Automatic planning method and device for pelvic fracture virtual reduction based on template registration

Abstract

The invention provides an automatic planning method and device for pelvic fracture virtual reduction based on template registration, and the method comprises the steps: carrying out the initial matching of a pelvic fracture block and a mirror image template of a pelvic affected side through a SAC-IA registration method based on the three-dimensional point cloud data of the pelvic fracture block and the three-dimensional point cloud data of the mirror image template of the pelvic affected side; obtaining an initial transformation matrix; updating the three-dimensional point cloud data of the pelvic fracture block based on the initial transformation matrix to obtain updated three-dimensional point cloud data of the pelvic fracture block; based on the updated three-dimensional point cloud data of the pelvic fracture block and the three-dimensional point cloud data of the mirror image template, performing final matching on the pelvic fracture block and the mirror image template by using an ICP registration method to obtain a final transformation matrix; and performing automatic virtual reduction on the three-dimensional model of the pelvic fracture block based on the initial transformation matrix and the final transformation matrix. According to the invention, the reset efficiency and the reset precision are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of pelvic reduction, and in particular to an automatic planning method and device for virtual reduction of pelvic fractures based on template registration. Background Art

[0002] The pelvis is a crucial biomechanical structure that supports and protects many vital organs and innervates the vital blood vessels and nerves that supply the lower extremities. Pelvic fractures, if not treated promptly and accurately, can lead to serious long-term consequences. Preoperative pelvic reduction surgery is a common but technically demanding procedure for pelvic fractures, making precise and rapid pelvic reduction techniques crucial for surgical success.

[0003] Currently, the main preoperative surgical planning technology is manual planning. In manual planning, orthopedic surgeons use professional medical software to reconstruct the pelvis in three dimensions and simulate fracture reduction surgery based on the printed patient-specific 3D model.

[0004] Manual planning takes time, requiring doctors to make repeated attempts and judgments. Furthermore, the position of the pelvic fracture fragments varies between observers and is susceptible to subjective judgment. This results in low reduction efficiency and accuracy, leading to suboptimal reduction results. Summary of the Invention

[0005] In view of this, an object of the present invention is to provide an automatic planning method and device for virtual reduction of pelvic fractures based on template registration, so as to improve the efficiency and accuracy of pelvic reduction.

[0006] In a first aspect, a method for automatically planning virtual reduction of a pelvic fracture based on template registration is provided, the method comprising: Acquire three-dimensional point cloud data of the pelvic fracture fragment and three-dimensional point cloud data of a pre-constructed mirror image template of the affected side of the pelvis; wherein the mirror image template of the affected side of the pelvis is constructed based on the self-symmetry of the pelvic sacrum; Based on the 3D point cloud data of the pelvic fracture fragment and the 3D point cloud data of the mirror image template of the affected pelvis, the SAC-IA registration method is used to perform initial matching between the pelvic fracture fragment and the mirror image template of the affected pelvis to obtain the initial transformation matrix; updating the three-dimensional point cloud data of the pelvic fracture fragment based on the initial transformation matrix to obtain updated three-dimensional point cloud data of the pelvic fracture fragment; Based on the updated 3D point cloud data of the pelvic fracture fragment and the 3D point cloud data of the mirror template, the ICP registration method is used to perform the final matching of the pelvic fracture fragment and the mirror template to obtain the final transformation matrix; The three-dimensional model of the pelvic fracture fragment is automatically virtually reduced based on the initial transformation matrix and the final transformation matrix.

[0007] Optionally, based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis, the SAC-IA registration method is used to perform initial matching on the pelvic fracture fragment and the mirror image template of the affected side of the pelvis, and the initial transformation matrix obtained includes: Step 1: Based on the 3D point cloud data of the pelvic fracture fragment and the 3D point cloud data of the mirror image template of the affected side of the pelvis, determine the corresponding point pairs of the pelvic fracture fragment and the mirror image template of the affected side of the pelvis; Step 2: Randomly select a preset number of point pairs from all corresponding point pairs; Step 3: Determine a first transformation matrix based on a preset number of point pairs; Step 4: updating the three-dimensional point cloud data of the pelvic fracture fragment based on the first transformation matrix to obtain an updated three-dimensional point cloud of the pelvic fracture fragment; Step 5: Match the updated 3D point cloud of the pelvic fracture fragment with the 3D point cloud of the mirror image template of the affected side of the pelvis, and determine the number of matching point cloud pairs; Repeat the above steps 2 to 5 until the preset number of iterations is reached, and determine the first transformation matrix with the largest number of matched point cloud pairs during the iteration as the initial transformation matrix.

