A fracture reduction planning method, device and medium based on three-dimensional modeling

Abstract

The application discloses a fracture reduction planning method and device based on three-dimensional modeling, and a medium, relates to the technical field of fracture reduction, and comprises the following steps: performing an outward expansion operation on a risk structure three-dimensional body and a distal three-dimensional bone piece, obtaining a risk structure protection shell and a moving bone piece movement shell, taking a bone closed configuration as a unlocking starting configuration, calculating a locking release sequence according to the local fitting depth of the locking nodes in a broken end complementary locking diagram, and releasing the locking nodes step by step under the condition that the moving bone piece movement shell does not invade the risk structure protection shell, obtaining a reverse unlocking pose sequence, and after the reverse unlocking pose sequence is inversely sequenced and interpolated to form a target reduction path, performing three-dimensional process envelope checking, and obtaining a fracture reduction planning result. The fracture reduction planning result is generated, and the digital accuracy and the surgical safety of the fracture reduction planning are improved.

Description

Technical Field

[0001] This invention relates to the field of fracture reduction technology, and in particular to a fracture reduction planning method, equipment and medium based on three-dimensional modeling. Background Technology

[0002] With the continuous evolution of computer-aided surgical technology and the deep integration of medical image processing and 3D modeling algorithms, orthopedic fracture reduction planning is undergoing a paradigm shift from traditional "hands-on experience" to "digital and precise navigation". In recent years, 3D reconstruction technology based on CT images has gradually become the standard configuration for preoperative planning. Through surface rendering (such as Marching-Cubes algorithm) or volume rendering technology, bone models with a three-dimensional feel can be generated to assist doctors in simulating fracture fragment separation and reduction, laying a solid digital foundation for the precise treatment of complex fractures (such as intra-articular fractures and comminuted fractures).

[0003] Nevertheless, existing fracture reduction methods still have room for improvement. First, they ignore the complex micro-geometric features of the fracture ends, resulting in insufficient reduction accuracy when facing asymmetrical or irregular fracture surfaces, making it impossible to achieve tight interlocking between bone fragments. Second, they lack effective modeling and avoidance verification of the spatial volume (i.e., motion envelope) occupied by the moving bone fragments in the reduction path, which can easily lead to the planned path being unable to be implemented due to collisions during actual surgery. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a fracture reduction planning method based on three-dimensional modeling to solve the problem of insufficient reduction accuracy and inability to achieve tight occlusion between bone fragments.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a fracture reduction planning method based on three-dimensional modeling, which includes, Continuous medical images covering the fracture ends, bone segments on both sides of the fracture ends, and the area of ​​surrounding structures to be avoided were acquired, and three-dimensional surface reconstruction was performed to obtain proximal three-dimensional bone slices, distal three-dimensional bone slices, and three-dimensional volumes of risk structures. Extract the fracture surface of the proximal and distal three-dimensional bone slices and cut them to form a set of fracture shell slices; Based on the normal, curvature, and adjacency relationship of the shell sheets in the set of shell sheets at the fault ends, a complementary locking diagram of the fault ends is constructed. Based on the complementary interlocking diagram of the fracture ends, the distal three-dimensional bone fragments are subjected to pose transformation to form a bony closed configuration; Perform an expansion operation on the three-dimensional volume of the risk structure and the distal three-dimensional bone fragment to obtain the protective shell of the risk structure and the motion shell of the moving bone fragment. Using the bony closed configuration as the unlocking starting configuration, the locking and unlocking sequence is calculated according to the local interlocking depth of the locking nodes in the complementary locking diagram of the fracture ends. The locking nodes are unlocked step by step under the condition that the moving bone plate shell does not intrude into the protective shell of the risk structure, and the reverse unlocking pose sequence is obtained. After the reverse unlocking pose sequence is interpolated in reverse order to form the target reduction path, a three-dimensional process envelope check is performed to obtain the fracture reduction planning result.

[0007] As a preferred embodiment of the fracture reduction planning method based on three-dimensional modeling described in this invention, the three-dimensional surface reconstruction specifically includes: The acquired continuous medical images are spatially arranged according to the acquisition sequence to form voxel data of the fracture ends and surrounding structures that need to be avoided. Extract continuous voxel regions within the bony grayscale range from the voxel data, and separate the continuous voxel regions into proximal bony voxel regions and distal bony voxel regions according to the fracture gap; Mesh reconstruction and surface smoothing were performed on the outer surface boundaries of the proximal and distal bone voxel regions, respectively, to obtain proximal and distal three-dimensional bone slices. Mark the boundaries of surrounding structures that are adjacent to the fracture ends and need to be avoided during reduction from the voxel data. Connect the boundaries of surrounding structures that need to be avoided along the acquisition sequence. Perform mesh reconstruction and surface smoothing on the connected closed boundaries to obtain the three-dimensional volume of the risk structure.

[0008] As a preferred embodiment of the fracture reduction planning method based on three-dimensional modeling described in this invention, the formation of the fracture end shell sheet set specifically includes: Triangular facets were extracted from the proximal and distal three-dimensional bone fragments, respectively. Open surfaces of the proximal three-dimensional bone fragments facing the distal three-dimensional bone fragments and open surfaces of the distal three-dimensional bone fragments facing the proximal three-dimensional bone fragments were selected along the fracture gap direction to form the proximal fracture end surface and the distal fracture end surface, respectively. The outer edge of the cortex is extracted along the outer contour of the proximal and distal fracture surfaces, and the opening of the medullary cavity is extracted along the inner contour of the proximal and distal fracture surfaces that communicates with the medullary cavity. Calculate the curvature value of each grid vertex on the proximal and distal fracture surfaces, and connect continuous grid edges where the curvature difference between adjacent grid vertices reaches a preset curvature inflection threshold to form a curvature inflection boundary; Using the outer edge of the cortex, the opening of the medullary cavity, and the curvature transition boundary as the dividing boundaries, the surface regions of the proximal and distal fracture ends are divided separately. The resulting surface regions are then associated and recorded according to their respective three-dimensional bone slices and spatial connections to form a set of fracture end shell slices.

