cup prosthesis

Abstract

The invention provides a bone mineral density evaluation method for an acetabular cup prosthesis implantation position. The bone mineral density evaluation method comprises the following steps: placing an acetabular cup prosthesis model at a to-be-implanted position in a CT (Computed Tomography) image; extracting grid data of an outer spherical shell of the acetabular cup prosthesis model; according to the grid data of the outer spherical shell, determining a pixel gray value of a part attached to the outer spherical shell of the acetabular cup prosthesis model in the CT image, and binding the pixel gray value with the grid data of the outer spherical shell; and according to a set pixel gray threshold value and a pixel gray value corresponding to the grid data of the outer spherical shell, quantitatively evaluating the bone density distribution condition of the implantation position of the acetabular cup prosthesis. According to the method, the grid data of the outer spherical shell of the prosthesis model is bound with the pixel gray value of the CT image, so that the bone mineral density distribution condition of the implant position of the prosthesis is quantitatively evaluated, the experience of a doctor is not needed, the bone mineral density distribution of the implant position can be quantitatively analyzed, the reading of multiple layers of CT images is not needed, and the preoperative planning efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly relates to a method for evaluating bone density at the implantation position of an acetabular cup prosthesis. Background Art

[0002] Total Hip Arthroplasty (THA) is an effective means for treating end-stage hip joint diseases. As a core component, the implantation position of the acetabular cup prosthesis directly determines the initial stability of the prosthesis and the long-term bone integration effect. Existing clinical practices have shown that the bonding strength between the acetabular cup prosthesis and the femur highly depends on the bone mineral density (BMD) of the implantation area. If the bone density in the implantation area is insufficient (such as in osteoporosis patients), it is prone to prosthesis micromotion and aseptic loosening, thereby affecting postoperative recovery and the service life of the prosthesis. Therefore, accurately evaluating the bone density distribution in the implantation area of the acetabular cup prosthesis before surgery is crucial for improving prosthesis stability and reducing the incidence of complications.

[0003] As Figure 1 shown, the current mainstream preoperative planning for the acetabular cup relies on the patient's pelvic CT images. Doctors manually adjust the position and angle of the three-dimensional model of the acetabular cup in the multi-planar reconstruction view (MPR) through a medical image processing system (such as Mimics, 3D Slicer, etc.). Figure 1 The blue area in [the figure] is the acetabular cup model; in the axial, coronal, and sagittal views, the HU values of the CT images around the projection area of the acetabular cup are observed layer by layer (i.e., the pixel gray value of the CT image, the higher the gray value, the higher the tissue density) to judge the bone density of the acetabular cup implantation position. This method of evaluating bone density highly depends on doctors' experience, is difficult to perform quantitative analysis, and requires viewing multiple layers of CT images, resulting in low efficiency. Summary of the Invention

[0004] The present invention proposes a method for evaluating bone density at the implantation position of an acetabular cup prosthesis, which solves the problems in the prior art that the evaluation method of judging the bone density of the acetabular cup implantation position by observing HU values highly depends on doctors' experience, is difficult to perform quantitative analysis, and requires viewing multiple layers of CT images, resulting in low efficiency.

[0005] The technical solution of the present invention is realized as follows: The present invention provides a method for evaluating bone density at the implantation position of an acetabular cup prosthesis, including the following steps: Place the acetabular cup prosthesis model at the position to be implanted in the CT image; Extract the mesh data of the outer spherical shell of the acetabular cup prosthesis model; According to the mesh data of the outer spherical shell, determine the pixel gray values at the positions in the CT image that fit the outer spherical shell of the acetabular cup prosthesis model, and bind the pixel gray values to the mesh data of the outer spherical shell; According to the set pixel gray threshold and the pixel gray values corresponding to the mesh data of the outer spherical shell, quantitatively evaluate the bone density distribution at the implant position of the acetabular cup prosthesis.

[0006] Specifically, the method of placing the acetabular cup prosthesis model at the position to be implanted in the CT image is as follows: In the preoperative planning stage, in the MPR viewing mode of the CT image, import and move the acetabular cup prosthesis model to ensure that the rotation center of the acetabular cup prosthesis model is at the acetabular fossa of the CT image regardless of whether it is observed from the coronal view, axial view or sagittal view of the CT image.