[0008] Optionally, based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis, determining corresponding point pairs of the pelvic fracture fragment and the mirror image template of the affected side of the pelvis includes: Calculate the feature descriptors of all points in the three-dimensional point cloud of the pelvic fracture fragment based on the three-dimensional point cloud data of the pelvic fracture fragment, which is called the first feature descriptor; Calculate the feature descriptors of all points in the 3D point cloud of the mirror template based on the 3D point cloud data of the mirror template on the affected side of the pelvis, which are called the second feature descriptors; Calculate the similarity between the first feature descriptor and all second feature descriptors of each point in the three-dimensional point cloud of the pelvic fracture fragment; The similarity is compared with a preset similarity threshold, and the points corresponding to the first feature descriptor and the second feature descriptor whose similarity is greater than the preset similarity threshold are determined as corresponding point pairs of the pelvic fracture fragment and the mirror image template of the affected side of the pelvis.

[0009] Optionally, the feature descriptors of all points in the three-dimensional point cloud of the pelvic fracture fragment are calculated based on the three-dimensional point cloud data of the pelvic fracture fragment, which are called first feature descriptors and include: For each point in the three-dimensional point cloud of the pelvic fracture fragment, a preset number of neighboring points are obtained within a preset radius; Calculate the multi-dimensional SPFH descriptor of each point based on the point cloud data of each point and its neighboring points; The multi-dimensional SPFH descriptor of each point is weightedly fused with the multi-dimensional SPFH descriptor of its neighboring points using the following formula to obtain the FPFH descriptor of each point:

[0010] in, for SPFH descriptor of the point; For neighboring points SPFH descriptor; is the total number of neighboring points; is the distance weight between each point and its neighboring points.

[0011] Optionally, matching the updated three-dimensional point cloud of the pelvic fracture fragment with the three-dimensional point cloud of the mirror image template of the affected side of the pelvis includes: The updated 3D point cloud of the pelvic fracture fragment is matched with the 3D point cloud of the mirror template of the affected side of the pelvis based on the greedy matching algorithm. The greedy matching process is as follows: For one of the points in the three-dimensional point cloud of the pelvic fracture fragment, search for the closest point in the three-dimensional point cloud of the mirror template of the affected side of the pelvis and determine it as a matching point pair; Two matching points are eliminated from the three-dimensional point cloud of the pelvic fracture fragment and the three-dimensional point cloud of the mirror template of the affected side of the pelvis, and matching is continued in the remaining point clouds until all points are matched.

[0012] Optionally, based on the updated 3D point cloud data of the pelvic fracture fragment and the 3D point cloud data of the mirror template, the pelvic fracture fragment and the mirror template are finally matched using the ICP registration method to obtain a final transformation matrix including: For each point in the updated 3D point cloud of the pelvic fracture fragment, search for the nearest point in the 3D point cloud of the mirror image template on the affected side of the pelvis and form a matching point pair; Based on all matching point pairs, the optimal transformation matrix is ​​estimated using the singular value decomposition method and the pre-built objective optimization function; Use the optimal transformation matrix to update the preset initial unit matrix to obtain the latest transformation matrix; The latest transformation matrix is ​​used to continue updating the three-dimensional point cloud data of the pelvic fracture fragment to obtain the updated three-dimensional point cloud of the pelvic fracture fragment. The above ICP registration process is repeated until the preset number of iterations is reached, and the latest transformation matrix corresponding to the minimum value of the target optimization function is determined as the final transformation matrix.

[0013] Optionally, the mirror image template of the affected side of the pelvis is constructed by: Based on the original three-dimensional point cloud data of the sacrum, the original three-dimensional point cloud of the sacrum is symmetrically transformed based on the symmetry transformation matrix to obtain a three-dimensional point cloud of a symmetrical model of the sacrum; Perform ICP registration on the 3D point cloud of the symmetrical model of the sacrum and the original 3D point cloud of the sacrum to obtain a registration transformation matrix; The three-dimensional point cloud of the healthy side ilium is transformed based on the symmetric transformation matrix and the registration transformation matrix to obtain a mirror template of the affected side ilium, and the mirror template of the ilium is used as a mirror template of the affected side of the pelvis.