[0009] As a preferred embodiment of the fracture reduction planning method based on three-dimensional modeling described in this invention, the construction of the complementary locking diagram of the fracture ends specifically includes: Read the shell slices belonging to the proximal three-dimensional bone slice and the shell slices belonging to the distal three-dimensional bone slice in the shell slice set of the fracture end respectively, calculate the average normal and average curvature of each shell slice, and read the adjacency relationship between each shell slice and the adjacent shell slices. The shell plates of the near-end three-dimensional bone fragment and the shell plates of the far-end three-dimensional bone fragment are paired and compared one by one. When the average normals of the two shell plates correspond in opposite directions, the average curvatures of the two shell plates are in a concave-convex fit relationship, and the adjacent shell plates connected to each of the two shell plates can form a corresponding relationship according to the same connection order, the two shell plates are associated as a locking node. The adjacency relationship between adjacent shell layers is converted into the locking connection edge between corresponding locking nodes, and all locking nodes and locking connection edges are associated and recorded according to the spatial position of the fracture end surface to form a complementary locking diagram of the fracture end.

[0010] As a preferred embodiment of the fracture reduction planning method based on three-dimensional modeling described in this invention, the formation of the bony closed configuration specifically includes: Read all locking nodes and locking connection edges in the complementary locking diagram of the broken ends, and read the shell slices of the proximal and distal three-dimensional bone slices of each locking node; The centroids of the proximal and distal shells are calculated based on the three-dimensional coordinates of each mesh vertex in the shell. The displacement differences of all distal shell centroids toward the corresponding proximal shell centroids are weighted and averaged to obtain the translation amount of the distal three-dimensional bone plate. After calculating the spatial extension direction vectors of the locking connection edge on the proximal and distal 3D bone pieces, the rotation amount of the distal 3D bone piece is calculated. Perform pose transformation on the distal 3D bone fragment based on translation and rotation, and perform triangular mesh penetration verification on the distal 3D bone fragment after pose transformation and the proximal 3D bone fragment. When there is no triangular mesh penetration between the distal 3D bone fragment and the proximal 3D bone fragment after pose transformation, and the proximal 3D bone fragment shell and the distal 3D bone fragment shell associated with each locking node achieve positional correspondence according to the locking connection edge order, the bony closure configuration is obtained.

[0011] As a preferred embodiment of the fracture reduction planning method based on three-dimensional modeling described in this invention, the acquisition of the risk structure protective shell and the moving bone fragment motion shell specifically includes: Triangular facets constituting the three-dimensional external surface of the risk structure and triangular facets constituting the external surface of the distal three-dimensional bone slice were collected respectively. Based on the spatial boundary formed by each triangular facet, spatial sampling points located outside the spatial boundary and whose shortest distance to the spatial boundary is no greater than the corresponding preset outward expansion distance are selected. The selected spatial sampling points are merged with the spatial sampling points inside the triangular facets, and three-dimensional surface reconstruction is performed on the outer boundary of the merged spatial sampling points to obtain the risk structure protective shell and the moving bone plate motion shell.

[0012] As a preferred embodiment of the fracture reduction planning method based on three-dimensional modeling described in this invention, the step of obtaining the reverse unlocking pose sequence specifically includes: Using the bony closure configuration as the unlocking starting configuration, the shell plates and normals associated with each locking node in the complementary locking diagram of the fracture ends are read one by one. The normal direction from the shell plate in the distal three-dimensional bone plate to the corresponding shell plate in the proximal three-dimensional bone plate is calculated to determine the projection distance of the shell plate in the distal three-dimensional bone plate into the boundary range of the corresponding shell plate in the proximal three-dimensional bone plate. The projection distance is used as the local engagement depth of the corresponding locking node, and the locking nodes are arranged in ascending order of local engagement depth to obtain the locking and unlocking order. Read the locking nodes one by one in the locking and unlocking order, and control the far-end three-dimensional bone piece to generate pose changes in the opposite direction of the normal corresponding to the currently read locking node, until the local interlocking depth corresponding to the currently read locking node is reduced to zero. After each pose change, the moving bone plate's outer shell is synchronously transformed to the corresponding position of the distal three-dimensional bone plate, and it is determined whether there is spatial overlap between the moving bone plate's outer shell and the risk structure protection shell. If there is no spatial overlap, the pose corresponding to the distal three-dimensional bone plate is recorded and the next locking node is unlocked until the local interlocking depth corresponding to all locking nodes is reduced to zero, and the reverse unlocking pose sequence is obtained.

[0013] As a preferred embodiment of the fracture reduction planning method based on three-dimensional modeling described in this invention, the step of obtaining the fracture reduction planning result specifically includes: Read the poses of the distal three-dimensional bone fragments in the reverse unlocking pose sequence in the reverse order of the unlocking node, take the pose corresponding to the bony closure configuration as the termination pose of the target reset path, and fill in the intermediate poses of the distal three-dimensional bone fragments between adjacent poses according to the preset translation interval to obtain the target reset path. The moving bone fragments are transformed one by one along the target reset path, and the three-dimensional space occupied by the moving bone fragments in all poses is merged to form a three-dimensional process envelope. The spatial overlap between the three-dimensional process envelope and the risk structure protective shell is determined. When there is no spatial overlap between the three-dimensional process envelope and the risk structure protective shell, the target reduction path, bony closure configuration, reverse unlocking pose sequence and three-dimensional process envelope are associated and recorded to obtain the fracture reduction planning result.

[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the fracture reduction planning method based on three-dimensional modeling as described in the first aspect of the present invention.

[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the fracture reduction planning method based on three-dimensional modeling as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: By constructing a complementary interlocking diagram of fracture ends, precise matching and bony closure of the microscopic geometric features of the fracture ends are achieved using normals, curvature, and adjacency relationships. This effectively solves the problem of insufficient accuracy caused by neglecting microscopic features in previous reduction methods, ensuring tight interlocking of bone fragments. At the same time, an innovative expansion calculation mechanism using a risk structure protective shell and a moving bone fragment shell transforms complex dynamic collision detection into static geometric voxel calculation, significantly reducing computational complexity. Combined with reverse unlocking path planning based on local interlocking depth sorting, a step-by-step unlocking process "from shallow to deep" is simulated, and interference checks are performed in real time to select a unique collision-free safe reduction path, thereby improving the digital accuracy of fracture reduction planning and surgical safety. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a fracture reduction planning method based on 3D modeling. Figure 2 A flowchart for forming an assembly of fractured shell layers; Figure 3 A flowchart for obtaining the bony closure configuration; Figure 4A flowchart for obtaining the locking and unlocking sequence. Detailed Implementation

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0020] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0021] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0022] Reference Figures 1-4 This is one embodiment of the present invention, which provides a fracture reduction planning method based on three-dimensional modeling, including the following steps: S1: Acquire continuous medical images covering the fracture ends, bone segments on both sides of the fracture ends, and the area where surrounding structures to be avoided are located, and perform three-dimensional surface reconstruction to obtain proximal three-dimensional bone slices, distal three-dimensional bone slices, and three-dimensional volumes of risk structures.

