Point cloud registration method, system and equipment for osteotomy around acetabulum
Through a point cloud registration method combining global and local characteristics, the problem of low registration accuracy in periacetabular osteotomy is solved, achieving higher surgical accuracy and lower complication risk.
Patent Information
- Application Number
- CN202411977852.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-31
- Publication Date
- 2025-06-03
AI Technical Summary
Peripheral osteotomy of the acetabular osteotomy is not accurate due to deep areas, widespread neurovascular, difficulty in surgical exposure and diverse hip lesions, which leads to low accuracy, which can easily cause serious complications, and lacks reliable and accurate registration methods.
A point cloud registration method is adopted to obtain the patient's hip joint image, perform segmentation and three-dimensional reconstruction, select feature points, and extract the registration feature map between point clouds using a fusion method of global and local features, and calculate the registration matrix to determine the registration parameters.
It improves the registration accuracy of periacetabular osteotomy, enhances the accuracy of the surgery, reduces the occurrence of complications, and provides a reliable registration method.
Smart Images
Figure CN120088296A_ABST
Abstract
Description
Technical Field
[0001] The present application belongs to the field of point cloud registration, and in particular, relates to a point cloud registration method, system, device and computer-readable storage medium for periacetabular osteotomy. Background Art
[0002] Periacetabular osteotomy (PAO) is a surgical procedure to treat hip dysplasia or insufficient acetabular depth. This procedure relieves hip pain and improves function by repositioning the acetabulum, improving acetabular coverage and stability. This procedure is one of the most complex orthopedic surgeries and is the crown jewel of orthopedic surgery.
[0003] However, since the periacetabular osteotomy is located deep and surrounded by nerves and blood vessels, surgical exposure is difficult and the exposure range is limited. In addition, the diversity of hip lesions, such as joint deformity, arthritis, and previous osteotomy history, makes periacetabular osteotomy inaccurate, which can easily cause severe complications of sciatic ramus and posterior column truncation, bringing disastrous consequences to patients. The use of navigation or robotic osteotomy is an important direction of development, but there is currently no reliable registration method and the registration accuracy is poor.
[0004] At present, the point cloud registration method in hip replacement surgery related technologies is to use the Landmark algorithm and the vtk_ICP algorithm to combine coarse and fine registration. However, due to the abnormally complex local anatomy of periacetabular osteotomy, which is located in the vascular and neural networks of the hip and pelvis, it is difficult to select points for registration. So far, there is still no reliable and accurate registration method and matching system in clinical practice.
[0005] Therefore, finding a reliable registration method for periacetabular osteotomy and improving the registration accuracy of periacetabular osteotomy are technical problems that need to be solved urgently by those skilled in the art. Summary of the invention
[0006] The embodiments of the present application provide a point cloud registration method, system, device and computer-readable storage medium for periacetabular osteotomy, which can achieve the registration of periacetabular osteotomy and improve the registration accuracy of periacetabular osteotomy.
[0007] In a first aspect, an embodiment of the present application provides a point cloud registration method for periacetabular osteotomy, comprising:
[0008] Obtain an image of the patient's hip joint;
[0009] Segment and 3D reconstruct the hip joint image to obtain a 3D model of the hip joint;
[0010] Select feature points based on the three-dimensional model of the hip joint;
[0011] Extract the features between point clouds using the fusion method of global and local features, and obtain the registration feature maps of the source point cloud and the target point cloud respectively;
[0012] Calculate the registration matrix based on the registration feature maps of the source point cloud and the target point cloud to determine the registration parameters.
[0013] Optionally, extracting the features between point clouds using the fusion method of global and local features, and obtaining the registration feature maps of the source point cloud and the target point cloud respectively, includes:
[0014] Use the method of convolution and multi-layer perceptron (MLP) for global feature extraction to effectively ensure the accuracy of global information extraction;
[0015] Fuse the Mask feature map of the current iteration with the feature map of the previous iteration to obtain the registration feature maps of the source point cloud and the target point cloud respectively.