[0007] Specifically, the method of extracting the mesh data of the outer spherical shell of the acetabular cup prosthesis model includes the following steps: Calculate the normal vector of each triangular patch of the acetabular cup prosthesis model; Split the triangular patches that are originally connected to each other but the included angle between adjacent normal vectors is greater than the preset value into independent triangular patches to obtain several pieces of mesh data, and each piece of mesh data contains several connected triangular patches; Take the mesh data with the largest occupied space volume among all the mesh data as the mesh data of the outer spherical shell of the acetabular cup prosthesis model.

[0008] Preferably, after extracting the mesh data of the outer spherical shell of the acetabular cup prosthesis model, use a mesh optimization method to process the mesh data of the outer spherical shell to make the distribution of mesh vertices uniform.

[0009] Specifically, the method of binding the pixel gray values to the mesh data of the outer spherical shell is as follows: Denote a certain mesh vertex of the outer spherical shell mesh data as a, traverse and calculate the voxel point a' closest to a among all the voxels in the CT image, obtain the pixel gray value b of the voxel point a', and bind the pixel gray value b to the mesh vertex a; Traverse all the mesh vertices in the outer spherical shell mesh data and bind each mesh vertex to the corresponding pixel gray value.

[0010] Specifically, the method of quantitatively evaluating the bone density distribution at the implant position of the acetabular cup prosthesis is as follows: Set a first pixel gray threshold Y1 and a second pixel gray threshold Y2, and Y1 > Y2; denote the pixel gray value bound to a certain mesh vertex in the mesh data of the outer spherical shell as b; If b > Y1, the implantation site corresponding to the grid vertex is cortical bone; if Y1 ≥ b ≥ Y2, the implantation site corresponding to the grid vertex is cancellous bone; if b < Y2, the implantation site corresponding to the grid vertex is trabecular bone; Traverse all grid vertices in the grid data of the outer spherical shell, and separately count the proportions of cortical bone, cancellous bone, and trabecular bone in the implantation sites corresponding to the grid vertices.

[0011] Preferably, the bone density evaluation method further includes: coloring the outer spherical shell of the acetabular cup prosthesis model according to the bone density distribution at the implantation position of the acetabular cup prosthesis, specifically including the following steps: Color each grid vertex according to the bone density category corresponding to each grid vertex in the grid data of the outer spherical shell; set the grid vertices corresponding to the cortical bone region to green, with an RGB value of [0, 1, 0]; set the grid vertices corresponding to the cancellous bone region to yellow, with an RGB value of [1, 1, 0]; set the grid vertices corresponding to the trabecular bone region to red, with an RGB value of [1, 0, 0]; Color each triangular patch in the grid data of the outer spherical shell according to the RGB value of each grid vertex in the grid data. The RGB value of the triangular patch is obtained by taking the average of the RGB values of the corresponding three grid vertices.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) By extracting the grid data of the outer spherical shell of the acetabular cup prosthesis model, binding the grid data of the outer spherical shell to the pixel gray values of the corresponding part in the CT image, and quantitatively evaluating the bone density distribution at the implantation position of the acetabular cup prosthesis by setting pixel gray value thresholds, the present invention does not rely on the experience of doctors, can quantitatively analyze the bone density distribution at the implantation site, and does not require viewing multiple CT images, improving the preoperative planning efficiency; (2) By converting the bone density quantitative data into a red-yellow-green three-color dynamic heat map and realizing continuous color rendering of the outer spherical shell of the acetabular cup based on vertex RGB interpolation, the present invention enables doctors to locate the bone density weak area in seconds, avoid high-risk mechanical sites in real time during preoperative planning, especially provide an intuitive basis for implant position correction for osteoporosis patients, and significantly reduce the risk of prosthesis loosening caused by insufficient bone support. Description of the Drawings

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0014] Figure 1 It is a CT image for bone density evaluation of the implantation position of an acetabular cup prosthesis in the prior art; Figure 2 It is a flowchart of a method for bone density evaluation of the implantation position of an acetabular cup prosthesis according to the present invention; Figure 3 It is a schematic diagram of placing an acetabular cup prosthesis model at the acetabular fossa in the CT image in an embodiment of the present invention; Figure 4 It is a schematic diagram of an acetabular cup prosthesis model in an embodiment of the present invention; Figure 5 It is a meshed effect diagram of a three-dimensional model of an acetabular cup prosthesis in an embodiment of the present invention; Figure 6 It is a schematic diagram of the normal vector distribution on a certain cross-section of an acetabular cup prosthesis model in an embodiment of the present invention; Figure 7 It is a schematic diagram of the principle of splitting triangular patches where there are mutations in adjacent normal vectors in an embodiment of the present invention; Figure 8 It is a schematic diagram of the outer spherical shell of an acetabular cup prosthesis model in an embodiment of the present invention; Figure 9 It is a meshed effect diagram of the outer spherical shell of an acetabular cup prosthesis model after uniform optimization of grid vertices in an embodiment of the present invention; Figure 10 It is an effect diagram of coloring the outer spherical shell of an acetabular cup prosthesis model according to the bone density distribution in an embodiment of the present invention. Detailed implementation manners