[0023] S1.1: It should be noted that a tomographic scan of the bone segment containing the fracture ends is performed using a CT image acquisition device. The scan start and end positions are set along the length of the bone segment containing the fracture ends, so that the scan range covers the fracture ends, the bone segments on both sides of the fracture ends, and the area of ​​the surrounding structures to be avoided. According to the preset slice thickness and preset slice spacing, tomographic images of each layer are continuously acquired from the scan start position to the scan end position, and the integrity of the fracture ends boundary, the bone segments on both sides of the fracture ends, and the boundaries of the surrounding structures to be avoided in each layer of tomographic images is checked. The tomographic images of each layer with complete boundaries are spatially arranged according to the scan position sequence to obtain continuous medical images covering the fracture ends, the bone segments on both sides of the fracture ends, and the area of ​​the surrounding structures to be avoided.

[0024] Furthermore, the layer thickness and interlayer spacing are set according to the anatomical structure of the bone segment where the fracture ends are located and the thickness of the bone cortex. For fractures of long bones of the limbs, the layer thickness is 0.5–1.0 mm and the interlayer spacing is 0.4–1.0 mm; for fractures of small bones or joints, the layer thickness is 0.3–0.6 mm and the interlayer spacing is 0.25–0.5 mm.

[0025] S1.2: It should be noted that the acquisition sequence, number of pixel rows, number of pixel columns, pixel spacing, preset slice thickness, and slice spacing of each slice image in continuous medical imaging are read. Each slice image is arranged sequentially from the scan start position to the scan end position according to the acquisition sequence. The pixel spacing is used as the horizontal and vertical spatial resolution, and the slice spacing is used as the inter-slice spatial resolution. The position of each pixel in each slice image is converted into a three-dimensional spatial position, and the gray value corresponding to each pixel is written into the voxel of the corresponding three-dimensional spatial position. For the three-dimensional spatial positions where no continuous gray-level boundary is formed between adjacent slice images, linear interpolation is performed to fill the gray values ​​of the same horizontal and vertical positions of adjacent layers, so that the fracture end boundary, the bone segment boundary on both sides of the fracture end, and the boundary of the surrounding structure to be avoided are continuously distributed in three-dimensional space, forming voxel data of the fracture end and the surrounding structure to be avoided.

[0026] S1.3: It should be noted that the gray values ​​of each voxel in the voxel data of the fracture ends and surrounding structures to be avoided are read. Voxels with gray values ​​within the bony gray range are marked as bony voxels. The bony voxels are then traversed according to the connection relationships of face adjacency, edge adjacency, and point adjacency in three-dimensional space. The connected bony voxels are merged to form a continuous voxel region. The break points where there is no connection between continuous voxel regions along the length direction of the bone segment where the fracture ends are located are detected and the break points are taken as the fracture gap. Using the fracture gap as the boundary, the continuous voxel region located on the proximal side of the fracture gap and continuously connected to the proximal bone segment is recorded as the proximal bony voxel region, and the continuous voxel region located on the distal side of the fracture gap and continuously connected to the distal bone segment is recorded as the distal bony voxel region.

[0027] Furthermore, the bone grayscale range is set based on the difference between the Heinz units of bone tissue and the Heinz units of surrounding soft tissue in CT images. For example, the bone grayscale range is set to 226 HU to 3071 HU.

[0028] S1.4: It should be noted that the positions of adjacent bone voxels and non-bone voxels in the proximal and distal bone voxel regions are read respectively, and the spatial boundary between bone voxels and non-bone voxels is recorded as the outer surface boundary; the boundary intersection points are extracted according to the adjacent voxel units where the outer surface boundary is located, and the boundary intersection points are connected in each voxel unit to generate triangular patches, forming the initial triangular mesh surfaces of the proximal and distal bone voxel regions respectively; vertex smoothing is performed on the initial triangular mesh surface, the average position of each mesh vertex and its adjacent vertices is calculated one by one, and the vertex is moved by a preset smoothing distance along the direction from the current position to the average position, while keeping the position of the outer surface boundary vertex at the fracture gap unchanged, repeating the iteration until the change in the normal angle between adjacent triangular patches is lower than the preset smoothing termination threshold, thereby obtaining the proximal three-dimensional bone slice and the distal three-dimensional bone slice.

[0029] Voxel data of the fracture ends and surrounding structures to be avoided are read. In each layer of tomographic images, the boundaries of the surrounding structures to be avoided are marked point by point along the gray-scale change edges between the surrounding structures to be avoided and adjacent tissues. The boundary points of the same surrounding structures to be avoided in adjacent layers of tomographic images are connected sequentially according to their spatial correspondence to form the closed spatial boundary of the surrounding structures to be avoided. The boundary intersections are extracted according to the adjacent voxel units where the closed spatial boundary is located, and the boundary intersections are connected in each voxel unit to generate triangular patches, forming the initial triangular mesh surface corresponding to the surrounding structures to be avoided. The same meshing reconstruction and surface smoothing are performed on the initial triangular mesh surface corresponding to the surrounding structures to be avoided to obtain the three-dimensional volume of the risk structure.

[0030] It should also be noted that, in order to ensure the consistency of surface smoothness and accuracy between the three-dimensional bone fragments at the fracture ends and the three-dimensional bodies of the surrounding structures to be avoided in three-dimensional space, and to avoid deviations in collision detection or three-dimensional process envelope calculation due to differences in surface accuracy during subsequent fracture reduction, the preset smoothing distance and preset smoothing termination threshold of the risk structure three-dimensional body are consistent with the preset smoothing distance and preset smoothing termination threshold of the proximal and distal three-dimensional bone fragments.

[0031] Furthermore, the smoothing distance is set based on the average side length of the three-dimensional bone fragment mesh and the surface complexity of the fracture ends to ensure that the smoothing process eliminates noise without changing the fracture gap and the geometric features of the bone fragments. For example, the smoothing distance is set to 0.2 mm.

[0032] The smoothing termination threshold is set based on the local curvature change of the bone fragment surface to ensure that the smoothing process eliminates noise and does not destroy the geometric features of the fracture ends. For example, the smoothing termination threshold is set to 5°.

[0033] S2: Extract the fracture surface of the proximal and distal three-dimensional bone slices and cut them to form a set of fracture shell slices.