[0016] Optionally, using the method of convolution and multi-layer perceptron (MLP) for global feature extraction to effectively ensure the accuracy of global information extraction, includes:
[0017] Aggregate local features into global features, and the aggregation methods are max pooling, average pooling or weighted pooling;
[0018] Input the global features into a multi-layer perceptron, and further extract high-dimensional features through a series of fully connected layers, activation functions, and normalization operations;
[0019] Use MLP to map the high-dimensional features to the dimensions required for the registration task.
[0020] Optionally, fusing the Mask feature map of the current iteration with the feature map of the previous iteration to obtain the registration feature maps of the source point cloud and the target point cloud respectively, includes:
[0021] The current Mask feature map and the previous iteration feature map are fused to generate new features in the following way:
[0022] Assign weights to each feature, and the weights are learned by the network or set manually to obtain the fused feature map;
[0023] Concatenate the fused feature map with the Mask feature map of the current iteration and the feature map of the previous iteration along the feature dimension respectively to obtain the registration feature maps of the source point cloud and the target point cloud.
[0024] Optionally, calculating the registration matrix based on the registration feature maps of the source point cloud and the target point cloud to determine the registration parameters, includes:
[0025] Input registration feature maps: source point cloud feature map and target point cloud feature map, which respectively represent the feature representations of each point in the source point cloud and the target point cloud;
[0026] Use Euclidean distance or cosine similarity to measure the similarity between the source point cloud and the target point cloud features, and generate a matching score matrix;
[0027] For each source point cloud feature, find the most similar point in the target point cloud features and record the matching point pairs;
[0028] Use the matching point pairs to calculate the registration matrix to determine the registration parameters.
[0029] Optionally, segment and three-dimensionally reconstruct the hip joint image to obtain a three-dimensional hip joint model, including:
[0030] Use the threshold segmentation method to segment the CT image data, separate the blood vessels and the background, and optimize the threshold segmentation result, and smooth the bone quality and surface prosthesis, etc., to obtain a better surface appearance;
[0031] Adopt a deep learning method, a deep learning network based on a feature fusion module, to fuse the global features and the target feature map to obtain the target image;
[0032] Use a conditional generative adversarial network and a bidirectional long short-term memory network for more accurate bone segmentation;
[0033] Extract the surface from the segmentation result, and use the unique surface reconstruction module of the software to three-dimensionally reconstruct the surface prosthesis;
[0034] Post-process the reconstructed surface, smooth the surface or perform transparency modification to obtain a clear three-dimensional model.
[0035] Optionally, after calculating the registration matrix to determine the registration parameters, it further includes:
[0036] According to the obtained registration parameters by matching, register the preoperative and intraoperative hip joint images to guide the osteotomy operation during the surgery;
[0037] Evaluate the accuracy of the registration by comparing the image differences before and after registration, and adjust the registration parameters as needed.
[0038] In a second aspect, an embodiment of the present application provides a point cloud registration system for periacetabular osteotomy, including:
[0039] An image acquisition module, configured to acquire hip joint images of a patient;
[0040] A three-dimensional reconstruction module, configured to segment and three-dimensionally reconstruct the hip joint image to obtain a three-dimensional hip joint model;
[0041] A feature point selection module, configured to select feature points based on a three-dimensional hip joint model;
[0042] A feature extraction module, configured to extract features between point clouds by using a fusion method of global and local features, and respectively obtain registration feature maps of a source point cloud and a target point cloud;
[0043] A registration matrix calculation module, configured to calculate a registration matrix according to the registration feature maps of the source point cloud and the target point cloud to determine registration parameters.
[0044] In a third aspect, an embodiment of the present application provides an electronic device, which includes: a processor and a memory storing computer program instructions;
[0045] When the processor executes the computer program instructions, the point cloud registration method for periacetabular osteotomy is implemented.
[0046] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the point cloud registration method for periacetabular osteotomy is implemented.
[0047] The point cloud registration method, system, device and computer-readable storage medium for periacetabular osteotomy in the embodiments of the present application can implement the registration of periacetabular osteotomy and improve the registration accuracy of periacetabular osteotomy.