[0014] In a second aspect, an automatic planning device for virtual reduction of pelvic fractures based on template registration is provided, the device comprising: An acquisition unit is used to acquire three-dimensional point cloud data of the pelvic fracture fragment and three-dimensional point cloud data of a pre-constructed mirror image template of the affected side of the pelvis; wherein the mirror image template of the affected side of the pelvis is constructed based on the self-symmetry of the pelvic sacrum; The first matching unit is configured to perform initial matching between the pelvic fracture fragment and the mirror image template of the affected side of the pelvis using a SAC-IA registration method based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis, thereby obtaining an initial transformation matrix; An updating unit, configured to update the three-dimensional point cloud data of the pelvic fracture block based on the initial transformation matrix to obtain updated three-dimensional point cloud data of the pelvic fracture block; The second matching unit is used to perform final matching on the pelvic fracture fragment and the mirror template using the ICP registration method based on the updated pelvic fracture fragment 3D point cloud data and the mirror template 3D point cloud data to obtain a final transformation matrix; The resetting unit is used for automatically performing virtual resetting on the three-dimensional model of the pelvic fracture fragment based on the initial transformation matrix and the final transformation matrix.

[0015] According to a third aspect, an electronic device is provided, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus. The memory is used to store computer programs; The processor is configured to implement any of the method steps described in the first aspect when executing the program stored in the memory.

[0016] In a fourth aspect, a computer-readable storage medium is provided, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, any method step described in the first aspect is implemented.

[0017] The present invention provides an automatic planning method and device for virtual reduction of a pelvic fracture based on template registration, which obtains three-dimensional point cloud data of a pelvic fracture fragment and three-dimensional point cloud data of a pre-constructed mirror template of the affected side of the pelvis; wherein the mirror template of the affected side of the pelvis is constructed based on the self-symmetry of the pelvic sacrum; based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror template of the affected side of the pelvis, the pelvic fracture fragment and the mirror template of the affected side of the pelvis are initially matched using the SAC-IA registration method to obtain an initial transformation matrix; based on the initial transformation matrix, the three-dimensional point cloud data of the pelvic fracture fragment are updated to obtain updated three-dimensional point cloud data of the pelvic fracture fragment; based on the updated three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror template, the pelvic fracture fragment and the mirror template are finally matched using the ICP registration method to obtain a final transformation matrix; and based on the initial transformation matrix and the final transformation matrix, the three-dimensional model of the pelvic fracture fragment is automatically virtually reduced. When performing automatic pelvic reduction, the present invention obtains the transformation matrix required for reduction through two stages: coarse registration and fine registration. Coarse registration enables the pelvic fracture fragment to quickly reach the appropriate position, ensuring registration efficiency. The ICP fine registration criterion further helps the pelvic fracture fragment to be reduced to a more precise position, thereby improving reduction accuracy. Therefore, the embodiment of the present invention not only realizes automatic reduction, but also takes into account reduction efficiency and reduction accuracy, thereby improving the overall reduction effect of the pelvic fracture fragment.

[0018] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without making any creative efforts.

[0020] Figure 1 A flow chart of an automatic planning method for virtual reduction of a pelvic fracture based on template registration provided by an embodiment of the present invention is shown; FIG2 (a) shows a schematic structural diagram of the pelvis before reduction provided by an embodiment of the present invention; FIG2( b ) shows a schematic structural diagram of the pelvis after reduction provided by an embodiment of the present invention; Figure 3 A schematic structural diagram of an automatic planning device for virtual reduction of pelvic fractures based on template registration provided by an embodiment of the present invention is shown; Figure 4A structural schematic diagram of an electronic device provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0021] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0022] It is considered that, in the manual planning process, firstly, the planning time is long, and the doctor needs to constantly try and judge; secondly, the reduction position of the pelvic fracture block has observer difference and is easily affected by subjective judgment. Therefore, the efficiency and accuracy of reduction are both low, and the reduction effect is not ideal.

[0023] Based on this, the embodiments of the present application provide an automatic planning method and device for pelvic fracture virtual reduction based on template registration, which will be described below through embodiments.

[0024] The embodiments of the present application provide an automatic planning method for pelvic fracture virtual reduction based on template registration, as shown in Figure 1 The method comprises the following steps: Step S101: Obtain three-dimensional point cloud data of a pelvic fracture block and three-dimensional point cloud data of a mirror template of a pelvic affected side pre-constructed.

[0025] In the embodiments of the present application, the three-dimensional point cloud data of the pelvic fracture block refers to position coordinate data of each point in the three-dimensional point cloud of the pelvic fracture in a world coordinate system.

[0026] In one example, the three-dimensional point cloud data of the pelvic fracture block can be obtained through an STL (Standard Triangle Language) three-dimensional model file of the pelvis.

[0027] The mirror template of the pelvic affected side is constructed based on the self-symmetry of the pelvic sacrum.

[0028] In a feasible implementation, the construction process of the mirror template of the pelvic affected side comprises: Step S101A: Based on the original three-dimensional point cloud data of the sacrum, the original three-dimensional point cloud of the sacrum is symmetrically transformed based on a symmetric transformation matrix to obtain a three-dimensional point cloud of a symmetric model of the sacrum.