[0034] S2.1: It should be noted that the triangular facets of the proximal and distal 3D bone slices are read separately, and the fracture gap between the proximal and distal 3D bone slices is read; based on the boundary position of the fracture gap in 3D space, the average coordinates of each spatial position within the fracture gap are calculated, and the average coordinates are recorded as the fracture gap reference point; the triangular facets in the proximal and distal 3D bone slices are read one by one, the centroid coordinates and normal direction of each triangular facet are calculated, and the fracture gap pointing vector is formed by pointing the centroid coordinates of each triangular facet to the fracture gap reference point; when the angle between the normal direction of the triangular facet and the fracture gap pointing vector is not greater than the preset orientation angle, and the triangular facet... When the shortest distance between the bone fragment and the fracture gap is not greater than the preset adjacent distance of the fracture ends, the corresponding triangular facet is recorded as a candidate triangular facet facing the fracture gap. The candidate triangular facets facing the fracture gap in the proximal three-dimensional bone fragment are connected and merged according to the edge connection relationship to form the proximal fracture end surface. The candidate triangular facets facing the fracture gap in the distal three-dimensional bone fragment are connected and merged according to the edge connection relationship to form the distal fracture end surface. Isolated triangular facets that are not continuously adjacent to the fracture gap in the proximal and distal fracture end surfaces are excluded, so that the proximal and distal fracture end surfaces respectively represent the three-dimensional surfaces of the proximal and distal three-dimensional bone fragments exposed to each other at the fracture ends.

[0035] Furthermore, the orientation angle is set based on the spatial distribution of the open surface of the fracture end and the direction of the fracture gap, and is set to 15°.

[0036] The adjacent distance between fracture ends is set based on the average side length of the fracture end surface mesh and the width of the fracture gap, and is set to 0.2 mm.

[0037] S2.2: It should be noted that the triangular facets and their edges in the proximal and distal fracture surfaces are read separately. The edges of the triangular facets connected to only one triangular facet are recorded as boundary edges, and the boundary edges are connected sequentially according to the continuity of the endpoints to form closed boundary loops. For the proximal and distal fracture surfaces, the connection relationship between each closed boundary loop and the corresponding three-dimensional bone fragment outer surface and internal medullary cavity surface is read separately. The closed boundary loops that are continuously connected to the corresponding three-dimensional bone fragment outer surface and distributed around the outside of the fracture surface are recorded as the outer edge of the cortex, and the closed boundary loops that are continuously connected to the corresponding three-dimensional bone fragment internal medullary cavity surface and distributed around the internal cavity of the fracture surface are recorded as medullary cavity openings. The outer edge of the cortex and the medullary cavity opening corresponding to the proximal fracture surface are associated and recorded with the proximal fracture surface, and the outer edge of the cortex and the medullary cavity opening corresponding to the distal fracture surface are associated and recorded with the distal fracture surface.

[0038] S2.3: It should be noted that the mesh vertices, triangular faces, and the connection relationships between triangular faces in the near-end and far-end surfaces should be read respectively. For each mesh vertex, the adjacent triangular faces with the corresponding mesh vertex as a common endpoint should be read, and the product of the normal angle between the adjacent triangular faces and the length of the common edge should be calculated. The ratio of the product to twice the cumulative area of ​​the adjacent triangular faces should be recorded as the curvature value of the corresponding mesh vertex. The curvature values ​​of two adjacent mesh vertices directly connected by mesh edges should be compared one by one, and the curvature difference between the two adjacent mesh vertices should be calculated. When the curvature difference reaches the preset curvature transition threshold, the mesh edge between the two adjacent mesh vertices should be recorded as the transition mesh edge. The transition mesh edges connected to each other should be connected sequentially according to the endpoint continuity relationship to form the curvature transition boundaries corresponding to the near-end and far-end surfaces.

[0039] Furthermore, the curvature inflection threshold is set based on the distribution of the curvature difference between adjacent mesh vertices in the smooth region and the fracture undulation region on the proximal and distal fracture surfaces. For example, the curvature inflection threshold is set to 15°.

[0040] S2.4: It should be noted that the cortical outer edge, medullary canal opening, and curvature transition boundary corresponding to the proximal and distal fracture surfaces are read separately. The cortical outer edge and medullary canal opening are used as the outer and inner termination boundaries of the fracture surface, respectively, and the curvature transition boundary is used as the dividing boundary inside the fracture surface. The connection relationship between adjacent triangular facets is broken along the curvature transition boundary, and the surface area located between the cortical outer edge and the medullary canal opening where the triangular facets are interconnected is retained. The surface areas corresponding to the proximal and distal fracture surfaces are obtained respectively. For each surface area, the spatial connection relationship between the proximal or distal three-dimensional bone slice, the surface area boundary, and the adjacent surface areas is recorded. All surface areas that have completed the association recording are summarized to form a fracture shell slice set.

[0041] S3: Construct a complementary locking diagram of the fracture ends based on the normal, curvature, and adjacency relationship of the shell sheets in the set of shell sheets at the fracture ends.

[0042] S3.1: It should be noted that the shell slices belonging to the proximal three-dimensional bone slice and the shell slices belonging to the distal three-dimensional bone slice in the shell slice set are read separately. The coordinates of the vertices of the triangular facets in each shell slice are traversed, the normal vectors of all vertices are calculated, and the average value is taken as the average normal vector of the corresponding shell slice. The average value of the vertex curvature values ​​is taken as the average curvature of the corresponding shell slice. At the same time, the relationship between each shell slice and its adjacent shell slices directly connected in space is read to form the adjacency relationship record between shell slices.

[0043] Read the shell slices belonging to the proximal 3D bone slice and the distal 3D bone slice in the shell slice set one by one; pair the shell slices of the proximal 3D bone slice with the shell slices of the distal 3D bone slice; for each pairing, read the average normal of the shell slice corresponding to the proximal 3D bone slice and the average normal of the shell slice corresponding to the distal 3D bone slice, and calculate the angle between the two average normals; when the difference between the angle between the two average normals and 180° is not greater than the preset normal matching angle, it is determined that the average normals of the two shell slices correspond in opposite directions; read the average curvature of the two shell slices in the pairing combination, and when the average curvatures of the two shell slices have opposite signs... Furthermore, when the difference in the absolute value of the average curvature is not greater than the preset curvature matching difference, the average curvature is determined to have a concave-convex matching relationship. According to the arrangement order around the shell sheet boundary, the adjacent shell sheets connected by the two shell sheets are read respectively, and the correspondence of the average normal and the concave-convex matching relationship of the average curvature of the adjacent shell sheets at the same arrangement position are compared. When the adjacent shell sheets at the same arrangement position all satisfy the correspondence, the adjacency relationship of the two shell sheets is determined to be mutually corresponding. The shell sheets of the proximal three-dimensional bone sheet and the shell sheets of the distal three-dimensional bone sheet that simultaneously satisfy the correspondence of the average normal in opposite directions, the concave-convex matching relationship of the average curvature, and the mutual correspondence of the adjacency relationship are associated and recorded to form a locking node.

