Orthopedic implantation post-operation X-ray film simulation method and system based on preoperative CT image, medium, program product and terminal

Abstract

The invention provides an orthopaedic implantation post-operation X-ray film simulation method and system based on a preoperative CT image, a medium, a program product and a terminal, and aims to accurately convert and present results of three-dimensional segmentation, key point identification and prosthesis positioning based on preoperative CT into a standard post-operation X-ray film simulation effect conforming to clinical routine. A doctor can intuitively and clearly foresee the form projection and the specific spatial position of selected prostheses including an acetabular cup, a lining, a ball head, a femoral stem and the like on a post-operation standard orthotopic X-ray film and the accurate relative anatomical relation between the prostheses and key bony markers of the pelvis and the femur in the pre-operation planning stage. By providing an intuitive and quantitative post-operation X-ray film simulation effect, a doctor is effectively assisted in selecting an optimal operation scheme and adjusting setting of parameters such as the size, the position and the angle of an implant, the predictability and the accuracy of operation planning are improved, and dependence on pure experience is reduced.

Description

Technical Field

[0001] This application relates to the field of preoperative X-ray simulation, and in particular to a method, system, medium, program product and terminal for simulating postoperative X-rays of orthopedic implantation based on preoperative CT images. Background Technology

[0002] In modern orthopedic surgery, especially total hip arthroplasty (THA), 3D reconstruction and surgical planning based on preoperative CT images have become crucial. Using technologies such as deep learning, precise segmentation and key point identification (e.g., teardrop shape, ischial tuberosity) of skeletal structures like the pelvis and femur can be performed to construct 3D models. Based on these models, surgeons can visually select and locate prosthetic components (e.g., acetabular cup, liner, ball head, femoral stem) in 3D space for preliminary implantation planning. However, current technologies have the following limitations:

[0003] 1. Disconnect between 3D Views and Clinical Gold Standards: Although 3D planning provides a stereoscopic perspective, its presentation differs significantly from the gold standard relied upon by orthopedic surgeons for postoperative evaluation—the standard 2D X-ray. Surgeons are accustomed to assessing the final position, shape, and anatomical relationship of the implant to bony landmarks, such as the acetabular cup tilt, anteversion angle, limb length, and offset, on anteroposterior radiographs, typically pelvic radiographs.

[0004] 2. Lack of intuitive 2D simulation output: Existing surgical planning systems based on 3D models cannot generate simulated postoperative X-ray views that conform to clinical practice and have diagnostic value. The planning results are difficult to directly map onto the plain radiograph view that doctors are most familiar with and use for interpreting the final outcome.

[0005] 3. Difficulty in predicting key clinical parameters: Due to the lack of accurate simulation radiographs, doctors cannot intuitively and quantitatively predict the specific shape of the selected prosthesis and its placement on the postoperative standard X-ray, the precise spatial position relative to key bony landmarks, and the interrelationships between various prosthesis components, such as the coverage relationship of the ball head in the acetabular cup and the femoral stem axis, during the preoperative planning stage.

[0006] 4. Optimization of surgical plans relies on experience: The above-mentioned defects make the comparison and optimization process of various surgical plans, including different prosthesis sizes, models, placement positions and angles, less intuitive, objective and quantitative. To a large extent, it still relies on the doctor's clinical experience to make judgments and selections, and it is difficult to accurately predict the specific differences of different plans on the final plain film before surgery. Summary of the Invention

[0007] In view of the shortcomings of the prior art described above, the purpose of this application is to provide a method, system, medium, program product and terminal for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images, to solve the problems of difficulty in predicting clinical parameters, lack of intuitive two-dimensional simulation output and insufficient experience in implantation schemes in modern orthopedic surgery based on three-dimensional reconstruction and surgical planning using preoperative CT images.

[0008] To achieve the above and other related objectives, the first aspect of this application provides a method for simulating postoperative X-ray images of orthopedic implants based on preoperative CT images, comprising: acquiring preoperative CT images including the pelvis and bilateral femurs and the affected side markers; segmenting each bone voxel region from the preoperative CT images; identifying key bony landmarks in the bone voxel regions; constructing a three-dimensional coordinate system accordingly to obtain three-dimensional bone voxel data; determining the size of the acetabular prosthesis based on the size of the affected acetabulum; matching a suitable acetabular cup assembly to obtain three-dimensional voxel data of the acetabular prosthesis; and performing the simulation based on the bone voxel region segmentation and key point identification results. The optimal femoral stem model in the prosthesis library is selected according to preset conditions to obtain the three-dimensional voxel data of the femoral stem prosthesis on the affected side. The standard posture of the pelvis and both femurs after surgery is set in the three-dimensional coordinate system, and the spatial transformation parameters from the original posture of the preoperative CT image to the standard posture are calculated based on this standard. The healthy femur and the prosthesis on the affected side are transformed to the standard posture accordingly. Under the standard posture, the three-dimensional voxel data of the bone and the three-dimensional voxel data of the prosthesis are fused to form a three-dimensional voxel array. The voxel values ​​of the three-dimensional voxel array are superimposed layer by layer to generate a two-dimensional projection matrix, and a simulated X-ray film under two-dimensional projection is generated accordingly.

[0009] In some embodiments of the first aspect of this application, the process of segmenting the skeletal voxel region from the preoperative CT image, identifying key bony landmarks in the skeletal voxel region, and constructing a three-dimensional coordinate system accordingly includes: using a deep learning segmentation network to identify and label the pelvic voxel region, the healthy femoral voxel region, and the affected femoral voxel region; using a pelvic key point detection network to locate the bilateral lower edge of the teardrop and the midpoint of the pubic symphysis in the pelvic voxel region; constructing a three-dimensional coordinate system with the midpoint of the line connecting the bilateral teardrops as the origin, the direction of the teardrop line as the X-axis, and the Z-axis perpendicular to the pubic symphysis plane and pointing towards the head; and using a femoral key point detection network to locate the femoral landmarks in the healthy femoral voxel region and the affected femoral voxel region.

[0010] In some embodiments of the first aspect of this application, the process of determining the size of the acetabular prosthesis based on the size of the affected acetabulum and matching a suitable acetabular cup assembly includes: determining the acetabular cup model by measuring the size of the affected acetabulum, matching a corresponding liner model based on the acetabular cup model, and selecting the ball head size based on the liner model.

[0011] In some embodiments of the first aspect of this application, the process of selecting the optimal femoral stem model from the prosthesis library according to preset conditions based on the results of bone voxel region segmentation and key point identification includes: locating the femoral head and identifying key points of the greater trochanter based on the femoral medullary cavity segmentation results; sequentially traversing all available femoral stem models in the prosthesis library in three-dimensional space; and selecting the optimal femoral stem prosthesis based on the spatial adaptability verification results and / or biomechanical matching evaluation results.

[0012] In some embodiments of the first aspect of this application, the spatial transformation parameters specifically include: a healthy-side transformation matrix for rotating the healthy femur around the center of the acetabulum; a translation vector for moving the center of the affected femoral head to the center of the acetabular liner; an affected-side rotation matrix for rotating the affected femoral prosthesis around the center of the acetabular liner; the translation vector is used as a parameter to generate a two-dimensional affine transformation matrix, which is then multiplied by the affected-side rotation matrix to generate the affected-side transformation matrix.