[0015] Next, the technical solutions of the present invention will be clearly and completely described in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0016] Refer to Figure 2 , the present invention provides a method for bone density evaluation of the implantation position of an acetabular cup prosthesis, including the following steps: Step 1, place the acetabular cup prosthesis model at the position to be implanted (acetabular fossa) in the CT image. The specific method is: In the preoperative planning stage, in the MPR viewing mode of the CT image, import and move the acetabular cup prosthesis model to ensure that the rotation center of the acetabular cup prosthesis model is at the acetabular fossa of the CT image regardless of whether it is observed from the coronal view, axial view or sagittal view of the CT image, as Figure 3 shown.

[0017] With the help of the MPR views (axial view, coronal view, sagittal view) of the patient's pelvic CT images or the pelvic reconstruction model, place the acetabular cup prosthesis model at the acetabular fossa of the patient's CT image. In this embodiment, a deep learning method (such as U-Net) can be used to reconstruct the pelvic three-dimensional model based on the patient's CT image (this method is an existing technology). The general process is as follows: Input: The patient's CT sequence in DICOM format (slice thickness ≤ 1 mm); Preprocessing: Normalize the CT image (map the voxel values to [0, 1]) and resample (unify the voxel spacing to 1 mm³); Network training: Use the labeled pelvic mask dataset (such as the public dataset TCIA), the loss function is DiceLoss, and the optimizer is Adam (learning rate 0.001); Output: A binary pelvic segmentation mask.

[0018] Convert the segmentation mask into a three-dimensional mesh model (this model is a closed three-dimensional surface) through the Marching Cubes algorithm, and the mesh resolution is the same as the original CT voxel spacing.

[0019] Step 2, extract the mesh data of the outer spherical shell of the acetabular cup prosthesis model. The specific method is as follows: Not all parts of the acetabular cup need to be in contact with the bone. Its effective bone contact surface is only the outermost hemispherical shell part. In order to accurately calculate the bone contact rate of the acetabular cup, first extract its outermost spherical shell (such as Figure 4 the red part in the figure), and the method is as follows: The essence of the three-dimensional model of the acetabular cup prosthesis is a mesh-like data composed of many triangular patches, such as Figure 5 shown; Calculate the normal vector of each triangular patch of the acetabular cup prosthesis model (obtained by the cross product of the vectors formed by any two sides of the triangle), and draw all the normal vectors in the form of blue arrows, such as Figure 6 shown as the normal vector distribution on a certain cross-section of the acetabular cup prosthesis model. It can be clearly seen that at the outermost spherical shell of the acetabular cup prosthesis model and at the adjacent parts of the inner arc, the transition between adjacent normal vectors is uniform (the included angle is very small), and there are mutations (the included angle is very large) between adjacent normal vectors in other parts.

[0020] Split the triangular patches that are originally connected to each other but the included angle between adjacent normal vectors is greater than the preset value (that is, there is a mutation between adjacent normal vectors. In this embodiment, the preset value is 5°, and it can be flexibly adjusted according to the actual situation in the specific implementation process) into independent triangular patches; such as Figure 7As shown, there are two adjacent and connected triangular patches in the original mesh data A, namely triangular patch i and triangular patch j. If there is a sudden change in the normal vectors of i and j, the original mesh data A is split into two independent mesh data B and C.

[0021] Traverse all the triangular patches of the acetabular cup prosthesis model and perform the above splitting steps to obtain several pieces of mesh data, and each piece of mesh data contains several mutually connected triangular patches; Select the mesh data with the largest occupied space volume among all the mesh data and denote it as shell, which is the mesh data of the outer spherical shell of the acetabular cup prosthesis model, as Figure 8 shown.