[0044] Furthermore, the normal mating angle is set based on the allowable surface orientation deviation between the shell plates of the proximal three-dimensional bone fragment and the shell plates of the distal three-dimensional bone fragment in the concave-convex mating state at the fracture ends. For example, the normal mating angle is set to 10°.

[0045] The curvature fit difference is set based on the allowable local curvature difference of the shell plates in the concave-convex complementary state, in order to filter out shell plates whose curved shapes cannot fit each other, for example, a curvature fit difference of 0.2.

[0046] S3.2: It should be noted that the shell plates of the proximal and distal three-dimensional bone fragments associated with each locking node are read one by one; for any two locking nodes, when there is an adjacency relationship between the shell plates of the proximal three-dimensional bone fragments associated with each locking node, and there is an adjacency relationship between the shell plates of the distal three-dimensional bone fragments associated with each locking node, a locking connection edge is established between the two locking nodes; according to the spatial arrangement of the shell plates of the proximal three-dimensional bone fragments associated with each locking node on the surface of the proximal fracture end, and the spatial arrangement of the shell plates of the distal three-dimensional bone fragments associated with each locking node on the surface of the distal fracture end, the positional correspondence of each locking node is recorded; according to the locking connection edge between the locking nodes, the connection order of each locking node is recorded, and all locking nodes and locking connection edges that have completed the positional correspondence and connection order recording are summarized to form a fracture end complementary locking diagram.

[0047] S4: Perform pose transformation on the distal three-dimensional bone fragments according to the complementary locking diagram of the fracture ends to form a bony closed configuration.

[0048] S4.1: It should be noted that, for each locking node, all mesh vertices in the shell slices corresponding to the proximal 3D bone slice and the shell slices corresponding to the distal 3D bone slice are read one by one, and the 3D coordinates of all mesh vertices are extracted respectively; for each locking node, the 3D coordinates of all mesh vertices in the shell slice corresponding to the proximal 3D bone slice are summed and then divided by the number of mesh vertices to obtain the centroid of the proximal shell slice; for the shell slice of the distal 3D bone slice, the 3D coordinates of all mesh vertices in the shell slice are summed and then divided by the number of mesh vertices to obtain the centroid of the distal shell slice; the proximal shell slice corresponding to each locking node is calculated. The three-dimensional coordinate difference between the centroid and the centroid of the distal shell is used to obtain the displacement difference corresponding to each locking node. For each locking node, the surface areas of the shell corresponding to the proximal three-dimensional bone piece and the shell corresponding to the distal three-dimensional bone piece are calculated and summed to obtain the shell area corresponding to each locking node. The quotient of the shell area corresponding to each locking node and the sum of the shell areas corresponding to all locking nodes is used to obtain the displacement difference weight corresponding to each locking node. Based on the displacement difference weight corresponding to each locking node, the displacement differences corresponding to each locking node are weighted and summed to obtain the translation amount of the distal three-dimensional bone piece.

[0049] S4.2: It should be noted that, for each locking connection edge, the proximal shell sheet centroids associated with the two locking nodes at both ends of the locking connection edge are read, and the two proximal shell sheet centroids are connected sequentially according to the connection order of the locking connection edge. The three-dimensional coordinate difference from the previous proximal shell sheet centroid to the next proximal shell sheet centroid is used as the spatial extension direction vector of the locking connection edge on the proximal three-dimensional bone sheet; the distal shell sheet centroids associated with the two locking nodes at both ends of the locking connection edge are read, and the two distal shell sheet centroids are connected sequentially according to the connection order of the locking connection edge. The locking connection edges are connected sequentially, and the difference in three-dimensional coordinates from the centroid of the preceding distal shell sheet to the centroid of the following distal shell sheet is used as the spatial extension direction vector of the locking connection edge on the distal three-dimensional bone sheet. After translating the centroid of the distal shell sheet according to the translation amount of the distal three-dimensional bone sheet, the cost function of the difference in the spatial extension direction vector between the locking connection edge on the distal three-dimensional bone sheet and the proximal three-dimensional bone sheet is constructed, and the least squares method is used to solve for the optimal rotation matrix. The optimal rotation matrix is ​​used as the rotation amount of the distal three-dimensional bone sheet.

[0050] S4.3: It should be noted that the centroids of all the distal shell plates associated with the locking nodes in the distal 3D bone plate are read, the average position of all the centroids of the distal shell plates is calculated, and the average position is used as the pose transformation center. Based on the pose transformation center, spatial rotation is performed on each triangular mesh vertex in the distal 3D bone plate according to the rotation amount, and then spatial translation is performed on each triangular mesh vertex after spatial rotation according to the translation amount to obtain the distal 3D bone plate after pose transformation. The triangular facets in the distal 3D bone plate and the proximal 3D bone plate after pose transformation are read one by one. When all the triangular facets of the distal 3D bone plate do not penetrate the triangular facets of the proximal 3D bone plate, and the proximal 3D bone plate shell plates associated with each locking node correspond one-to-one with the distal 3D bone plate shell plates according to the locking connection edge order, the pose state of the distal 3D bone plate is recorded as a bony closed configuration.

[0051] S5: Perform an expansion operation on the three-dimensional volume of the risk structure and the distal three-dimensional bone fragment to obtain the protective shell of the risk structure and the moving shell of the moving bone fragment.

[0052] The triangular facets constituting the three-dimensional external surface of the risk structure and the three-dimensional external surface of the distal bone fragment are read separately, and the expansion distances of the three-dimensional risk structure and the distal bone fragment are preset respectively. According to the spatial resolution corresponding to the voxel data of the fracture ends and surrounding structures to be avoided, spatial sampling points are generated in the spatial range occupied by the three-dimensional risk structure and the distal bone fragment, as well as outside the spatial range. Based on the spatial boundary formed by the triangular facets constituting the three-dimensional outer surface of the risk structure, spatial sampling points located outside the spatial boundary with a shortest distance from the spatial boundary not exceeding the outer extension distance of the three-dimensional risk structure are selected. The selected spatial sampling points are merged with the spatial sampling points inside the three-dimensional risk structure. Three-dimensional surface reconstruction is performed on the outer boundary formed by the merged spatial sampling points to obtain the protective shell of the risk structure. Based on the spatial boundary formed by the triangular facets constituting the outer surface of the distal three-dimensional bone fragment, spatial sampling points located outside the spatial boundary with a shortest distance from the spatial boundary not exceeding the outer extension distance of the distal three-dimensional bone fragment are selected. The selected spatial sampling points are merged with the spatial sampling points inside the distal three-dimensional bone fragment. Three-dimensional surface reconstruction is performed on the outer boundary formed by the merged spatial sampling points to obtain the moving shell of the moving bone fragment.