[0013] In some embodiments of the first aspect of this application, the process of transforming the healthy femur and the affected prosthesis to a standard posture includes: not applying any spatial transformation to the pelvis, leaving it unchanged; applying a healthy-side transformation matrix to the vertex coordinates of the healthy femur; and applying an affected-side transformation matrix to the vertex coordinates of the affected femur and all attached prostheses.

[0014] In some embodiments of the first aspect of this application, the process of fusing the three-dimensional voxel data of the bone and the three-dimensional voxel data of the prosthesis to generate a three-dimensional voxel array, and then superimposing the voxel values ​​of the three-dimensional voxel array layer by layer to generate a two-dimensional projection matrix, thereby generating a simulated X-ray film under two-dimensional projection, includes: fusing the three-dimensional voxel data of the pelvis, the healthy femur and the prosthesis component of the affected side to generate a complete postoperative lower limb three-dimensional voxel array; scanning the fused three-dimensional voxel array layer by layer along the coronal plane of the three-dimensional coordinate system, superimposing the data of each layer to generate a two-dimensional projection matrix; and normalizing and mapping the two-dimensional matrix to a preset grayscale range to generate a simulated two-dimensional X-ray film.

[0015] To achieve the above and other related objectives, a second aspect of this application provides a postoperative X-ray simulation system for orthopedic implantation based on preoperative CT images, comprising: a modeling module for acquiring preoperative lower limb CT image data of a patient, performing automatic segmentation on the CT images, performing key point recognition based on the segmentation results, and establishing a three-dimensional coordinate system accordingly; a prosthesis selection module for determining the size of the acetabular prosthesis according to the size of the affected acetabulum, matching a suitable acetabular cup assembly to obtain three-dimensional voxel data of the acetabular prosthesis; and selecting prostheses according to preset conditions based on the bone voxel region segmentation and key point recognition results. The optimal femoral stem model from the prosthesis library is used to obtain the three-dimensional voxel data of the femoral stem prosthesis on the affected side; the posture transformation module is used to calculate the spatial transformation parameters from the original posture of the preoperative CT image to the standard posture according to the set postoperative three-dimensional coordinate system standard posture, and accordingly transform the healthy femur and the prosthesis on the affected side to the standard posture; the plain film generation module is used to fuse the three-dimensional voxel data of the bone and the three-dimensional voxel data of the prosthesis to form a three-dimensional voxel array under the standard posture, and to superimpose the voxel values ​​of the three-dimensional voxel array layer by layer to generate a two-dimensional projection matrix, thereby generating a simulated X-ray film under two-dimensional projection.

[0016] To achieve the above and other related objectives, a third aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images.

[0017] To achieve the above and other related objectives, a fourth aspect of this application provides a computer program product comprising computer program code that, when executed on a computer, causes the computer to implement the method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images.

[0018] To achieve the above and other related objectives, a fifth aspect of this application provides an electronic terminal, including a memory, a processor, and a computer program stored in the memory; the processor executes the computer program to implement the method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images.

[0019] As described above, the method, system, medium, program product, and terminal for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images of this application have the following beneficial effects:

[0020] Based on the results of preoperative CT three-dimensional segmentation, key point identification, and prosthesis positioning, this application accurately converts and presents the results into a standard postoperative X-ray simulation effect that conforms to clinical practice. This allows doctors to intuitively observe the morphological projection, specific spatial position, and precise relative anatomical relationship between the selected prosthesis and key bony landmarks of the pelvis and femur on the postoperative X-ray during the preoperative planning stage.

[0021] Based on the generated simulated radiographs, doctors are provided with a basis for quantitative assessment, which effectively assists them in selecting the best surgical plan and adjusting parameters such as implant size, position, and angle, thereby improving the predictability and accuracy of surgical planning and reducing reliance on mere experience. Attached Figure Description

[0022] Figure 1 The diagram shown is a flowchart illustrating a method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images, according to one embodiment of this application.

[0023] Figure 2A The diagram shows a flowchart illustrating the generation of a three-dimensional model of the pelvis and femur using a method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images, as described in one embodiment of this application.

[0024] Figure 2B The diagram shows the entire process of segmenting voxel regions from CT images, identifying key points, and then constructing a three-dimensional coordinate system in one embodiment of this application.

[0025] Figure 3 The diagram shows a flowchart of a method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images in one embodiment of this application, illustrating the prosthesis selection process.

[0026] Figure 4 The diagram shows a flowchart illustrating the postoperative prosthesis posture simulation method based on preoperative CT images and postoperative X-ray simulation of orthopedic implantation in one embodiment of this application.

[0027] Figure 5 The diagram shows a flowchart illustrating the generation of simulated X-ray images in an embodiment of this application, illustrating a method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images.

[0028] Figure 6 The diagram shown is a structural schematic of a postoperative X-ray simulation system for orthopedic implantation based on preoperative CT images, according to one embodiment of this application.

[0029] Figure 7 The diagram shows a schematic of the structure of an electronic terminal for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images, according to an embodiment of this application. Detailed Implementation

[0030] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0031] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," "fixing," and "holding" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0032] It should be noted that, in the embodiments of this application, the words "exemplary" or "for example" indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0033] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0034] In the embodiments of this application, the terms "first" and "second" are used to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that the terms "first" and "second" do not necessarily imply that they are different.

[0035] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the invention.

[0036] Before providing a further detailed description of the present invention, the nouns and terms used in the embodiments of the present invention are explained, and the nouns and terms used in the embodiments of the present invention are subject to the following interpretations:

[0037] <1> Preoperative CT images: These are three-dimensional images of the target area of ​​the patient obtained by computed tomography (CT) scans before surgery, used for diagnosis, evaluation, and surgical planning.

[0038] <2> Teardrop: A teardrop-shaped imaging landmark formed by the overlap of the medial rim of the acetabulum and the bony structures of the obturator foramen on an anteroposterior pelvic X-ray. It is often used for imaging localization and measurement of the hip joint.

[0039] <3> Midpoint of the pubic symphysis: This refers to the geometric center point where the left and right pubic bones connect anteriorly. It is a commonly used reference point in pelvic imaging and surgical localization.

[0040] <4> The acetabulum is a crescent-shaped depression on the hip bones on both sides of the pelvis. It connects to the femoral head to form the hip joint and bears the responsibility of transmitting and distributing body weight.

[0041] <5> Acetabular cup: A metal or polymer hemispherical prosthesis implanted in the acetabulum during total hip replacement surgery to replace the natural acetabulum and work with the femoral head prosthesis to form an artificial hip joint.

[0042] <6> Femoral stem: A rod-shaped prosthetic component implanted in the femoral medullary cavity during total hip replacement surgery. It is used to fix and support the femoral head prosthesis, enabling the transmission and support of lower limb forces.

[0043] <7> Coronal plane: a section in human anatomy, referring to the vertical plane that divides the human body into anterior and posterior parts.

[0044] <8> Femoral shaft axis: refers to the center line extending longitudinally along the femoral shaft, which is an important reference axis for describing femoral posture and for orthopedic measurements and surgical positioning.

[0045] <9> Liner: A liner component placed inside the acetabular cup during artificial hip replacement. It is usually made of polyethylene, ceramic or metal and is used to contact the femoral head prosthesis and reduce friction and wear.

[0046] <10> Hip joint: A ball-and-socket joint formed by the femoral head and the acetabulum, responsible for supporting the lower limbs, movement, and weight transmission.