[0029] Step S101B: The three-dimensional point cloud of the symmetric model of the sacrum is ICP registered with the original three-dimensional point cloud of the sacrum to obtain a registration transformation matrix.

[0030] The process of ICP (Iterative Closest Point) registration can refer to the process of ICP registration in the following embodiments, which will not be repeated here.

[0031] Step S101C: Based on the symmetric transformation matrix and the registration transformation matrix, the three-dimensional point cloud of the healthy side ilium is converted to obtain a mirror image template of the affected side ilium, and the mirror image template of the ilium is taken as the mirror image template of the affected side of the pelvis.

[0032] According to the structure of the pelvis, it can be understood that the ilium is located on both sides of the sacrum and is generally symmetrical. The sacrum itself has self-symmetry, and by using the self-symmetry of the sacrum of each patient to construct the mirror image template of the affected side of the pelvis, the constructed mirror image template is more personalized and conforms to the actual pelvis structure of each patient, providing a more accurate mirror image template for subsequent reduction, thereby improving the accuracy of the reduction.

[0033] Step S102: Based on the three-dimensional point cloud data of the pelvis fracture block and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis, an SAC-IA (Sample Consensus Initial Alignment) registration method is used to perform initial matching on the pelvis fracture block and the mirror image template of the affected side of the pelvis to obtain an initial transformation matrix.

[0034] In this step, SAC-IA is an initial registration method, which first uses the SAC-IA registration method for coarse registration to quickly bring the pelvis fracture block to a good initial position.

[0035] In a feasible implementation, the SAC-IA registration method includes the following steps: First step: Based on the three-dimensional point cloud data of the pelvis fracture block and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis, corresponding point pairs of the pelvis fracture block and the mirror image template of the affected side of the pelvis are determined.

[0036] In a specific implementation, based on the three-dimensional point cloud data of the pelvis fracture block and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis, the corresponding point pairs of the pelvis fracture block and the mirror image template of the affected side of the pelvis include: Step S102A: Calculate the feature descriptors of all points in the three-dimensional point cloud of the pelvic fracture fragment based on the three-dimensional point cloud data of the pelvic fracture fragment, which are called first feature descriptors.

[0037] In a specific example, the process of determining the first feature descriptor includes: Step S102A1: For each point in the three-dimensional point cloud of the pelvic fracture fragment, a preset number of neighboring points are obtained within a preset radius.

[0038] In one example, for each point in the 3D point cloud of the pelvic fracture fragment, , first extract its Neighbors .

[0039] Step S102A2: Calculate a multi-dimensional SPFH descriptor for each point based on the point cloud data of each point and its neighboring points.

[0040] In this step, the SPFH descriptor captures the local geometric information of the point by calculating the angular relationship between each point and the surface normal vector of its neighboring points, and encodes this information into a fixed-length vector. In an example, the three angular features between each point (called the center point) and its neighboring points are mainly calculated: ,in, It represents the angle between the vector from the center point to the neighboring point and the normal of the neighboring point; Represents the ratio of the normal of the center point to the length of the projection vector from the center point to the neighboring points; Represents the angle calculated from the normals of two points and the vectors from the center point to the nearest neighbor.

[0041] The calculation formulas for these three angle features are as follows: (1); (2); (3); in, The position vector representing the center point; Represents the position vector of the neighboring points; The normal vector representing the center point; Represents the normal vector of the neighboring point.

[0042] After obtaining the three angle features, they are discretized into histograms, and each angle feature is divided into 11 intervals, forming a SPFH descriptor with a dimension of 33.

[0043] Specifically, the process of discretizing these three angle features includes: Step 1: First determine the value range of each angle eigenvalue.

[0044] In this step, find the minimum and maximum values ​​of all calculated angle eigenvalues. This step is necessary because point cloud data from different scenes may result in different distribution ranges of angle eigenvalues. The value range is determined based on the maximum and minimum values ​​of each angle eigenvalue.

[0045] Step 2: Divide each angle feature into 11 intervals based on the value range of each angle feature.

[0046] In this step, the simplest approach is to evenly divide the value range into 11 intervals of equal width. For example, if the value range of a feature is from 0 to 1, the width of each interval is 1 / 11. Another approach is frequency-based segmentation, which divides the intervals according to the actual frequency distribution of the angle eigenvalues ​​so that each interval contains roughly the same number of eigenvalues. This method is more complex, but it can better adapt to the situation where the eigenvalues ​​are unevenly distributed, thereby providing better discrimination ability.

[0047] Step 3: After the interval division is completed, each angle eigenvalue will be assigned to a corresponding interval, and then this information will be encoded into the form of a histogram.