[0053] Furthermore, the three-dimensional external expansion distance of the risk structure is set based on the spatial avoidance requirements between the distal three-dimensional bone fragment and the surrounding structures to be avoided during fracture reduction, so that the protective shell of the risk structure covers the protection range outside the surrounding structures to be avoided. For example, the three-dimensional external expansion distance of the risk structure is 2.0 mm.

[0054] The outward expansion distance of the distal three-dimensional bone piece is set based on the verification margin of the space occupied by the distal three-dimensional bone piece during the repositioning process, so that the moving shell of the moving bone piece covers the motion verification range outside the distal three-dimensional bone piece. For example, the outward expansion distance of the distal three-dimensional bone piece is 1.0 mm.

[0055] S6: Using the bony closed configuration as the unlocking starting configuration, calculate the locking release sequence according to the local interlocking depth of the locking nodes in the complementary locking diagram of the fracture ends, and release the locking nodes step by step under the condition that the moving bone plate does not intrude into the protective shell of the risk structure, to obtain the reverse unlocking pose sequence.

[0056] S6.1: It should be noted that, using the bony closure configuration as the unlocking starting configuration, the proximal and distal three-dimensional bone shell plates associated with each locking node in the complementary locking diagram of the fracture ends are read one by one; the mesh vertices in the distal three-dimensional bone shell plate are read one by one, and the adjacent triangular facets with the corresponding mesh vertices as common endpoints are read, the average normal of the adjacent triangular facets is calculated, and the local normal of the corresponding mesh vertex is obtained; the corresponding mesh vertex is spatially projected along the local normal toward the proximal three-dimensional bone shell plate, and the intersection point of the spatial projection and the surface of the proximal three-dimensional bone shell plate is read; when the intersection point is within the boundary range of the proximal three-dimensional bone shell plate, the distance between the corresponding mesh vertex and the intersection point along the local normal is calculated, and the distance is calculated. The calculated distance is recorded as the projected distance of the distal 3D bone shell layer entering the boundary range of the proximal 3D bone shell layer. When the spatial projection does not intersect with the surface of the proximal 3D bone shell layer, or the intersection point is not located within the boundary range of the proximal 3D bone shell layer, the surface point on the surface of the proximal 3D bone shell layer with the shortest distance to the corresponding mesh vertex is read. The direction of the line connecting the corresponding mesh vertex and the shortest distance surface point is taken as the supplementary projection direction, and the distance from the corresponding mesh vertex to the shortest distance surface point along the supplementary projection direction is calculated. The calculated distance is recorded as the projected distance of the distal 3D bone shell layer entering the boundary range of the proximal 3D bone shell layer. All projected distances are associated with and recorded with the corresponding locking nodes.

[0057] Furthermore, the local normal of a mesh vertex is obtained by averaging the normals of the adjacent triangular facets of the corresponding mesh vertex, and the normal of a shell sheet is obtained by averaging the local normals of each mesh vertex within the shell sheet.

[0058] S6.2: It should be noted that, the total projection distance of each locking node associated record is read one by one, the total projection distance of the records associated with the same locking node is summed, and the summation result is divided by the number of the corresponding projection distance to obtain the local interlocking depth of the corresponding locking node; the local interlocking depth of each locking node is compared in turn according to the number of the locking nodes in the complementary locking diagram of the fracture ends, the locking node with the smallest local interlocking depth is arranged first, and the remaining locking nodes are arranged in order of increasing local interlocking depth; when the local interlocking depths of two locking nodes are the same, the surface areas of the distal three-dimensional bone shell plates associated with the two locking nodes are compared, the locking node corresponding to the distal three-dimensional bone shell plate with the smallest surface area is arranged first to obtain the locking release order.

[0059] S6.3: It should be noted that, according to the locking and unlocking order, the locking node that is first in the sequence and has not yet been unlocked is read, and the read locking node is taken as the current locking node; the normal direction of the distal 3D bone shell associated with the current locking node towards the proximal 3D bone shell is read, and the current pose of the distal 3D bone is read; based on the current pose of the distal 3D bone, a translational candidate pose is formed along the opposite direction of the normal direction corresponding to the current locking node, and a rotational candidate pose is formed around the center of the current pose of the distal 3D bone. The translational candidate pose and the rotational candidate pose are combined to form a set of candidate unlocking poses corresponding to the current locking node; the candidate unlocking poses in the set of candidate unlocking poses are read one by one, and the local interlocking depth corresponding to the current locking node is recalculated based on the candidate unlocking poses, and the local interlocking depth corresponding to the lock that has not yet been unlocked is recalculated; when the candidate unlocking pose makes the local interlocking depth corresponding to the current locking node less than the local interlocking depth corresponding to the current locking node under the current pose of the distal 3D bone, and the lock that has not yet been unlocked is... When the local interlocking depth corresponding to the locking node is not greater than the local interlocking depth corresponding to the locking node that has not yet been released under the current pose of the distal 3D bone piece, the moving bone piece's motion shell is synchronously transformed to the position corresponding to the candidate unlocking pose, and it is determined whether the moving bone piece's motion shell and the risk structure protection shell have spatial overlap; when the moving bone piece's motion shell and the risk structure protection shell do not have spatial overlap, the candidate unlocking pose is recorded as a feasible unlocking pose; from the feasible unlocking poses, the feasible unlocking pose with the largest reduction in the local interlocking depth corresponding to the current locking node is selected as the next unlocking pose of the distal 3D bone piece; when the local interlocking depth corresponding to the current locking node under the next unlocking pose of the distal 3D bone piece is greater than zero, a new set of candidate unlocking poses is formed based on the next unlocking pose of the distal 3D bone piece, and the feasible unlocking pose selection is repeated; when the local interlocking depth corresponding to the current locking node under the next unlocking pose of the distal 3D bone piece decreases to zero, the current locking node is marked as released, and the next unlocking pose of the distal 3D bone piece is used as the starting pose when releasing the next locking node.

[0060] S6.4: It should be noted that after the distal 3D bone piece completes one movement according to the reverse pose change in the normal direction of the current locking node, the moving bone piece's outer shell is synchronously adjusted to the current position of the distal 3D bone piece, and the spatial overlap between the moving bone piece's outer shell and the risk structure protection shell is determined one by one; when there is no spatial overlap, the current position and pose of the distal 3D bone piece are recorded as the pose after the current locking node is released, and the next locking node in the locking and unlocking sequence is read. The pose change, synchronous transformation of the moving bone piece's outer shell, and spatial overlap judgment steps are repeated until the local interlocking depth corresponding to all locking nodes is reduced to zero, and a complete reverse unlocking pose sequence is obtained.