[0047] <11> Voxel: The smallest unit with a fixed volume in three-dimensional space, equivalent to a three-dimensional pixel, used to represent volume information in medical imaging, 3D reconstruction, and computer graphics.

[0048] <12> Femur: The long bone in the human thigh, it is the thickest and longest bone in the lower limb, connecting the hip joint and the knee joint, and is responsible for support and movement.

[0049] <13> Deep learning segmentation networks: Models that use neural networks to divide images into different regions, thereby achieving pixel-level target recognition and understanding.

[0050] <14> Keypoint detection network: A deep learning model used to automatically locate key positions or landmarks of objects in images or videos.

[0051] <13> Affine transformation matrix: A matrix representation used to perform linear transformations such as translation, rotation, scaling, and shearing on points or graphics in two-dimensional or three-dimensional space, while preserving straight lines and parallel relationships.

[0052] <14> DICOM (Digital Imaging and Communications in Medicine) is an international standard for medical images and related information. It defines a medical image format that meets clinical needs and can be used for data exchange. DICOM data is typically stored in a medical image format, such as CT images and X-ray images.

[0053] To facilitate understanding of the embodiments of this application, firstly, in conjunction with Figure 1 Detailed explanation. Figure 1 This document illustrates a flowchart of a method for simulating postoperative X-ray images of orthopedic implants based on preoperative CT images, according to an embodiment of the present invention. The method for simulating postoperative X-ray images of orthopedic implants based on preoperative CT images in this embodiment mainly includes the following steps:

[0054] Step S11: Obtain preoperative CT images containing the pelvis and bilateral femurs, along with the affected side markers. Segment each bone voxel region from the preoperative CT images, identify key bony landmarks in the bone voxel regions, and construct a three-dimensional coordinate system accordingly to obtain three-dimensional bone voxel data.

[0055] The above step S11 includes the acquisition and preprocessing of data, acquiring the patient's preoperative lower limb CT image data, and the scanning range of the CT image must include the entire pelvic region and the proximal femurs of both sides, and specifying the left or right femur that needs to be replaced as the affected femur.

[0056] It should be understood that the CT image data involved in the embodiments of this application can be in one or more of the following formats: DICOM, NIfTI, RAW, and JPEG. DICOM (Digital Imaging and Communications in Medicine) is an international standard for medical images and related information, defining a medical image format that meets clinical needs and can be used for data exchange. NIfTI (Neuroimaging Informatics Technology Initiative) is a relatively simple three-dimensional image storage format, facilitating rapid loading and analysis by image processing software, and enabling image analysis and segmentation. RAW format is the raw data format of CT images, storing unprocessed scan data, typically containing a large amount of raw signal information, suitable for research on image reconstruction algorithms. JPEG format refers to a lossy compressed image format; although it reduces image quality, it results in small file sizes and fast transmission, suitable for scenarios where high image quality is not required.

[0057] Furthermore, to segment the bone voxel region from the preoperative CT image, identify key bony landmarks within the bone voxel region, and construct a three-dimensional coordinate system accordingly, the specific sub-steps are as follows: Figure 2A As shown:

[0058] Step S11a: Use a deep learning segmentation network to identify and label the pelvic voxel region, the healthy femoral voxel region, and the affected femoral voxel region.

[0059] For example, deep learning segmentation networks used for identifying and labeling voxel regions mainly include: U-Net segmentation network, DeepLab series segmentation networks, Mask R-CNN segmentation network, V-Net segmentation network, FCN segmentation network, or HRNet segmentation network, etc. Among them, the U-Net segmentation network adopts an encoder-decoder structure, extracting features through convolution and pooling, and then restoring resolution through upsampling and convolution to achieve high-precision segmentation. The DeepLab series segmentation networks are semantic segmentation networks based on convolutional neural networks, introducing dilated convolution and multi-scale feature fusion, which can effectively handle objects of different scales and is suitable for medical image segmentation with complex backgrounds and multi-scale targets. The Mask R-CNN segmentation network is a region-based convolutional neural network, combining the object detection capabilities of Faster R-CNN and the segmentation capabilities of fully convolutional networks, which can generate accurate segmentation masks for each instance, suitable for accurate segmentation of multiple targets in medical images. The V-Net segmentation network is a 3D version of the network, also adopting an encoder-decoder structure, optimized for 3D data, and can effectively identify and label 3D regions. Fully Convolutional Network (FCN) is a semantic segmentation network that replaces the fully connected layers of traditional convolutional neural networks with convolutional layers, achieving end-to-end pixel-level segmentation. It can handle input images of arbitrary size and generate segmentation masks of corresponding sizes, making it suitable for large-scale image data segmentation. High-Resolution Network (HRNet) segments images by processing high-resolution and low-resolution features in parallel, preserving image details. In medical image segmentation, it better captures regional details, achieving more accurate segmentation and making it suitable for high-precision segmentation tasks.

[0060] To facilitate understanding by those skilled in the art, the following description uses the U-Net segmentation network as an example to illustrate the specific process of identifying and labeling voxel regions of the pelvis, the healthy femur, and the affected femur: First, the input preoperative lower limb CT images of the patient are preprocessed, including normalization and cropping, to ensure that the images meet the network input requirements. Next, the preprocessed images are input into the U-Net network. The encoder part of the network extracts image features through multi-layer convolution and pooling operations. These features capture the texture and shape information of different regions in the image. Then, the decoder part gradually restores the image resolution through upsampling and convolution operations, while fusing the features extracted by the encoder to generate a high-resolution segmentation mask. During the training phase, the network learns how to distinguish different regions such as the pelvis, the healthy femur, and the affected femur using a large amount of labeled image data. Each region is labeled with a different pixel value in the mask. Finally, the generated segmentation mask is applied to the original image to obtain the labeled voxel regions of the pelvis, the healthy femur, and the affected femur, thereby achieving accurate identification and labeling of these regions. It should be noted that the above explanation only uses the U-Net segmentation network as an example. Although the specific segmentation process of other segmentation models has its own characteristics, the basic principles are similar, so they will not be elaborated here.

[0061] It should be noted that this embodiment simulates postoperative X-ray images of orthopedic implantation. The acquired data includes the pelvis and both femurs. The segmentation results of the bone voxel region usually include the voxel regions of the pelvis and both femurs. A voxel is a "volume pixel", which is the smallest unit in three-dimensional space, just like a pixel in a two-dimensional image. In three-dimensional medical imaging such as CT and MRI, each voxel represents a small cubic region, containing the density or grayscale information of that region. Each voxel unit represents a small volume block of bone in space.

[0062] Step S11b: Use the pelvic key point detection network to locate the lower edge of the bilateral teardrops and the midpoint of the pubic symphysis in the pelvic voxel region; construct a three-dimensional coordinate system with the midpoint of the line connecting the bilateral teardrops as the origin, the direction of the teardrop line as the X-axis, and the direction perpendicular to the pubic symphysis plane pointing to the head as the Z-axis.

[0063] For example, pelvic keypoint detection networks used to identify and label pelvic keypoints mainly include: CPM (Convolutional Pose Machines), Stacked Hourglass Network, Open Pose, HRNet, and TokenPose. CPM progressively optimizes keypoint prediction through layer-by-layer convolution; Stacked Hourglass Network utilizes repeated encoding and decoding of contextual features to achieve multi-scale keypoint localization; OpenPose establishes spatial relationships between joints through Part Affinity Fields when detecting multi-person skeletons; HRNet maintains fine-grained features through a high-resolution parallel network, thereby achieving higher accuracy in keypoint prediction; and TokenPose, based on Transformer, models the relationships between keypoints through a global attention mechanism, improving robustness under complex poses and occlusion conditions. Overall, these networks have evolved from convolutional to attention mechanisms, significantly improving pelvic keypoint detection in terms of accuracy, speed, and adaptability to multiple scenes.