[0048] The point cloud registration method for periacetabular osteotomy includes:
[0049] Obtain a hip joint image of a patient;
[0050] Segment and three-dimensionally reconstruct the hip joint image to obtain a three-dimensional hip joint model;
[0051] Based on the three-dimensional hip joint model, select feature points;
[0052] Use a fusion method of global and local features to extract features between point clouds, and respectively obtain registration feature maps of a source point cloud and a target point cloud;
[0053] According to the registration feature maps of the source point cloud and the target point cloud, calculate a registration matrix to determine registration parameters. Description of the Drawings
[0054] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.
[0055] Figure 1 It is a schematic flow chart of a point cloud registration method for periacetabular osteotomy provided by an embodiment of the present application;
[0056] Figure 2 It is a schematic planning diagram of periacetabular osteotomy provided by an embodiment of the present application;
[0057] Figure 3 It is a schematic diagram of feature point extraction provided by an embodiment of the present application;
[0058] Figure 4 It is a schematic structural diagram of a point cloud registration system for periacetabular osteotomy provided by an embodiment of the present application;
[0059] Figure 5 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners
[0060] The features and exemplary embodiments of various aspects of the present application will be described in detail below. For the purpose of making the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by showing examples of the present application.
[0061] It should be noted that, in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover a non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, elements defined by the statement "comprising..." do not preclude the existence of additional identical elements in the process, method, article or device comprising the said elements.
[0062] To solve the problems of the prior art, embodiments of the present application provide a point cloud registration method, system, device and computer-readable storage medium for periacetabular osteotomy. The point cloud registration method for periacetabular osteotomy provided by the embodiments of the present application will be introduced first below.
[0063] Figure 1The flowchart of the point cloud registration method for periacetabular osteotomy provided by an embodiment of the present application is shown. Figure 2 It is a schematic planning diagram of periacetabular osteotomy provided by an embodiment of the present application. As Figure 1 shown, the point cloud registration method for periacetabular osteotomy includes:
[0064] S101. Obtain the hip joint image of the patient;
[0065] S102. Segment and three-dimensionally reconstruct the hip joint image to obtain a three-dimensional model of the hip joint;
[0066] S103. Select feature points based on the three-dimensional model of the hip joint;
[0067] Figure 3 It is a schematic diagram of feature point extraction provided by an embodiment of the present application.
[0068] S104. Use the fusion method of global and local features to extract features between point clouds, and obtain the registration feature maps of the source point cloud and the target point cloud respectively;
[0069] Using the fusion method of global and local features to extract features between point clouds is more conducive to taking into account both global and local information during the registration process, making the registration more accurate.
[0070] S105. Calculate the registration matrix according to the registration feature maps of the source point cloud and the target point cloud to determine the registration parameters.
[0071] In one embodiment, using the fusion method of global and local features to extract features between point clouds and obtaining the registration feature maps of the source point cloud and the target point cloud respectively includes:
[0072] Use the method of convolution and multi-layer perceptron MLP to extract global features, effectively ensuring the accuracy of global information extraction;
[0073] Fuse the Mask feature map of the current iteration with the feature map of the previous iteration at the same time to obtain the registration feature maps of the source point cloud and the target point cloud respectively.
[0074] In this embodiment, the global feature extraction module (GFEM) uses the method of convolution and multi-layer perceptron MLP to extract features, which can effectively ensure the accuracy of global information extraction.
[0075] In one embodiment, using the method of convolution and multi-layer perceptron MLP to extract global features and effectively ensuring the accuracy of global information extraction includes:
[0076] Aggregate local features into global features, and the aggregation method is max pooling, average pooling or weighted pooling;
[0077] Input the global features into a multi-layer perceptron, and further extract high-dimensional features through a series of fully connected layers, activation functions, and normalization operations;
[0078] Use the MLP to map the high-dimensional features to the dimensions required for the registration task.