[0061] S7: After interpolating the reverse unlocking pose sequence in reverse order to form the target reset path, perform three-dimensional process envelope verification to obtain the fracture reduction planning result.

[0062] S7.1: It should be noted that, according to the reverse order of the unlocking and locking nodes, the poses corresponding to the distal three-dimensional bone pieces in the reverse unlocking pose sequence are read, and the pose corresponding to the bony closure configuration is taken as the termination pose of the target reset path; between adjacent poses, linear interpolation calculation is performed on the distal three-dimensional bone pieces according to the preset translation interval to generate supplementary intermediate poses, and the starting pose, interpolated intermediate pose and termination pose are connected in sequence to obtain the complete target reset path.

[0063] Read each pose corresponding to the distal 3D bone fragment in the target reset path one by one, and read the translation and rotation corresponding to each pose; perform spatial position transformation on the moving bone fragment's motion shell according to the translation and rotation corresponding to each pose to obtain the 3D space occupied by the moving bone fragment's motion shell in each pose; merge the 3D space occupied by the moving bone fragment's motion shell in adjacent poses and the supplemented intermediate poses in the pose arrangement order in the target reset path, so that the spatial range traversed by the distal 3D bone fragment in the target reset path from the initial pose to the final pose is continuously covered; associate and record the outer boundary of the merged 3D space to form a 3D process envelope.

[0064] Furthermore, the translation interval is set based on the average side length of the triangular facet of the distal three-dimensional bone piece, so that the movement distance between adjacent poses in the target reset path does not exceed half of the average side length of the triangular facet of the distal three-dimensional bone piece, for example, the preset translation interval is 0.5mm.

[0065] S7.2: It should be noted that the three-dimensional spatial regions occupied by the three-dimensional process envelope and the three-dimensional spatial regions occupied by the risk structure protective shell are read separately. According to the spatial resolution corresponding to the voxel data of the fracture ends and the surrounding structures to be avoided, the spatial sampling points in the three-dimensional process envelope and the spatial sampling points in the risk structure protective shell are compared one by one to see if they have the same three-dimensional spatial position. When there are no spatial sampling points in the three-dimensional process envelope that have the same three-dimensional spatial position as the risk structure protective shell, it is determined that there is no spatial overlap between the three-dimensional process envelope and the risk structure protective shell. The sequence of changes in the pose of the distal three-dimensional bone fragments corresponding to the target reduction path, the termination pose of the distal three-dimensional bone fragments corresponding to the bony closure configuration, the unlocking process of the locking node corresponding to the reverse unlocking pose sequence, and the spatial occupancy range corresponding to the three-dimensional process envelope are associated and recorded to obtain the fracture reduction planning result.

[0066] This embodiment also provides a computer device applicable to fracture reduction planning methods based on three-dimensional modeling, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the fracture reduction planning method based on three-dimensional modeling as proposed in the above embodiment.

[0067] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0068] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the fracture reduction planning method based on three-dimensional modeling as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0069] In summary, this invention effectively solves the problem of insufficient accuracy caused by neglecting microscopic features in previous reduction methods by constructing a complementary interlocking diagram of fracture ends and using normal, curvature, and adjacency relationships to achieve precise matching and bony closure of the microscopic geometric features of fracture ends. This ensures tight interlocking of bone fragments. At the same time, it innovatively introduces an expansion calculation mechanism of risk structure protective shell and moving bone fragment movement shell, transforming complex dynamic collision detection into static geometric voxel calculation, which greatly reduces computational complexity. Combined with reverse unlocking path planning based on local interlocking depth sorting, it simulates a step-by-step unlocking process "from shallow to deep" and performs interference checks in real time, thereby selecting a unique collision-free safe reduction path, improving the digital accuracy of fracture reduction planning and surgical safety.

[0070] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A fracture reduction planning method based on three-dimensional modeling, characterized in that: include, Continuous medical images covering the fracture ends, bone segments on both sides of the fracture ends, and the area of ​​surrounding structures to be avoided were acquired, and three-dimensional surface reconstruction was performed to obtain proximal three-dimensional bone slices, distal three-dimensional bone slices, and three-dimensional volumes of risk structures. Extract the fracture surface of the proximal and distal three-dimensional bone slices and cut them to form a set of fracture shell slices; Based on the normal, curvature, and adjacency relationship of the shell sheets in the set of shell sheets at the fault ends, a complementary locking diagram of the fault ends is constructed. Based on the complementary interlocking diagram of the fracture ends, the distal three-dimensional bone fragments are subjected to pose transformation to form a bony closed configuration; Perform an expansion operation on the three-dimensional volume of the risk structure and the distal three-dimensional bone fragment to obtain the protective shell of the risk structure and the motion shell of the moving bone fragment. Using the bony closed configuration as the unlocking starting configuration, the locking and unlocking sequence is calculated according to the local interlocking depth of the locking nodes in the complementary locking diagram of the fracture ends. The locking nodes are unlocked step by step under the condition that the moving bone plate shell does not intrude into the protective shell of the risk structure, and the reverse unlocking pose sequence is obtained. After the reverse unlocking pose sequence is interpolated in reverse order to form the target reduction path, a three-dimensional process envelope check is performed to obtain the fracture reduction planning result.

2. The fracture reduction planning method based on three-dimensional modeling according to claim 1, characterized in that: The three-dimensional surface reconstruction is specifically performed as follows: The acquired continuous medical images are spatially arranged according to the acquisition sequence to form voxel data of the fracture ends and surrounding structures that need to be avoided. Extract continuous voxel regions within the bony grayscale range from the voxel data, and separate the continuous voxel regions into proximal bony voxel regions and distal bony voxel regions according to the fracture gap; Mesh reconstruction and surface smoothing were performed on the outer surface boundaries of the proximal and distal bone voxel regions, respectively, to obtain proximal and distal three-dimensional bone slices. Mark the boundaries of surrounding structures that are adjacent to the fracture ends and need to be avoided during reduction from the voxel data. Connect the boundaries of surrounding structures that need to be avoided along the acquisition sequence. Perform mesh reconstruction and surface smoothing on the connected closed boundaries to obtain the three-dimensional volume of the risk structure.