[0048] Specifically, for each angle feature , creating a vector of length 11, representing the frequency or weight of the feature value falling in each interval. Finally, these three vectors of length 11 are combined to form a vector of dimension 33, namely the SPFH descriptor.

[0049] Step S102A3: The multi-dimensional SPFH descriptor of each point is weightedly fused with the multi-dimensional SPFH descriptors of its neighboring points using the following formula to obtain the FPFH descriptor of each point: (4); in, for SPFH descriptor of the point; For neighboring points SPFH descriptor; is the total number of neighboring points; is the distance weight between each point and its neighboring points.

[0050] Through this step, the inter-point information is utilized and the robustness of the descriptor is enhanced.

[0051] Step S102B: Calculate the feature descriptors of all points in the three-dimensional point cloud of the mirror template based on the three-dimensional point cloud data of the mirror template on the affected side of the pelvis, which are called second feature descriptors.

[0052] The calculation method of the second feature descriptor is the same as that of the first feature descriptor, and will not be repeated here.

[0053] Step S102C: Calculate the similarity between the first feature descriptor of each point in the three-dimensional point cloud of the pelvic fracture fragment and all the second feature descriptors.

[0054] In one example, the Euclidean distance or Manhattan distance between the first feature descriptor and the second feature descriptor is calculated to represent the similarity between the feature descriptors of two different point clouds.

[0055] Step S102D: Compare the similarity with a preset similarity threshold, and determine the points corresponding to the first feature descriptor and the second feature descriptor whose similarity is greater than the preset similarity threshold as corresponding point pairs of the mirror image template of the pelvic fracture fragment and the affected side of the pelvis.

[0056] By matching and comparing one by one, all corresponding point pairs in the three-dimensional point cloud of the pelvic fracture fragment and the mirror image template of the affected side of the pelvis are found.

[0057] Step 2: Randomly select a preset number of point pairs from all corresponding point pairs.

[0058] Step 3: Determine a first transformation matrix based on a preset number of point pairs.

[0059] In this step, the first transformation matrix is ​​estimated using the least squares method.

[0060] Step 4: Update the three-dimensional point cloud data of the pelvic fracture fragment based on the first transformation matrix to obtain an updated three-dimensional point cloud of the pelvic fracture fragment.

[0061] In this step, the first transformation matrix is ​​applied to all 3D point clouds of the pelvic fracture fragment and its suitability is verified. For example, the distance between the transformed points and the target point cloud (the 3D point cloud of the mirrored template of the affected side of the pelvis) is calculated. If this distance is less than a set threshold, the transformation matrix is ​​considered valid. Otherwise, the first transformation matrix needs to be re-estimated.

[0062] Step 5: Match the updated 3D point cloud of the pelvic fracture fragment with the 3D point cloud of the mirror template of the affected side of the pelvis, and determine the number of matched point cloud pairs.

[0063] In a specific embodiment, a greedy matching method can be used for matching. The greedy matching process is: Step A: For one of the points in the three-dimensional point cloud of the pelvic fracture fragment, search for the closest point in the three-dimensional point cloud of the mirror image template on the affected side of the pelvis and determine it as a matching point pair.

[0064] In this step, the nearest neighbor points can be found by calculating the Euclidean distance.

[0065] Step B: Eliminate two matching points in the three-dimensional point cloud of the pelvic fracture fragment and the three-dimensional point cloud of the mirror template of the affected side of the pelvis, and continue matching in the remaining point clouds until all points are matched.

[0066] During the matching process, some unreasonable matching point pairs may appear. These can be removed through optimization methods. For example, if the distance between a matching point pair is significantly larger than the distance between other matching point pairs, or if the geometric relationship between them does not meet the constraints of the rigid transformation, they can be removed from the matching set.

[0067] Repeat the above steps 2 to 5 until the preset number of iterations is reached, and determine the first transformation matrix with the largest number of matched point cloud pairs during the iteration as the initial transformation matrix.

[0068] Step S103: updating the three-dimensional point cloud data of the pelvic fracture block based on the initial transformation matrix to obtain updated three-dimensional point cloud data of the pelvic fracture block.

[0069] Step S104: Based on the updated 3D point cloud data of the pelvic fracture fragment and the 3D point cloud data of the mirror template, the ICP registration method is used to perform final matching on the pelvic fracture fragment and the mirror template to obtain a final transformation matrix.

[0070] The next step is to further fine-tune the registration after the rough registration. The two processes of rough registration and fine registration greatly improve the registration accuracy and ensure the registration efficiency.