[0079] In one embodiment, the Mask feature map of the current iteration is simultaneously fused with the feature map of the previous iteration to obtain the registration feature maps of the source point cloud and the target point cloud respectively, including:
[0080] The current Mask feature map and the previous round feature map are fused in the following way to generate new features:
[0081] Assign weights to each feature, and the weights are learned by the network or set manually to obtain the fused feature map;
[0082] The fused feature map is concatenated with the Mask feature map of the current iteration and the feature map of the previous iteration along the feature dimension respectively to obtain the registration feature maps of the source point cloud and the target point cloud.
[0083] In this embodiment, the Mask extraction module not only obtains the Mask feature map of the current iteration, but also fuses it with the feature map of the previous iteration to obtain the registration feature maps of the source point cloud and the target point cloud respectively, providing richer feature information for the following registration matrix calculation to determine the optimal registration parameters.
[0084] In one embodiment, according to the registration feature maps of the source point cloud and the target point cloud, calculate the registration matrix to determine the registration parameters, including:
[0085] Input the registration feature maps: the source point cloud feature map and the target point cloud feature map, which respectively represent the feature representations of each point of the source point cloud and the target point cloud;
[0086] Use the Euclidean distance or cosine similarity to measure the similarity between the features of the source point cloud and the target point cloud, and generate a matching score matrix;
[0087] For each source point cloud feature, find the most similar point in the target point cloud features and record the matching point pairs;
[0088] Use the matching point pairs to calculate the registration matrix to determine the registration parameters.
[0089] In one embodiment, segment and three-dimensionally reconstruct the hip joint image to obtain a three-dimensional hip joint model, including:
[0090] Use the threshold segmentation method to segment the CT image data, separate the blood vessels and the background, and optimize the threshold segmentation result, and smooth the bone quality and surface prosthesis, etc., to obtain a better surface appearance;
[0091] Using a deep learning method, a deep learning network based on a feature fusion module fuses the global feature and the target feature map to obtain a target image;
[0092] Using a conditional generative adversarial network and a bidirectional long short-term memory network for more accurate bone segmentation;
[0093] Extract the surface from the segmentation result, and use the unique surface reconstruction module of the software to perform three-dimensional reconstruction on the surface prosthesis;
[0094] Perform post-processing on the reconstructed surface, such as smoothing the surface or performing transparency modification, to obtain a clear three-dimensional model.
[0095] In one embodiment, after calculating the registration matrix to determine the registration parameters, it further includes:
[0096] According to the obtained registration parameters, register the preoperative and intraoperative hip joint images to guide the osteotomy operation during the surgery;
[0097] Evaluate the accuracy of the registration by comparing the image differences before and after registration, and adjust the registration parameters as needed.
[0098] This application uses a deep learning positioning and registration algorithm to accurately locate, segment, and identify the local features of the pelvis and acetabulum, which can assist doctors in planning the periacetabular osteotomy surgery.
[0099] The periacetabular osteotomy registration algorithm of deep learning in this application is applied in the medical field, so that due to the deep location of the periacetabular osteotomy, the surrounding is full of nerves and blood vessels, the surgical exposure is difficult and the range is limited, and the diversity of hip diseases, resulting in the lack of an effective registration method for periacetabular osteotomy and the problem of low accuracy is solved, and it is widely used in precision medicine.
[0100] Figure 4 It is a schematic structural diagram of a point cloud registration system for periacetabular osteotomy provided by an embodiment of this application; the point cloud registration system for periacetabular osteotomy includes:
[0101] An image acquisition module 301, configured to acquire hip joint images of a patient;
[0102] A three-dimensional reconstruction module 302, configured to segment and three-dimensionally reconstruct the hip joint image to obtain a three-dimensional hip joint model;
[0103] A feature point selection module 303, configured to select feature points based on the three-dimensional hip joint model;
[0104] The feature extraction module 304 is used to extract features between point clouds in a way that fuses global and local features, and respectively obtain the registration feature maps of the source point cloud and the target point cloud.
[0105] The registration matrix calculation module 305 is used to calculate the registration matrix according to the registration feature maps of the source point cloud and the target point cloud to determine the registration parameters.