[0064] To facilitate understanding by those skilled in the art, the following description uses the CPM keypoint recognition network as an example to specifically describe its identification and labeling of skeletal keypoints such as the midpoint of the teardrop shape at the lower edge of both sides of the pelvis and the midpoint of the pubic symphysis. Taking CPM as an example, the network generally divides the identification of keypoints on pelvic images into three main stages: First, the input pelvic X-ray or CT image is fed into a convolutional feature extraction network to obtain multi-scale feature maps containing edges, textures, and bony structural contours. In the first stage of prediction, the network generates initial heatmaps for each keypoint based on these features. For example, in a teardrop-shaped high-density image visible on the medial side of both acetabulums, the network forms a high-response region at its lower edge and generates a concentrated probability distribution on the midline of the pubic symphysis image to indicate the midpoint of the pubic symphysis. Subsequently, CPM combines these preliminary heatmaps with the original features... Figure 1 The data is then fed into the next stage of the convolutional module, where the results are optimized by expanding the receptive field and introducing spatial constraints. For example, the position of the midpoint of the lower edge of the teardrop is further corrected by utilizing the left-right symmetry of the pelvis to avoid offset caused by local occlusion or bone abnormalities. Simultaneously, through comprehensive analysis of the pixel gradients at the upper and lower edges of the pubic symphysis, the data gradually converges to the geometric center. After multiple iterations, the final heatmap output by CPM exhibits sharp probability peaks at key points, and the coordinates of the point with the maximum response are used to accurately mark anatomical locations such as the midpoint of the lower edge of the teardrop on both sides of the pelvis and the midpoint of the pubic symphysis. It should be noted that the above explanation only uses the CPM keypoint recognition network as an example. While other recognition algorithms have different specific recognition processes, their basic principles are similar, and the achieved results are the same; therefore, they will not be elaborated upon further.

[0065] Furthermore, the construction process of the three-dimensional coordinate system is as follows: This process establishes a spatial rectangular coordinate system with specific anatomical points of the pelvis as references. First, the midpoints of the lower edges of the teardrops on the medial sides of the left and right acetabulums are extracted, and the midpoint of the line connecting the two points is taken as the origin of the coordinate system. Then, the direction of this line is defined as the X-axis, representing the left-right direction of the pelvis. Next, the midsagittal plane of the pelvis is determined through the plane where the pubic symphysis is located, and the direction perpendicular to this plane and pointing towards the patient's head is taken as the Z-axis, representing the head-to-foot direction. Finally, the Y-axis is determined according to the right-hand rule, that is, when the right hand is extended and the thumb points in the positive direction of the X-axis and the index finger points in the positive direction of the Z-axis, the direction pointed by the middle finger is the positive direction of the Y-axis, representing the anterior-posterior direction. The three-dimensional coordinate system established in this way has a clear anatomical meaning and can serve as a unified reference framework for the spatial positioning and measurement of key points of the pelvis, and determine a relative reference system for subsequent skeletal voxel region division.

[0066] Step S11c: Use the femoral key point detection network to locate the femoral landmarks in the healthy femoral voxel region and the affected femoral voxel region.

[0067] The detection algorithms used to identify and label femoral keypoints mainly include CPM, Stacked Hourglass Network, HRNet, and Transformer-based keypoint detection methods such as TokenPose. CPM progressively optimizes keypoint coordinates through multi-stage convolution, achieving coarse-to-fine prediction; Stacked Hourglass Network repeatedly encodes and decodes multi-scale contextual information to improve keypoint localization accuracy; HRNet preserves skeletal details through parallel high-resolution features, improving keypoint recognition accuracy and robustness; and Transformer methods model the spatial relationships between keypoints through a global attention mechanism, enhancing robustness to complex poses or partial occlusion. Overall, these algorithms have evolved from convolution to attention mechanisms, achieving a balance between speed, accuracy, and adaptability to complex scenes, providing a reliable technical means for the automated identification of femoral keypoints. The specific femoral keypoint detection process is similar to the S11b pelvic keypoint detection process described above and will not be explained in detail here.

[0068] It is important to note that femoral landmarks are crucial anatomical reference points used to describe the geometry and spatial positioning of the femur. The femoral head center point refers to the geometric center of the femoral head, typically obtained by fitting a sphere or ellipsoid, and is used to characterize the center of the hip joint. The medullary canal axis, formed by connecting the centers of the medullary canals within the femoral shaft, represents the longitudinal direction of the femoral shaft and is used to guide axial alignment in orthopedic surgery, prostheses, and joint replacements. These landmarks provide a unified reference framework for radiographic measurements, 3D reconstruction, and surgical planning, enabling the accurate quantification and standardization of the femoral length, angles, and spatial orientation.

[0069] It should also be noted that the segmentation of different bone voxel regions and the identification of key points in this application are for postoperative prosthesis implantation simulation and are not affected or interfered with by the different internal methods and steps of voxel region segmentation and key point identification. Therefore, existing voxel region segmentation and key point identification methods, as well as new methods that may emerge in the future, can all be applied to this application, and no specific limitations are made here.

[0070] For ease of understanding, this embodiment further incorporates Figure 2B The explanation demonstrates the entire process of segmenting voxel regions from CT images and then constructing a three-dimensional coordinate system. The specific process includes the following:

[0071] The process corresponding to step S11a above is as follows: First, input the preoperative CT data; second, use a deep learning segmentation network to perform bone voxel recognition on the input preoperative CT data, identifying the pelvic voxel region, left femoral voxel region, and right femoral voxel region respectively. Accordingly, assign a label to each voxel region, that is, assign a pelvic voxel label to the pelvic voxel region, a left femoral voxel label to the left femoral voxel region, and a right femoral voxel label to the right femoral voxel region.

[0072] The process corresponding to step S11b above is as follows: First, input the pelvic voxel data obtained based on a deep learning segmentation network. Second, use a pelvic keypoint detection network to identify pelvic landmarks and obtain the positional information of landmarks such as the lower edge of the bilateral teardrops and the midpoint of the pubic symphysis. Finally, construct a three-dimensional coordinate system of the pelvis based on the obtained landmark positional information. The origin of the coordinate system is the midpoint of the line connecting the bilateral teardrops, the X-axis is the direction of the teardrop line, the Z-axis is perpendicular to the pubic symphysis plane, and the Y-axis is determined by the right-hand rule.

[0073] The process corresponding to step S11c above is as follows: First, input the left and right femoral voxel data obtained based on the deep learning segmentation network. Second, use the femoral key point detection network to identify femoral landmarks and obtain the position information of the femoral head center point and the medullary canal axis, respectively.

[0074] It should be noted that the method for constructing the three-dimensional coordinate system is not fixed, including but not limited to different coordinate axis orientations based on the same key point, different key point positions based on the same coordinate axis orientation, and different key points based on different coordinate axis orientations. The selection of key points can be determined according to different key point recognition needs, and the orientation of the three-dimensional coordinate system coordinate axes can also be selected differently as needed. No specific limitations are made here.