[0071] In a feasible embodiment, the final matching of the pelvic fracture fragment and the mirror image template using the ICP registration method includes the following steps: Step S104A: for each point in the updated 3D point cloud of the pelvic fracture fragment, search for the nearest point in the 3D point cloud of the mirror image template of the affected side of the pelvis, and form a matching point pair; Step S104B: Based on all matching point pairs, the optimal transformation matrix is ​​estimated using a singular value decomposition method and a pre-constructed target optimization function.

[0072] In the embodiment of the present invention, the transformation matrix is ​​composed of a rotation matrix and a translation vector, which can be calculated by singular value decomposition.

[0073] In an embodiment of the present invention, the objective optimization function is: (5); in, is the rotation matrix; is the translation vector; For the The 3D point cloud of the pelvic fracture fragment during the secondary optimization; For the The three-dimensional point cloud of the coarse registration of the pelvic fracture fragment during the secondary optimization; is the total number of optimizations.

[0074] Step S104C: using the optimal transformation matrix to update the preset initial identity matrix to obtain the latest transformation matrix.

[0075] Step S104D: Use the latest transformation matrix to continue updating the three-dimensional point cloud data of the pelvic fracture fragment to obtain the updated three-dimensional point cloud of the pelvic fracture fragment. Repeat the above process until the preset number of iterations is reached, and determine the latest transformation matrix corresponding to the minimum value of the target optimization function as the final transformation matrix.

[0076] Step S105: automatically virtually repositioning the three-dimensional model of the pelvic fracture fragment based on the initial transformation matrix and the final transformation matrix.

[0077] The initial and final transformation matrices are sequentially applied to the 3D model of the pelvic fracture fragment. First, a rough reduction is performed according to the initial transformation matrix, helping the pelvic fracture fragment quickly reach the appropriate position. The final transformation matrix then assists in the reduction and adjustment, ensuring the accuracy of the reduction. This allows for automatic virtual reduction throughout the entire process. Figure 2(a) shows a schematic diagram of the pelvic structure before reduction; Figure 2(b) shows a schematic diagram of the pelvic structure after reduction.

[0078] It can be understood from the above embodiments that when the present invention performs automatic pelvic reduction, the transformation matrix required for reduction is obtained through two stages of coarse registration and fine registration. The coarse registration can quickly enable the pelvic fracture fragment to reach the appropriate position, ensuring the registration efficiency. The ICP fine registration criterion further helps the pelvic fracture fragment to be reduced to a more precise position, thereby improving the reduction accuracy. Therefore, the embodiment of the present invention not only realizes automatic reduction, but also takes into account the reduction efficiency and reduction accuracy, thereby improving the overall reduction effect of the pelvic fracture fragment.

[0079] Based on the same inventive concept, an automatic planning device for virtual reduction of pelvic fractures based on template registration is provided, such as Figure 3 As shown, the device includes: The acquisition unit 301 is used to acquire the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the pre-constructed mirror template of the affected side of the pelvis; wherein the mirror template of the affected side of the pelvis is constructed based on the self-symmetry of the pelvic sacrum.

[0080] The first matching unit 302 is used to perform initial matching between the pelvic fracture fragment and the mirror image template of the affected side of the pelvis based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis using the SAC-IA registration method to obtain an initial transformation matrix.

[0081] The updating unit 303 is configured to update the three-dimensional point cloud data of the pelvic fracture block based on the initial transformation matrix to obtain updated three-dimensional point cloud data of the pelvic fracture block.

[0082] The second matching unit 304 is used to perform final matching on the pelvic fracture block and the mirror template using the ICP registration method based on the updated pelvic fracture block 3D point cloud data and the mirror template 3D point cloud data to obtain a final transformation matrix.

[0083] The resetting unit 305 is configured to automatically and virtually resize the three-dimensional model of the pelvic fracture fragment based on the initial transformation matrix and the final transformation matrix.

[0084] Based on the same technical concept, an embodiment of the present invention further provides an electronic device, such as Figure 4 As shown, it includes a processor 401 , a communication interface 402 , a memory 403 and a communication bus 404 , wherein the processor 401 , the communication interface 402 and the memory 403 communicate with each other via the communication bus 404 .

[0085] Memory 403, used for storing computer programs; The processor 401 is configured to implement the steps of the automatic planning method for virtual reduction of pelvic fractures based on template registration when executing the program stored in the memory 403 .

[0086] The communication bus mentioned in the electronic devices mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. This communication bus can be divided into address buses, data buses, control buses, etc. For ease of illustration, only a single thick line is used in the figure, but this does not mean that there is only one bus or only one type of bus.

[0087] The communication interface is used for communication between the above electronic device and other devices.

[0088] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage. Alternatively, the memory may be at least one storage device located away from the processor.