3. The fracture reduction planning method based on three-dimensional modeling according to claim 2, characterized in that: The formation of the fracture end shell sheet assembly specifically includes: Triangular facets were extracted from the proximal and distal three-dimensional bone fragments, respectively. Open surfaces of the proximal three-dimensional bone fragments facing the distal three-dimensional bone fragments and open surfaces of the distal three-dimensional bone fragments facing the proximal three-dimensional bone fragments were selected along the fracture gap direction to form the proximal fracture end surface and the distal fracture end surface, respectively. The outer edge of the cortex is extracted along the outer contour of the proximal and distal fracture surfaces, and the opening of the medullary cavity is extracted along the inner contour of the proximal and distal fracture surfaces that communicates with the medullary cavity. Calculate the curvature value of each grid vertex on the proximal and distal fracture surfaces, and connect continuous grid edges where the curvature difference between adjacent grid vertices reaches a preset curvature inflection threshold to form a curvature inflection boundary; Using the outer edge of the cortex, the opening of the medullary cavity, and the curvature transition boundary as the dividing boundaries, the surface regions of the proximal and distal fracture ends are divided separately. The resulting surface regions are then associated and recorded according to their respective three-dimensional bone slices and spatial connections to form a set of fracture end shell slices.

4. The fracture reduction planning method based on three-dimensional modeling according to claim 3, characterized in that: The construction of the complementary interlocking diagram at the fracture ends is specifically as follows: Read the shell slices belonging to the proximal three-dimensional bone slice and the shell slices belonging to the distal three-dimensional bone slice in the shell slice set of the fracture end respectively, calculate the average normal and average curvature of each shell slice, and read the adjacency relationship between each shell slice and the adjacent shell slices. The shell plates of the near-end three-dimensional bone fragment and the shell plates of the far-end three-dimensional bone fragment are paired and compared one by one. When the average normals of the two shell plates correspond in opposite directions, the average curvatures of the two shell plates are in a concave-convex fit relationship, and the adjacent shell plates connected to each of the two shell plates can form a corresponding relationship according to the same connection order, the two shell plates are associated as a locking node. The adjacency relationship between adjacent shell layers is converted into the locking connection edge between corresponding locking nodes, and all locking nodes and locking connection edges are associated and recorded according to the spatial position of the fracture end surface to form a complementary locking diagram of the fracture end.

5. The fracture reduction planning method based on three-dimensional modeling according to claim 4, characterized in that: The formation of the bony closed structure specifically refers to: Read all locking nodes and locking connection edges in the complementary locking diagram of the broken ends, and read the shell slices of the proximal and distal three-dimensional bone slices of each locking node; The centroids of the proximal and distal shells are calculated based on the three-dimensional coordinates of each mesh vertex in the shell. The displacement differences of all distal shell centroids toward the corresponding proximal shell centroids are weighted and averaged to obtain the translation amount of the distal three-dimensional bone plate. After calculating the spatial extension direction vectors of the locking connection edge on the proximal and distal 3D bone pieces, the rotation amount of the distal 3D bone piece is calculated. Perform pose transformation on the distal 3D bone fragment based on translation and rotation, and perform triangular mesh penetration verification on the distal 3D bone fragment after pose transformation and the proximal 3D bone fragment. When there is no triangular mesh penetration between the distal 3D bone fragment and the proximal 3D bone fragment after pose transformation, and the proximal 3D bone fragment shell and the distal 3D bone fragment shell associated with each locking node achieve positional correspondence according to the locking connection edge order, the bony closure configuration is obtained.

6. The fracture reduction planning method based on three-dimensional modeling according to claim 5, characterized in that: The acquisition of the risk structure protective shell and the movable bone plate movement shell specifically refers to: Triangular facets constituting the three-dimensional external surface of the risk structure and triangular facets constituting the external surface of the distal three-dimensional bone slice were collected respectively. Based on the spatial boundary formed by each triangular facet, spatial sampling points located outside the spatial boundary and whose shortest distance to the spatial boundary is no greater than the corresponding preset outward expansion distance are selected. The selected spatial sampling points are merged with the spatial sampling points inside the triangular facets, and three-dimensional surface reconstruction is performed on the outer boundary of the merged spatial sampling points to obtain the risk structure protective shell and the moving bone plate motion shell.

7. The fracture reduction planning method based on three-dimensional modeling according to claim 6, characterized in that: The specific steps for obtaining the reverse unlock pose sequence are as follows: Using the bony closure configuration as the unlocking starting configuration, the shell plates and normals associated with each locking node in the complementary locking diagram of the fracture ends are read one by one. The normal direction from the shell plate in the distal three-dimensional bone plate to the corresponding shell plate in the proximal three-dimensional bone plate is calculated to determine the projection distance of the shell plate in the distal three-dimensional bone plate into the boundary range of the corresponding shell plate in the proximal three-dimensional bone plate. The projection distance is used as the local engagement depth of the corresponding locking node, and the locking nodes are arranged in ascending order of local engagement depth to obtain the locking and unlocking order. Read the locking nodes one by one in the locking and unlocking order, and control the far-end three-dimensional bone piece to generate pose changes in the opposite direction of the normal corresponding to the currently read locking node, until the local interlocking depth corresponding to the currently read locking node is reduced to zero. After each pose change, the moving bone plate's outer shell is synchronously transformed to the corresponding position of the distal three-dimensional bone plate, and it is determined whether there is spatial overlap between the moving bone plate's outer shell and the risk structure protection shell. If there is no spatial overlap, the pose corresponding to the distal three-dimensional bone plate is recorded and the next locking node is unlocked until the local interlocking depth corresponding to all locking nodes is reduced to zero, and the reverse unlocking pose sequence is obtained.

8. The fracture reduction planning method based on three-dimensional modeling according to claim 7, characterized in that: The specific steps for obtaining the fracture reduction planning results are as follows: Read the poses of the distal three-dimensional bone fragments in the reverse unlocking pose sequence in the reverse order of the unlocking node, take the pose corresponding to the bony closure configuration as the termination pose of the target reset path, and fill in the intermediate poses of the distal three-dimensional bone fragments between adjacent poses according to the preset translation interval to obtain the target reset path. The moving bone fragments are transformed one by one along the target reset path, and the three-dimensional space occupied by the moving bone fragments in all poses is merged to form a three-dimensional process envelope. The spatial overlap between the three-dimensional process envelope and the risk structure protective shell is determined. When there is no spatial overlap between the three-dimensional process envelope and the risk structure protective shell, the target reduction path, bony closure configuration, reverse unlocking pose sequence and three-dimensional process envelope are associated and recorded to obtain the fracture reduction planning result.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the fracture reduction planning method based on three-dimensional modeling as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the fracture reduction planning method based on three-dimensional modeling as described in any one of claims 1 to 8.

Images