[0075] Furthermore, as the smallest unit in a three-dimensional coordinate system, a voxel contains spatial coordinate position information and attribute information such as grayscale value and density. The size of the voxel determines the resolution of the overall skeleton voxel region. The size of a single voxel can be flexibly determined according to the needs of different usage scenarios and hardware conditions, and no specific limitation is made here.

[0076] Step S12: Determine the size of the acetabular prosthesis based on the size of the affected acetabulum, and match the appropriate acetabular cup assembly to obtain the three-dimensional voxel data of the acetabular prosthesis; based on the bone voxel region segmentation and key point recognition results, select the optimal femoral stem model from the prosthesis library according to preset conditions to obtain the three-dimensional voxel data of the affected femoral stem prosthesis.

[0077] It's important to note that total hip arthroplasty is a surgical procedure that replaces a diseased hip joint with an implanted artificial prosthesis. The core of this procedure lies in the rational design of the prosthesis structure. The prosthesis typically consists of four parts: the femoral stem is implanted into the femoral medullary cavity for fixation and weight-bearing; the femoral head is mounted at the top of the femoral stem as the articular ball; the acetabular cup is embedded in the acetabulum as a load-bearing shell; and the liner is fitted inside the acetabular cup and contacts the femoral head to reduce friction. These four parts work together: the femoral stem is fixed within the femur; the femoral head acts as the articular ball; the acetabular cup is embedded in the acetabulum; and the liner reduces friction and forms the articular surface, ultimately enabling the joint to bear weight and perform its movement functions. In clinical practice, the choice of prosthesis size must be individualized based on the patient's anatomical characteristics of the acetabulum and femur to ensure joint stability, smooth movement, and prolong the prosthesis's lifespan.

[0078] In this embodiment, the implanted prosthesis consists of two parts in the selection process: an acetabular prosthesis and a femoral stem prosthesis. The acetabular prosthesis specifically comprises an acetabular cup, a liner, and a ball head. The acetabular cup, liner, ball head, and femoral stem prosthesis together constitute the prosthesis implanted in total hip arthroplasty. The selection process for the acetabular cup assembly and the femoral stem prosthesis is as follows: Figure 3 As shown, the process is as follows:

[0079] 1. Selection process for acetabular prosthesis: First, based on the segmented femoral voxel region of the affected side as described above and the constructed three-dimensional coordinate system, generate three-dimensional voxel data of the affected femur. Determine the acetabular cup model by measuring the size of the affected acetabulum. Select the liner model corresponding to the selected acetabular cup model. Finally, select the ball head model with a suitable size based on the liner model.

[0080] It should be noted that in human anatomy, the acetabulum is a depression in the pelvis, and the femoral head is a spherical structure at the proximal end of the femur, forming a ball-and-socket joint with the acetabulum. The structure and function of the prosthesis are as follows: the acetabular cup is implanted into the acetabular fossa, forming a new joint socket; the liner is placed inside the acetabular cup, with one side contacting the acetabular cup, acting as the articular cartilage of the prosthesis, and the other side contacting the ball-and-socket prosthesis, providing a smooth articular surface and reducing friction and wear; the ball-and-socket prosthesis replaces the original femoral head and is fixed to the proximal femur via the femoral stem. The entire prosthesis assembly together replaces the natural acetabular-femoral head joint, restoring the function of the ball-and-socket joint.

[0081] Furthermore, to ensure a proper fit between the various components of the acetabular prosthesis, the prosthesis selection process follows a chain-matching sequence: acetabular cup, liner, and ball head. This standardized sequence guarantees a proper fit between the components and ensures a good fit between the acetabular prosthesis and the human body, thus guaranteeing matching accuracy, reducing the risk of prosthesis mismatch, and optimizing joint stability.

[0082] 2. Femoral stem selection process: First, based on the three-dimensional voxel data of the affected femur segmented in the above process, obtain the three-dimensional data of the femoral medullary cavity. Then, based on the key point identification of the greater trochanter, traverse all femoral stem models in the prosthesis library. Determine the spatial adaptability of the femoral stem by judging whether the prosthesis stem is fully in place and without cortical bone penetration. Comprehensively evaluate parameters such as prosthesis axis alignment and medullary cavity filling rate to determine biomechanical matching and select the optimal model.

[0083] It should be noted that spatial fit and biomechanical fit are two independent parameters. In determining the optimal femoral stem type, one can choose either the highest biomechanical fit or the highest spatial fit, or a hyperparameter can be used to control the weights of both. The choice of hyperparameter can be based on the patient's specific situation. Considering different bone densities, CT image resolutions, and segmentation accuracies, the choice of hyperparameter will adapt to different patient conditions. For example, when CT image resolution and segmentation accuracy are higher, it is beneficial for accurately reconstructing the geometry of the femoral medullary cavity, resulting in higher reliability of spatial fit. In this case, the weight of spatial fit should be appropriately increased. When bone density is higher, the weight of biomechanical fit should be appropriately increased to prevent stress shielding and maintain long-term stability. The specific weight selection method needs to be flexibly adjusted according to the patient's actual situation; the details of various possible divisions will not be elaborated here.

[0084] S13: Set the standard posture of the pelvis and bilateral femurs in the three-dimensional coordinate system after surgery, and calculate the spatial transformation parameters from the original posture of the preoperative CT image to the standard posture based on this standard, and transform the healthy femur and the prosthesis of the affected side to the standard posture accordingly.

[0085] It is worth noting that the standard postoperative posture of the pelvis and both femurs in the three-dimensional coordinate system is one of the most commonly used positions and projection methods in orthopedics and radiology, especially in hip replacement surgery and pelvic fracture assessment. In this embodiment, the specific posture is as follows: First, the patient lies supine or standing with their back pressed against the probe; the pelvis is in a standard anteroposterior position, meaning the pelvic plane is parallel to the coronal plane, without rotation or tilt; both femurs are in a natural, neutral, extended position, i.e., the femoral shaft axis is vertically downward in the coronal plane. A standard anteroposterior pelvic X-ray is an important basis for measuring the acetabular angle, lower limb length difference, and prosthesis selection, clearly showing the overall condition of the pelvis, ischium, ilium, and acetabulum.

[0086] In this embodiment of the invention, the three-dimensional spatial transformation parameters of the affected femur and the healthy femur from the original posture of the preoperative CT image to the standard posture are calculated.

[0087] It is worth noting that the computational space transformation parameters and application parameter transformations described in the embodiments of this invention currently face the following limitations for orthopedic surgeons in modern medical fields, especially in hip replacement surgery: (1) Disconnection between three-dimensional views and clinical gold standards: Although three-dimensional planning can provide a stereoscopic perspective, its presentation method differs significantly from the gold standard that orthopedic surgeons rely on for postoperative evaluation—standard two-dimensional X-ray films. Doctors are accustomed to evaluating the final position, shape, and anatomical relationship with bony landmarks, such as the tilt angle, anteversion angle, limb length, and eccentricity of the implant, on anteroposterior radiographs, usually pelvic anteroposterior radiographs. (2) Lack of intuitive two-dimensional simulation output: Surgical planning systems based on existing three-dimensional models cannot generate simulated postoperative X-ray views that conform to clinical practice and have diagnostic value. The planning results are difficult to directly map onto the radiograph perspective that doctors are most familiar with and use for final effect interpretation. (3) Difficulty in predicting key clinical parameters: Due to the lack of accurate simulation radiographs, doctors cannot intuitively and quantitatively predict the specific shape of the selected prosthesis and its placement on the postoperative standard X-ray, the precise spatial position relative to key bony landmarks, and the interrelationship between various prosthesis components, such as the coverage relationship of the ball head in the acetabular cup and the axis of the femoral stem, etc.