[0089] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0090] The computer program product of the automatic planning method for virtual reduction of pelvic fractures based on template alignment provided in an embodiment of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the method described in the previous method embodiment. The specific implementation can be found in the method embodiment and will not be repeated here.

[0091] The device for automatic planning of virtual reduction of pelvic fractures based on template registration provided in the embodiment of the present invention can be specific hardware on the device or software or firmware installed on the device. The implementation principle and technical effects of the device provided in the embodiment of the present invention are the same as those in the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the device embodiment, reference can be made to the corresponding content in the aforementioned method embodiment. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can all refer to the corresponding processes in the aforementioned method embodiment, and will not be repeated here.

[0092] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interface, the indirect coupling or communication connection of the device or unit may be electrical, mechanical or other forms.

[0093] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment of the present application.

[0094] In addition, the functional units in the embodiments provided by the present application can be integrated in one processing unit, or each unit can exist alone physically, or two or more units can be integrated in one unit.

[0095] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0096] It should be noted that: similar reference numerals and letters represent similar items in the following drawings, therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings, and in addition, the terms "first", "second", "third" and the like are only used to distinguish description, and cannot be understood as indicating or implying relative importance.

[0097] Finally, it should be noted that: the above-described embodiments are only specific embodiments of the present application, which are used to illustrate the technical solutions of the present application, and are not limited thereto, the protection scope of the present application is not limited thereto, although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art within the technical range disclosed by the present application can still modify or easily think of changes to the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications, changes or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application. They should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An automatic planning method for virtual reduction of pelvic fractures based on template registration, characterized in that: The method comprises: Acquire three-dimensional point cloud data of the pelvic fracture fragment and three-dimensional point cloud data of a pre-constructed mirror image template of the affected side of the pelvis; wherein the mirror image template of the affected side of the pelvis is constructed based on the self-symmetry of the pelvic sacrum; Based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror template of the affected side of the pelvis, the SAC-IA registration method is used to perform initial matching on the pelvic fracture fragment and the mirror template of the affected side of the pelvis to obtain an initial transformation matrix; updating the three-dimensional point cloud data of the pelvic fracture fragment based on the initial transformation matrix to obtain updated three-dimensional point cloud data of the pelvic fracture fragment; Based on the updated 3D point cloud data of the pelvic fracture fragment and the 3D point cloud data of the mirror template, the ICP registration method is used to perform a final match between the pelvic fracture fragment and the mirror template to obtain a final transformation matrix; The three-dimensional model of the pelvic fracture fragment is automatically virtually reset based on the initial transformation matrix and the final transformation matrix.

2. The method according to claim 1, characterized in that The three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis are used to perform initial matching on the pelvic fracture fragment and the mirror image template of the affected side of the pelvis using the SAC-IA registration method to obtain an initial transformation matrix including: Step 1: Based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis, determining corresponding point pairs of the pelvic fracture fragment and the mirror image template of the affected side of the pelvis; Step 2: Randomly select a preset number of point pairs from all corresponding point pairs; Step 3: determining a first transformation matrix based on the preset number of point pairs; Step 4: updating the three-dimensional point cloud data of the pelvic fracture fragment based on the first transformation matrix to obtain an updated three-dimensional point cloud of the pelvic fracture fragment; Step 5: Match the updated 3D point cloud of the pelvic fracture fragment with the 3D point cloud of the mirror image template of the affected side of the pelvis, and determine the number of matching point cloud pairs; Repeat the above steps 2 to 5 until the preset number of iterations is reached, and determine the first transformation matrix with the largest number of matched point cloud pairs during the iteration as the initial transformation matrix.

3. The method according to claim 2, characterized in that Determining corresponding point pairs of the pelvic fracture fragment and the mirror image template of the affected side of the pelvis based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis includes: Calculating feature descriptors of all points in the three-dimensional point cloud of the pelvic fracture fragment based on the three-dimensional point cloud data of the pelvic fracture fragment, which is called a first feature descriptor; Calculating feature descriptors of all points in the three-dimensional point cloud of the mirror template based on the three-dimensional point cloud data of the mirror template of the affected side of the pelvis, which are called second feature descriptors; Calculate the similarity between the first feature descriptor and all second feature descriptors of each point in the three-dimensional point cloud of the pelvic fracture fragment; The similarity is compared with a preset similarity threshold, and points corresponding to the first feature descriptor and the second feature descriptor whose similarity is greater than the preset similarity threshold are determined as corresponding point pairs of the mirror image template of the pelvic fracture fragment and the affected side of the pelvis.