[0022] Step 3: Determine the pixel gray value of the part in the CT image that fits the outer spherical shell of the acetabular cup prosthesis model according to the mesh data of the outer spherical shell, and bind the pixel gray value to the mesh data of the outer spherical shell, which specifically includes the following steps: Step 301: Process the mesh data shell using a mesh optimization method (such as Delaunay, Voronoi, etc.) to make the distribution of mesh vertices more uniform, as Figure 9 shown; The mesh optimization method adopted in this embodiment is a prior art, and its general process is as follows: 1) Initialization: Select or generate N points from the original mesh M as the initial point set S = {s_1, s_2,..., s_N}; 2) Calculate the Restricted Voronoi Diagram (RVD): Calculate the ordinary Voronoi diagram of the point set S in three-dimensional space, and divide the entire space into Voronoi cells V_i; Restrict each Voronoi cell V_i to the original surface mesh M, that is, calculate V_i ∩ M, and this intersection R_i is the set of all points on the surface M that are closer to the point s_i (i = 1, 2,..., N) than to any other point in S, obtaining a partition {R_1, R_2,..., R_N} on the surface M, and each region R_i corresponds to the point s_i (i = 1, 2,..., N); 3) Calculate the Centroid on Surface: For each restricted Voronoi region R_i, calculate its centroid c_i on the surface M; 4) Move the point to the centroid (Lloyd Relaxation): Move each point s_i to its corresponding new position c_i; 5) Iteration: Use the moved point set S_new = {s_1_new, s_2_new,..., s_N_new} as the new initial point set; when the iteration converges, the point set S will be very close to the CVT state, and its distribution on the surface M will also be significantly improved.

[0023] Step 302: Denote a grid vertex of the shell after grid homogenization as a. Traverse and calculate the voxel point a' in all voxels (pixels in three-dimensional space) of the CT image that is closest to a, obtain the pixel gray value b of the voxel point a' (i.e., the HU value of the pixel in the CT image), and bind the pixel gray value b to the grid vertex a. Step 303: Traverse all grid vertices in the shell, repeat Step 302, and bind each grid vertex to the corresponding pixel gray value.

[0024] Step 4: According to the set pixel gray value threshold and the pixel gray values corresponding to the grid data of the outer spherical shell, quantitatively evaluate the bone density distribution of the acetabular cup prosthesis implantation position. The specific method is as follows: Set a first pixel gray value threshold Y1 and a second pixel gray value threshold Y2, and Y1 > Y2; denote the pixel gray value bound to a certain grid vertex in the grid data of the outer spherical shell as b. Since the HU value of the CT pixel is positively correlated with the bone density, the above thresholds can be determined with reference to the general HU values of human bones. For example, the HU value of cortical bone is above 700, the range of 100 - 700 HU is cancellous bone, and below 100 HU is porous bone. If b > Y1, the implantation site corresponding to this grid vertex is cortical bone; if Y1 ≥ b ≥ Y2, the implantation site corresponding to this grid vertex is cancellous bone; if b < Y2, the implantation site corresponding to this grid vertex is porous bone. Traverse all grid vertices in the grid data of the outer spherical shell, and respectively count the proportions of the implantation sites corresponding to the grid vertices as cortical bone, cancellous bone, and porous bone; for example, 50% of the grid vertices correspond to the cortical bone area, 40% of the grid vertices correspond to the cancellous bone area, and 10% of the grid vertices correspond to the porous bone area. This percentage is for the reference of surgical planners to further adjust and improve the acetabular cup plan.

[0025] Preferably, to further visually present the bone density at the acetabular cup position, the bone density evaluation method further includes: Step 5: Color the outer spherical shell of the acetabular cup prosthesis model according to the bone density distribution of the acetabular cup prosthesis implantation position. The specific method is as follows: Color each mesh vertex according to the bone density category corresponding to each mesh vertex of the shell; for example: set the mesh vertices corresponding to the cortical bone region to green, with an RGB value of [0, 1, 0]; set the mesh vertices corresponding to the cancellous bone region to yellow, with an RGB value of [1, 1, 0]; set the mesh vertices corresponding to the osteoporotic bone region to red, with an RGB value of [1, 0, 0]. Color each triangular face of the shell according to the RGB value of each mesh vertex of the shell. The RGB value of the triangular face is obtained by taking the average of the RGB values of the corresponding three mesh vertices, thereby realizing the visual presentation of bone density, as Figure 10 shown; doctors can intuitively understand that the bone density in the acetabular cup combination area is mostly osteoporotic bone or cancellous bone, and the effect of performing surgery according to this plan will be poor, and it is necessary to readjust the acetabular cup plan to optimize the bone density distribution at the combination position.

[0026] The bone density evaluation method of the present invention can assist doctors in obtaining the percentage and visual presentation of the bone density distribution at the implantation position in real time when performing the acetabular cup placement plan, so as to efficiently plan the acetabular cup to a more reasonable position.