[0088] In view of this, the main purpose of the three-dimensional spatial transformation from the original posture of the preoperative CT image to the standard posture is to generate X-rays that can be used by orthopedic surgeons and their clinical experience to judge the preoperative surgical plan of orthopedic surgery. By providing intuitive and quantitative simulation effects of postoperative X-rays, it can effectively assist doctors in optimizing the selection of surgical plans and setting implant parameters (adjusting size, position, and angle), improve the predictability and accuracy of surgical planning, and reduce reliance on mere experience.

[0089] Specifically, such as Figure 4 As shown, the posture transformation calculation module is divided into two parts: the affected side path and the healthy side path.

[0090] It's important to note that the posture changes on the affected side differ from those on the healthy side. This is because the femoral head on the healthy side is naturally located within the acetabular fossa, and its anatomical center is the true center of rotation for joint movement. Since the anatomically, the center of the femoral head coincides with the acetabular center, no additional translation is needed. However, on the affected side, in the three-dimensional coordinate system, it has been replaced with an artificial prosthesis. The femoral stem and femoral head may differ due to various factors, such as the femoral head not being in the precise original position of the femoral neck, differences in surgical implantation angle and position, and the prosthesis's femoral head center not being equal to the acetabular prosthesis liner center. These factors will not be listed here. If rotation were directly performed at the prosthesis's femoral head center, the femoral head would "disengage" from the acetabular liner. While direct rotation is possible, it does not conform to the actual biomechanical kinematics of the joint. Therefore, to ensure proper contact between the prosthesis and the acetabulum, accurately simulate hip joint movement, and achieve a reasonable matching posture in three-dimensional space, translational correction is necessary before rotation.

[0091] Furthermore, the two-part process, the affected side pathway and the healthy side pathway, is detailed below:

[0092] (1) The rotation transformation of the healthy side is as follows: First, the coordinates of the healthy femur and all its vertices in three-dimensional coordinates are rotated around the center of the healthy femoral head joint, with the center of the healthy femoral head joint as the rotation center point, so that the axis of the healthy femoral shaft is vertically downward in the coronal plane coordinate system of the pelvis. The position of the center of the healthy femoral head remains unchanged before and after the transformation. Second, the transformed positions of the healthy femur and all its vertices are recorded and combined to generate the healthy side transformation matrix T_healthy_femur_rotation, which only contains rotation.

[0093] (2) The affected side rotation transformation is as follows: First, the center of the affected femoral head is translated from its current position to the rotation center of the affected acetabular liner. The translation vector V_translation is calculated to move the center of the affected femoral head from its current position to coincide with the rotation center of the affected acetabular liner. Then, the coordinates of all pseudofemoral stems, femoral heads, and all vertices of the affected femur in three-dimensional coordinates are taken as the center of rotation point of the hip joint after the translation of the affected femoral head, i.e., the new center of rotation. Rotation is performed around the new center of rotation, so that the axis of the affected femoral shaft is vertically downward in the coronal plane coordinate system of the pelvis. The position of the center of the healthy femoral head remains unchanged before and after the transformation. Finally, the transformed positions of the healthy femur and all its vertices are recorded and combined to generate the affected side rotation matrix T_affected_femur_rotation, which only contains rotation. The translation vector V_translation and the affected side rotation matrix T_affected_femur_rotation are combined to generate the affected side transformation matrix T_affected_femur, which contains rotation and translation. The formula for calculating T_affected_femur is as follows:

[0094] T_affected_femur=T_affected_femur_rotation×[I|V_translation] Formula (1)

[0095] In formula (1), [I|V_translation] is the affine transformation matrix containing the translation vector V_translation, I is the identity matrix, and V_translation = (t x ,t y ,t z ) T The specific expanded form of [I|V_translation] is:

[0096]

[0097] Furthermore, based on the generated healthy-side transformation matrix T_healthy_femur_rotation and affected-side transformation matrix T_affected_femur, a transformation of the three-dimensional voxel model is performed. For the pelvis and acetabular prosthesis, since they are fixed together and the pelvis needs to maintain the stability of the three-dimensional coordinate system, the positions of the pelvis and acetabular prosthesis remain unchanged. The healthy-side transformation matrix T_healthy_femur_rotation is applied to the healthy-side femur, and the affected-side transformation matrix T_affected_femur is applied to the affected-side femur and prosthesis. Finally, the three parts are obtained: the pelvis that remains in place, the healthy-side femur that remains vertical with the acetabular center coinciding, and the affected-side femur that precisely coincides with the ball head center.

[0098] S14: Under the standard posture, the three-dimensional voxel data of the skeleton and the three-dimensional voxel data of the prosthesis are fused to form a three-dimensional voxel array. The voxel values ​​of the three-dimensional voxel array are superimposed layer by layer to generate a two-dimensional projection matrix, thereby generating a simulated X-ray film under two-dimensional projection.

[0099] Specifically, after step S13, the coordinate information of three regions—the pelvic voxel region, the healthy femur voxel region, and the affected femur voxel region—is obtained. At this point, the three regions are in a standard pose and are independent of each other. The final simulated X-ray image generation is then completed in three stages: a three-dimensional voxel data fusion stage, a coronal projection calculation stage, and a two-dimensional image generation stage. The process for each stage is as follows: Figure 5 As shown, the specific process is as follows:

[0100] First, in the 3D voxel data fusion stage, based on the standard posture, voxel data of the pelvis, bilateral femurs, acetabular cup, liner, ball head, and femoral stem prosthesis components are integrated to generate a unified fused voxel array, providing a spatially continuous anatomical model for projection calculations. Then, in the coronal projection calculation stage, the integrated 3D voxel array is scanned layer by layer along the coronal plane. For example, for each coordinate point in the 3D coordinate system, several S(x, z) values ​​are obtained by superimposing all (x, z) coordinate points with the same Y-axis coordinate along the Y-axis direction. The calculation method for S(x, z) is as follows:

[0101]

[0102] Where V(x,y,z) represents the value of a voxel in the 3D volume data, represents the density or intensity at spatial location (x,y,z), S(x,z) represents the accumulated 2D projection result, ∑ y This indicates that voxel values ​​are accumulated along the Y-axis. Then, a one-step normalization process is performed on all S(x,z) values, mapping them to a given grayscale range. In medical image analysis, voxel value normalization is a crucial preprocessing step. Its main purpose is to eliminate inconsistencies in intensity caused by different scanning devices, imaging parameters, and individual patient differences, making the grayscale values ​​of the same anatomical structure comparable in different images. It also maps the input data to a suitable numerical range. Common normalization methods include: min-max normalization, which linearly scales voxel values ​​to the 0,1 or -1,1 range, often used to unify different... The scanning dynamic range; Z-score normalization, by subtracting the mean of voxel intensity and dividing by the standard deviation, transforms the data into a zero-mean, unit-variance distribution, thus highlighting the relative differences between structures; quantile or percentile normalization, which truncates voxel values ​​according to the upper and lower percentiles of the distribution and linearly scales them, is used to suppress extreme noise or outliers; logarithmic normalization or histogram equalization is often used to improve image contrast and make details clearer. Through these methods, normalization not only improves the comparability between different image data, but also provides more stable input for tasks such as segmentation, registration, and 3D reconstruction. Figure 5 As shown, for example, all S(x,z) are mapped to the interval 0-255, and a linear normalization method is used. The calculation method is as follows:

[0103] Gray=(S-min) / (max-min)×255 Formula (4)

[0104] Where Gray is the normalized S(x,z) value used to generate the X-ray, S is the current stacked voxel value, and min and max are the minimum and maximum voxel values ​​under the same Y coordinate, respectively.