[0106] Figure 5 The structural schematic diagram of the electronic device provided by the embodiment of the present application is shown.
[0107] The electronic device may include a processor 401 and a memory 402 storing computer program instructions.
[0108] Specifically, the above-mentioned processor 401 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.
[0109] The memory 402 may include a mass storage for data or instructions. By way of example and not limitation, the memory 402 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disc, a magneto-optical disc, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In a suitable case, the memory 402 may include a removable or non-removable (or fixed) medium. In a suitable case, the memory 402 may be inside or outside the electronic device. In a specific embodiment, the memory 402 may be a non-volatile solid state memory.
[0110] In one embodiment, the memory 402 may be a read only memory (ROM). In one embodiment, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or a flash memory or a combination of two or more of these.
[0111] The processor 401 reads and executes the computer program instructions stored in the memory 402 to implement any one of the point cloud registration methods for periacetabular osteotomy in the above embodiments.
[0112] In one example, the electronic device may further include a communication interface 403 and a bus 410. Among them, as Figure 4As shown, a processor 401, a memory 402, and a communication interface 403 are connected via a bus 410 and communicate with each other.
[0113] The communication interface 403 is mainly used to implement communication between various modules, systems, units, and / or devices in the embodiments of the present application.
[0114] The bus 410 includes hardware, software, or both, and couples the components of the electronic device to each other. By way of example and not limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses or a combination of two or more of these. In suitable cases, the bus 410 may include one or more buses. Although the embodiments of the present application describe and illustrate specific buses, the present application contemplates any suitable bus or interconnect.
[0115] In addition, in combination with the point cloud registration method for periacetabular osteotomy in the above embodiments, the embodiments of the present application can be implemented by providing a computer-readable storage medium. Computer program instructions are stored on the computer-readable storage medium; when the computer program instructions are executed by a processor, any one of the point cloud registration methods for periacetabular osteotomy in the above embodiments is implemented.
[0116] It should be clear that the present application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications, and additions, or change the order between steps after understanding the spirit of the present application.
[0117] The functional modules shown in the above structural block diagrams can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a functional card, and so on. When implemented in software, the elements of the present application are programs or code segments for performing the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted via a data signal carried in a carrier wave over a transmission medium or a communication link. A "machine-readable medium" can include any medium that can store or transmit information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, and so on. The code segment can be downloaded via a computer network such as the Internet, an intranet, and so on.
[0118] It should also be noted that the exemplary embodiments mentioned in the present application describe some methods or systems based on a series of steps or systems. However, the present application is not limited to the order of the above steps, that is, the steps can be executed in the order mentioned in the embodiments, can be different from the order in the embodiments, or several steps can be executed simultaneously.
[0119] Aspects of the present application have been described above with reference to the flowcharts and / or block diagrams of methods, systems, and computer program products according to embodiments of the present application. It should be understood that each block in the flowcharts and / or block diagrams, and the combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing system to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing system enable the implementation of the functions / actions specified in one or more blocks of the flowcharts and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each block in the block diagrams and / or flowcharts, and the combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware for performing the specified functions or actions, or by a combination of dedicated hardware and computer instructions.
[0120] As described above, this is only the specific implementation manner of the present application. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, modules, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein. It should be understood that the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present application.
Claims
1. A point cloud registration method for periacetabular osteotomy, characterized in that: include: Obtain images of the patient's hip joint; Segment and 3D reconstruct the hip joint image to obtain a 3D model of the hip joint; Select feature points based on the three-dimensional model of the hip joint; The features between point clouds are extracted by fusing global and local features, and the registration feature maps of the source point cloud and the target point cloud are obtained respectively; According to the registration feature maps of the source point cloud and the target point cloud, the registration matrix is calculated to determine the registration parameters.