[0027] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for evaluating bone density at the implantation position of an acetabular cup prosthesis, characterized in that, Including the following steps: Place the acetabular cup prosthesis model at the position to be implanted in the CT image; Extract the mesh data of the outer spherical shell of the acetabular cup prosthesis model; According to the mesh data of the outer spherical shell, determine the pixel gray value of the part in the CT image that fits the outer spherical shell of the acetabular cup prosthesis model, and bind the pixel gray value to the mesh data of the outer spherical shell; According to the set pixel gray threshold and the pixel gray value corresponding to the mesh data of the outer spherical shell, quantitatively evaluate the bone density distribution of the acetabular cup prosthesis implantation position.

2. The method for evaluating bone density at the implantation position of an acetabular cup prosthesis according to claim 1, wherein, The method for placing the acetabular cup prosthesis model at the position to be implanted in the CT image is as follows: In the preoperative planning stage, in the MPR viewing mode of the CT image, import and move the acetabular cup prosthesis model to ensure that the rotation center of the acetabular cup prosthesis model is at the acetabular fossa of the CT image regardless of whether it is observed from the coronal view, axial view or sagittal view of the CT image.

3. The method for evaluating bone density at the implantation position of an acetabular cup prosthesis according to claim 1, wherein The method for extracting the mesh data of the outer spherical shell of the acetabular cup prosthesis model includes the following steps: Calculate the normal vector of each triangular patch of the acetabular cup prosthesis model; Split the triangular patches that are originally connected to each other but the included angle between adjacent normal vectors is greater than the preset value into independent triangular patches to obtain several pieces of mesh data, and each piece of mesh data contains several connected triangular patches; Take the mesh data with the largest occupied space volume among all the mesh data as the mesh data of the outer spherical shell of the acetabular cup prosthesis model.

4. The bone density evaluation method for the implantation position of an acetabular cup prosthesis according to claim 1, characterized in that, After extracting the mesh data of the outer spherical shell of the acetabular cup prosthesis model, use the mesh optimization method to process the mesh data of the outer spherical shell to make the distribution of mesh vertices uniform.

5. The method for evaluating bone density at the implantation position of an acetabular cup prosthesis according to claim 1, wherein, The method for binding the pixel gray value to the mesh data of the outer spherical shell is as follows: Denote a certain mesh vertex of the outer spherical shell mesh data as a, traverse and calculate the voxel point a' closest to a among all voxels in the CT image, obtain the pixel gray value b of the voxel point a', and bind the pixel gray value b to the mesh vertex a; Traverse all mesh vertices in the outer spherical shell mesh data, and bind each mesh vertex to the corresponding pixel gray value.

6. The bone density evaluation method for the implantation position of an acetabular cup prosthesis according to claim 1, characterized in that, The method for quantitatively evaluating the bone density distribution of the acetabular cup prosthesis implantation position is as follows: Set the first pixel gray threshold Y1 and the second pixel gray threshold Y2, and Y1>Y2; denote the pixel gray value bound to a certain mesh vertex in the mesh data of the outer spherical shell as b; If b>Y1, the implantation part corresponding to this mesh vertex is cortical bone; if Y1≥b≥Y2, the implantation part corresponding to this mesh vertex is cancellous bone; If b<Y2, the implantation part corresponding to this mesh vertex is osteoporotic bone; Traverse all mesh vertices in the mesh data of the outer spherical shell, and respectively count the proportions of the implantation parts corresponding to the mesh vertices as cortical bone, cancellous bone and osteoporotic bone.

7. The method for evaluating bone density at the implantation position of an acetabular cup prosthesis according to claim 6, characterized in that, The bone density evaluation method further includes: coloring the outer spherical shell of the acetabular cup prosthesis model according to the bone density distribution of the acetabular cup prosthesis implantation position, specifically including the following steps: Color each grid vertex according to the bone density category corresponding to each grid vertex in the grid data of the outer spherical shell; set the grid vertices corresponding to the cortical bone region to green with an RGB value of [0, 1, 0]; set the grid vertices corresponding to the cancellous bone region to yellow with an RGB value of [1, 1, 0]; set the grid vertices corresponding to the osteoporotic bone region to red with an RGB value of [1, 0, 0]. Color each triangular patch in the grid data of the outer spherical shell according to the RGB value of each grid vertex in the grid data. The RGB value of the triangular patch is obtained by taking the average of the RGB values of the corresponding three grid vertices.

Images