[0105] It should be noted that the normalization process shown in formula (4) can also adopt any other normalization method, and the choice can be made according to the actual situation. No specific limitation is made here. The purpose of normalization is to realize the density accumulation effect and finally realize the transformation from three-dimensional image to two-dimensional image. The density accumulation effect simulation generally refers to the study of the density change of a certain substance or structure in three-dimensional space as a result of the path or superposition process. It is used to describe the attenuation of rays when passing through the medium, the overall strength after the material is stacked, or the volume density accumulation reflected by the voxel grayscale projection in medical images. Its core idea is: to traverse the voxels or material units layer by layer along a certain direction, and to integrate or superimpose the density or attenuation coefficient of each layer to obtain the cumulative effect in that direction. For example, in CT imaging, the linear attenuation coefficients of different voxels will gradually accumulate when X-rays pass through the human body to form projection data; in three-dimensional reconstruction or simulation, this density accumulation can also be approximated by the summation or exponential superposition of the voxel array on a specified axis. In this way, researchers can simulate the energy loss of rays passing through tissues of different thicknesses, or estimate the overall density distribution of multilayer material composites, thereby providing numerical support for image reconstruction, dosing calculation, or material design.

[0106] Finally, the two-dimensional image generation stage converts the normalized voxel data into a two-dimensional projection matrix, and outputs a simulated X-ray film that meets clinical diagnostic standards through linear grayscale mapping, thus completing the conversion from three-dimensional space to two-dimensional image.

[0107] like Figure 6 The diagram illustrates a structural schematic of a postoperative X-ray simulation system for orthopedic implantation based on preoperative CT images, according to an embodiment of the present invention. The X-ray simulation system 600 in this embodiment includes the following modules: a modeling module 601, a prosthesis selection module 602, a posture transformation module 603, and a plain film generation module 604.

[0108] The modeling module 601 is used to acquire the patient's preoperative lower limb CT image data, perform automatic segmentation on the CT images, perform key point recognition based on the segmentation results, and establish a three-dimensional coordinate system accordingly.

[0109] The prosthesis selection module 602 is used to determine the size of the acetabular prosthesis based on the size of the affected acetabulum and match the appropriate acetabular cup assembly to obtain the three-dimensional voxel data of the acetabular prosthesis; based on the bone voxel region segmentation and key point recognition results, the optimal femoral stem model in the prosthesis library is selected according to preset conditions to obtain the three-dimensional voxel data of the femoral stem prosthesis on the affected side.

[0110] The posture transformation module 603 is used to calculate the spatial transformation parameters from the original posture of the preoperative CT image to the standard posture according to the set postoperative three-dimensional coordinate system standard posture, and accordingly transform the healthy femur and the affected prosthesis to the standard posture.

[0111] The flat film generation module 604 is used to fuse the three-dimensional voxel data of the skeleton and the three-dimensional voxel data of the prosthesis to form a three-dimensional voxel array under the standard posture, and to superimpose the voxel values ​​of the three-dimensional voxel array layer by layer to generate a two-dimensional projection matrix, thereby generating a simulated X-ray film under two-dimensional projection.

[0112] It should be understood that the specific process of each module performing the above-mentioned steps has been described in detail in the above method embodiments, and will not be repeated here for the sake of brevity.

[0113] It should also be understood that the module division in the embodiments of this application is illustrative and only represents a logical functional division; in actual implementation, there may be other division methods. Furthermore, the functional modules in the various embodiments of this application can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0114] Figure 7 This is a schematic block diagram of the electronic terminal provided in an embodiment of this application. Figure 7 As shown, the electronic terminal includes at least one processor 701, a memory 702, at least one network interface 703, and a user interface 705. The various components in the device are coupled together via a bus system 704. It is understood that the bus system 704 is used to implement communication between these components. In addition to a data bus, the bus system 704 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 7 The general will label all buses as bus systems.

[0115] The user interface 705 may include a monitor, keyboard, mouse, trackball, clicker, button, touchpad, or touch screen.

[0116] It is understood that memory 702 can be volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM) or programmable read-only memory (PROM), used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM) and synchronous static random access memory (SSRAM). The memories described in the embodiments of this invention are intended to include, but are not limited to, these and any other suitable categories of memory.

[0117] In this embodiment of the invention, the memory 702 is used to store various types of data to support the operation of the electronic terminal 700. Examples of this data include: any executable program for operation on the electronic terminal 700, such as the operating system 7021 and application program 7022; the operating system 7021 contains various system programs, such as the framework layer, core library layer, driver layer, etc., for implementing various basic services and handling hardware-based tasks. The application program 7022 may contain various applications, such as a media player, browser, etc., for implementing various application services. The implementation of the automatic farmland irrigation method based on region division provided in this embodiment of the invention can be included in the application program 7022.

[0118] The methods disclosed in the above embodiments of the present invention can be applied to or implemented by processor 701. Processor 701 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in processor 701 or by instructions in software form. The processor 701 may be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 701 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. General-purpose processor 701 may be a microprocessor or any conventional processor, etc. The steps of the accessory optimization method provided in the embodiments of the present invention can be directly reflected as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a storage medium, which is located in memory. The processor reads the information in the memory and combines it with its hardware to complete the steps of the aforementioned method.

[0119] In an exemplary embodiment, the electronic terminal 700 may be used by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), or complex programmable logic devices (CPLDs) to execute the aforementioned method.

[0120] According to the method provided in the embodiments of this application, this application also provides a computer program product, which includes: computer program code, which, when run on a computer, causes the computer to execute a method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images according to any of the embodiments shown.

[0121] According to the method provided in the embodiments of this application, this application also provides a computer-readable storage medium storing program code, which, when run on a computer, causes the computer to perform a method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images according to any of the embodiments shown.

[0122] As used in this specification, the terms "component," "module," "system," etc., are used to refer to computer-related entities, hardware, firmware, combinations of hardware and software, software, or software in execution. For example, a component can be, but is not limited to, a process running on a processor, a processor, an object, an executable file, an execution thread, a program, and / or a computer. As illustrated, applications running on computing devices and computing devices can both be components. One or more components may reside in a process and / or an execution thread, and components may be located on a single computer and / or distributed among two or more computers. Furthermore, these components can be executed from various computer-readable media on which various data structures are stored. Components can communicate, for example, via local and / or remote processes based on signals having one or more data packets (e.g., data from two components interacting with another component between a local system, a distributed system, and / or a network, such as the Internet interacting with other systems via signals).

[0123] Those skilled in the art will recognize that the various illustrative logical blocks and steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0124] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0125] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and 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. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0126] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0127] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0128] In the above embodiments, the functions of each functional unit can be implemented entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. A computer program product includes one or more computer instructions (programs). When the computer program instructions (programs) are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., high-density digital video discs, DVDs), or semiconductor media (e.g., solid-state disks, SSDs, etc.).