2. The point cloud registration method for periacetabular osteotomy according to claim 1, characterized in that: The features between point clouds are extracted by fusing global and local features, and the registration feature maps of the source point cloud and the target point cloud are obtained respectively, including: Use convolution and multi-layer perceptron MLP to extract global features, effectively ensuring the accuracy of global information extraction; The Mask feature map of this iteration is fused with the feature map of the previous iteration to obtain the registration feature maps of the source point cloud and the target point cloud respectively.
3. The point cloud registration method for periacetabular osteotomy according to claim 2, characterized in that: Use convolution and multi-layer perceptron MLP to extract global features, effectively ensuring the accuracy of global information extraction, including: Aggregate local features into global features, the aggregation method is maximum pooling, average pooling or weighted pooling; The global features are input into a multi-layer perceptron, and high-dimensional features are further extracted through a series of fully connected layers, activation functions, and normalization operations; Use MLP to map high-dimensional features to the dimensions required for the registration task.
4. The point cloud registration method for periacetabular osteotomy according to claim 3, characterized in that: The Mask feature map of this iteration is fused with the feature map of the previous iteration to obtain the registration feature maps of the source point cloud and the target point cloud, including: The current Mask feature map and the previous round feature map are fused to generate new features in the following way: Assign weights to each feature. The weights are learned through the network or set manually to obtain the fused feature map. The fused feature map is concatenated with the Mask feature map of this iteration and the feature map of the previous iteration along the feature dimension to obtain the registration feature maps of the source point cloud and the target point cloud.
5. The point cloud registration method for periacetabular osteotomy according to claim 4, characterized in that: According to the registration feature map of the source point cloud and the target point cloud, the registration matrix is calculated to determine the registration parameters, including: Input registration feature map: source point cloud feature map, target point cloud feature map, which represent the feature representation of each point in the source point cloud and target point cloud respectively; Use Euclidean distance or cosine similarity to measure the similarity between the source point cloud and the target point cloud features and generate a matching score matrix; For each source point cloud feature, find the most similar point in the target point cloud feature and record the matching point pair; Using the matched point pairs, a registration matrix is calculated to determine the registration parameters.
6. The point cloud registration method for periacetabular osteotomy according to claim 5, characterized in that: The hip joint image is segmented and 3D reconstructed to obtain a 3D model of the hip joint, including: The CT image data is segmented using the threshold segmentation method to separate the blood vessels and the background, and the threshold segmentation results are optimized to smooth the bone and surface prosthesis to obtain a better surface appearance; Using deep learning methods, the deep learning network based on the feature fusion module is used to fuse the global features and the target feature map to obtain the target image. Use conditional generative adversarial networks and bidirectional long short-term memory networks for more accurate bone segmentation; Extract the surface from the segmentation results and use the software's unique surface reconstruction module to perform three-dimensional reconstruction of the surface prosthesis; The reconstructed surface is post-processed to smooth the surface or make it transparent to obtain a clear three-dimensional model.
7. The point cloud registration method for periacetabular osteotomy according to claim 6, characterized in that: After calculating the registration matrix to determine the registration parameters, it also includes: According to the matching registration parameters, the preoperative and intraoperative hip joint images are registered to guide the osteotomy operation during the operation; By comparing the image differences before and after registration, the accuracy of registration can be evaluated and the registration parameters can be adjusted as needed.
8. A point cloud registration system for periacetabular osteotomy, characterized in that: The system comprises: An image acquisition module, used for acquiring a hip joint image of a patient; A three-dimensional reconstruction module is used to segment and reconstruct the hip joint image to obtain a three-dimensional model of the hip joint; A feature point selection module is used to select feature points based on the three-dimensional model of the hip joint; The feature extraction module is used to extract features between point clouds by fusing global and local features, and obtain the registration feature maps of the source point cloud and the target point cloud respectively; The registration matrix calculation module is used to calculate the registration matrix to determine the registration parameters according to the registration feature map of the source point cloud and the target point cloud.
9. An electronic device, characterized in that: The electronic device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the point cloud registration method for periacetabular osteotomy as described in any one of claims 1-7 is implemented.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the point cloud registration method for periacetabular osteotomy according to any one of claims 1 to 7 is implemented.