4. The method according to claim 3, characterized in that The feature descriptors of all points in the three-dimensional point cloud of the pelvic fracture fragment calculated based on the three-dimensional point cloud data of the pelvic fracture fragment, referred to as the first feature descriptor, include: For each point in the three-dimensional point cloud of the pelvic fracture fragment, obtaining a preset number of neighboring points within a preset radius; Calculate the multi-dimensional SPFH descriptor of each point based on the point cloud data of each point and its neighboring points; The multi-dimensional SPFH descriptor of each point is weightedly fused with the multi-dimensional SPFH descriptor of its neighboring points using the following formula to obtain the FPFH descriptor of each point: in, for SPFH descriptor of the point; For neighboring points SPFH descriptor; is the total number of neighboring points; is the distance weight between each point and its neighboring points.

5. The method according to claim 2, characterized in that The matching of the updated three-dimensional point cloud of the pelvic fracture fragment with the three-dimensional point cloud of the mirror image template of the affected side of the pelvis includes: The updated 3D point cloud of the pelvic fracture fragment is matched with the 3D point cloud of the mirror template of the affected side of the pelvis based on the greedy matching algorithm. The process of the greedy matching algorithm is as follows: For one of the points in the three-dimensional point cloud of the pelvic fracture fragment, searching for the closest point in the three-dimensional point cloud of the mirror image template of the affected side of the pelvis, and determining it as a matching point pair; Two matching points are removed from the three-dimensional point cloud of the pelvic fracture fragment and the three-dimensional point cloud of the mirror image template of the affected side of the pelvis, and matching is continued in the remaining point clouds until all points are matched.

6. The method according to claim 1, characterized in that The updated 3D point cloud data of the pelvic fracture fragment and the 3D point cloud data of the mirror template are used to perform final matching on the pelvic fracture fragment and the mirror template using the ICP registration method to obtain the final transformation matrix including: For each point in the updated three-dimensional point cloud of the pelvic fracture fragment, searching for the nearest point in the three-dimensional point cloud of the mirror image template of the affected side of the pelvis, and forming a matching point pair; Based on all matching point pairs, the optimal transformation matrix is ​​estimated using the singular value decomposition method and the pre-built objective optimization function; Use the optimal transformation matrix to update the preset initial unit matrix to obtain the latest transformation matrix; The latest transformation matrix is ​​used to continue updating the three-dimensional point cloud data of the pelvic fracture fragment to obtain the updated three-dimensional point cloud of the pelvic fracture fragment. The above ICP alignment process is repeated until the preset number of iterations is reached, and the latest transformation matrix corresponding to the minimum value of the target optimization function is determined as the final transformation matrix.

7. The method according to claim 1, characterized in that The process of constructing the mirror image template of the affected side of the pelvis includes: Based on the original three-dimensional point cloud data of the sacrum, the original three-dimensional point cloud of the sacrum is symmetrically transformed based on the symmetry transformation matrix to obtain a three-dimensional point cloud of a symmetrical model of the sacrum; Perform ICP registration on the 3D point cloud of the symmetrical model of the sacrum and the original 3D point cloud of the sacrum to obtain a registration transformation matrix; The three-dimensional point cloud of the healthy side ilium is transformed based on the symmetric transformation matrix and the registration transformation matrix to obtain a mirror template of the affected side ilium, and the mirror template of the ilium is used as a mirror template of the affected side of the pelvis.

8. An automatic planning device for virtual reduction of pelvic fractures based on template registration, characterized in that: The device comprises: An acquisition unit, configured to acquire three-dimensional point cloud data of the pelvic fracture fragment and three-dimensional point cloud data of a pre-constructed mirror image template of the affected side of the pelvis; wherein the mirror image template of the affected side of the pelvis is constructed based on the self-symmetry of the pelvic sacrum; A first matching unit is configured to perform initial matching on the pelvic fracture fragment and the mirror image template of the affected side of the pelvis using a SAC-IA registration method based on the three-dimensional point cloud data of the pelvic fracture fragment and the three-dimensional point cloud data of the mirror image template of the affected side of the pelvis, to obtain an initial transformation matrix; an updating unit, configured to update the three-dimensional point cloud data of the pelvic fracture block based on the initial transformation matrix to obtain updated three-dimensional point cloud data of the pelvic fracture block; A second matching unit is configured to perform a final matching between the pelvic fracture fragment and the mirror template using an ICP registration method based on the updated pelvic fracture fragment 3D point cloud data and the mirror template 3D point cloud data to obtain a final transformation matrix; A resetting unit is used to automatically virtually resize the three-dimensional model of the pelvic fracture fragment based on the initial transformation matrix and the final transformation matrix.

9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; The memory is used to store computer programs; The processor is configured to implement the method steps described in any one of claims 1 to 7 when executing the program stored in the memory.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method steps according to any one of claims 1 to 7 are implemented.

Images