[0129] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0130] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0131] In summary, this application provides a method, system, medium, program product, and terminal for simulating postoperative X-ray images of orthopedic implants based on preoperative CT images. It accurately converts and presents the results of three-dimensional segmentation, key point identification, and prosthesis positioning based on preoperative CT into a standard postoperative X-ray simulation effect that conforms to clinical practice. This allows surgeons to intuitively and clearly "foresee" the morphological projection, specific spatial position, and precise relative anatomical relationship between the selected prosthesis (including the acetabular cup, liner, ball head, and femoral stem) and key bony landmarks of the pelvis and femur on a standard postoperative anteroposterior X-ray during the preoperative planning stage. By providing intuitive and quantifiable postoperative X-ray simulation effects, it effectively assists surgeons in selecting the optimal surgical plan and adjusting parameters such as implant size, position, and angle, improving the predictability and accuracy of surgical planning and reducing reliance on mere experience. Therefore, this application effectively overcomes the various shortcomings of existing technologies and has high industrial applicability.

[0132] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A method for simulating postoperative X-ray images of orthopedic implants based on preoperative CT images, characterized in that, include: Preoperative CT images of the pelvis and bilateral femurs and the affected side markers are acquired. Each bone voxel region is segmented from the preoperative CT images, key bony landmarks in the bone voxel regions are identified, and a three-dimensional coordinate system is constructed accordingly to obtain three-dimensional bone voxel data. The size of the acetabular prosthesis is determined based on the size of the affected acetabulum, and a suitable acetabular cup assembly is matched to obtain three-dimensional voxel data of the acetabular prosthesis. Based on the results of bone voxel region segmentation and key point recognition, the optimal femoral stem model in the prosthesis library is selected according to preset conditions to obtain three-dimensional voxel data of the affected femoral stem prosthesis. The standard posture of the pelvis and bilateral femurs after surgery is set in the three-dimensional coordinate system. Based on this standard, the spatial transformation parameters from the original posture of the preoperative CT image to the standard posture are calculated, and the healthy femur and the prosthesis of the affected side are transformed to the standard posture accordingly. Under the standard posture, the three-dimensional voxel data of the skeleton and the three-dimensional voxel data of the prosthesis are fused to form a three-dimensional voxel array. The voxel values ​​of the three-dimensional voxel array are superimposed layer by layer to generate a two-dimensional projection matrix, thereby generating a simulated X-ray film under two-dimensional projection.

2. The method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images according to claim 1, characterized in that, The process of segmenting the bone voxel region from the preoperative CT image, identifying key bony landmarks within the bone voxel region, and constructing a three-dimensional coordinate system accordingly includes: A deep learning segmentation network was used to identify and label the pelvic voxel region, the healthy femur voxel region, and the affected femur voxel region. The lower edge of the bilateral teardrops and the midpoint of the pubic symphysis in the pelvic voxel region are located using a pelvic key point detection network; a three-dimensional coordinate system is constructed with the midpoint of the line connecting the two teardrops as the origin, the direction of the teardrop line as the X-axis, and the Z-axis perpendicular to the pubic symphysis plane and pointing to the head as the Z-axis. The femoral landmarks in the healthy and affected femoral voxel regions were located using a femoral key point detection network.

3. The method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images according to claim 1, characterized in that, The process of determining the acetabular prosthesis size based on the size of the affected acetabulum and matching a suitable acetabular cup assembly includes: The acetabular cup model is determined by measuring the size of the affected acetabulum, the corresponding liner model is matched based on the acetabular cup model, and the ball head size is selected based on the liner model.

4. The method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images according to claim 1, characterized in that, The process of selecting the optimal femoral stem model from the prosthesis library based on the results of bone voxel region segmentation and key point recognition according to preset conditions includes: Based on the femoral medullary cavity segmentation results, the femoral head is located and key points of the greater trochanter are identified. All available femoral stem models in the prosthesis library are sequentially traversed in three-dimensional space. The optimal femoral stem prosthesis is selected based on the spatial adaptability verification results and / or biomechanical matching evaluation results.

5. The method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images according to claim 1, characterized in that, The spatial transformation parameters specifically include: The transformation matrix of the healthy femur rotating around the center of the acetabulum; The translational vector from the center of the affected acetabular head to the center of the acetabular liner; The rotation matrix of the affected femoral prosthesis around the center of the acetabular liner; The translation vector is used as a parameter to generate a two-dimensional affine transformation matrix, which is then multiplied by the affected side rotation matrix to generate the affected side transformation matrix.

6. The method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images according to claim 5, characterized in that, The process of transforming the healthy femur and the affected prosthesis to a standard position includes: No spatial transformation should be applied to the pelvis; it should remain unchanged. Apply the healthy side transformation matrix to the vertex coordinates of the healthy femur; Apply the affected side transformation matrix to the vertex coordinates of the affected femur and all attached prostheses.

7. The method for simulating postoperative X-ray images of orthopedic implantation based on preoperative CT images according to claim 1, characterized in that, The process of fusing the 3D voxel data of the skeleton and the 3D voxel data of the prosthesis to generate a 3D voxel array, and then layering the voxel values ​​of the 3D voxel array to generate a 2D projection matrix, thereby generating a simulated X-ray image under 2D projection, includes: The three-dimensional voxel data of the pelvis, the healthy femur and the prosthesis component on the affected side are fused to generate a complete three-dimensional voxel array of the lower limb after surgery; The fused three-dimensional voxel array is scanned layer by layer along the coronal plane of the three-dimensional coordinate system, and the data of each layer are superimposed to generate a two-dimensional projection matrix; The two-dimensional matrix is ​​normalized and mapped to a preset grayscale range, thereby generating a simulated two-dimensional X-ray film.

8. A simulation system for postoperative X-ray images of orthopedic implants based on preoperative CT images, characterized in that, The system includes: The modeling module is used to acquire the patient's preoperative lower extremity CT image data, perform automatic segmentation on the CT images, perform key point recognition based on the segmentation results, and establish a three-dimensional coordinate system accordingly. The prosthesis selection module determines the size of the acetabular prosthesis based on the size of the affected acetabulum and matches the appropriate acetabular cup assembly to obtain three-dimensional voxel data of the acetabular prosthesis. Based on the bone voxel region segmentation and key point recognition results, the optimal femoral stem model in the prosthesis library is selected according to preset conditions to obtain three-dimensional voxel data of the affected femoral stem prosthesis. The posture transformation module is used to calculate the spatial transformation parameters from the original posture of the preoperative CT image to the standard posture according to the set postoperative three-dimensional coordinate system standard posture, and accordingly transform the healthy femur and the affected prosthesis to the standard posture. Flat film generation module: used to fuse the three-dimensional voxel data of the skeleton and the three-dimensional voxel data of the prosthesis to form a three-dimensional voxel array under the standard posture, and to superimpose the voxel values ​​of the three-dimensional voxel array layer by layer to generate a two-dimensional projection matrix, thereby generating a simulated X-ray film under two-dimensional projection.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the postoperative X-ray simulation method for orthopedic implantation based on preoperative CT images as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes computer program code that, when run on a computer, causes the computer to implement the postoperative X-ray simulation method for orthopedic implantation based on preoperative CT images as described in any one of claims 1 to 7.

11. An electronic terminal, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the method for simulating postoperative X-ray images of orthopedic implants based on preoperative CT images as described in any one of claims 1 to 7.

Images