Automatic extraction method and related apparatus for bone tumor boundaries including capsule and reaction zone

By generating and expanding the capsule boundary model on bone tumor CT images, and combining it with surface intersection technology, the boundary of bone tumors containing the capsule and reaction zone is automatically extracted, which solves the problem of difficult boundary identification in traditional bone tumor resection surgery and achieves precise R0 resection and minimal damage.

CN119090910BActive Publication Date: 2026-11-17BEIHANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411154441.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-22
Publication Date
2026-11-17
Estimated Expiration
2044-08-22

AI Technical Summary

Technical Problem

Traditional bone tumor resection surgery has difficulty accurately identifying the reaction zone of bone tumors, making R0 resection difficult and potentially leading to tumor recurrence or functional impairment.

Method used

By acquiring CT images of the bone containing the bone tumor, a trained segmentation model is used to generate the boundary of the bone tumor capsule and a healthy bone surface model. The capsule boundary is then extended outward, and combined with surface intersection technology, the boundary of the bone tumor containing the capsule and the reaction zone is automatically extracted.

Benefits of technology

It achieves precise R0 resection of bone tumors, reduces damage to surrounding normal tissues, provides minimal surgical boundaries, and ensures the safety and effectiveness of the surgery.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119090910B_ABST
    Figure CN119090910B_ABST
Patent Text Reader

Abstract

This application discloses an automatic extraction method and related apparatus for bone tumor boundaries including a capsule and a reaction zone, relating to the field of medical image data processing technology. Using CT images obtained from scanning the bone containing the bone tumor as input, a bone tumor capsule boundary model and a healthy bone surface model are generated using a trained first segmentation model and a trained second segmentation model, respectively. The bone tumor capsule boundary model is expanded outward to obtain a preliminary bone tumor boundary model including the capsule and the reaction zone. The preliminary bone tumor boundary model and the healthy bone surface model are subjected to surface intersection calculation. The portion of the preliminary bone tumor boundary model located outside the healthy bone surface model is removed, resulting in a bone tumor boundary model including the capsule and the reaction zone, which facilitates precise R0 resection of the bone tumor.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of medical image data processing technology, and in particular to a method and related apparatus for automatically extracting the boundary of a bone tumor, including a capsule and a reaction zone. Background Technology

[0002] Bone tumors are abnormal tissue masses or clumps formed by the uncontrolled division of bone cells. They can affect any bone in the body and can develop anywhere from the bone surface to the bone marrow. Bone tumors can weaken or destroy healthy bone tissue, making it more prone to fractures. Surgical removal of bone tumors is the primary treatment. The resection margin of bone tumors directly affects the quality of the surgery. Inadequate resection may lead to tumor recurrence, secondary metastasis, amputation, or even death, while excessive resection can easily lead to functional impairment, prosthesis loosening, or even secondary surgery. Therefore, accurate and safe resection of bone tumors is the focus of clinical treatment.

[0003] According to the International Union for Cancer Control (UICC), R0 (completely tumor-free margin) resection is used to indicate the complete removal of a tumor. Precise R0 resection is increasingly sought after to achieve it with minimal tissue damage. However, traditional R0 resections are difficult to perform precisely due to the challenge of identifying the resection margin. Therefore, determining the minimally invasive surgical margins required for complete tumor removal is crucial for achieving precise R0 resection.

[0004] Currently, automated surgical path planning methods for bone tumor resection mainly rely on the direct automatic identification of anatomically defined bone tumor features, primarily the primary lesion and capsule of the bone tumor. In other words, current bone tumor boundaries are primarily based on models of the bone tumor capsule boundary. However, in actual clinical bone tumor surgery, in addition to the anatomically visible primary lesion and capsule, the reaction zone should also be considered when planning the surgical path to achieve precise R0 resection. For anatomically invisible reaction zones, some researchers use near-infrared fluorescent probes for detection, but their detection accuracy in clinical applications for bone tumors remains unstable.

[0005] Therefore, there is an urgent need for an automated extraction technology for bone tumor boundaries that includes the capsule and reaction zone. Summary of the Invention

[0006] The purpose of this application is to provide an automatic method and related device for extracting the boundary of a bone tumor, including the capsule and the reaction zone. This method can automatically extract the boundary of a bone tumor, including the capsule and the reaction zone, to obtain the minimally damaging surgical boundary for complete tumor resection, which is beneficial for achieving precise R0 resection of bone tumors.

[0007] To achieve the above objectives, this application provides the following solution:

[0008] In a first aspect, this application provides an automatic method for extracting the boundary of a bone tumor including a capsule and a reaction zone, the automatic method for extracting the boundary of a bone tumor including a capsule and a reaction zone comprising:

[0009] Obtain CT images of the bone containing the bone tumor;

[0010] Using the CT image as input, a bone tumor capsule boundary model is generated using a trained first segmentation model; the bone tumor capsule boundary model includes the primary lesion and capsule area of ​​the bone tumor;

[0011] Using the CT image as input, a healthy bone surface model is generated using a trained second segmentation model; the healthy bone surface model includes the surface of the bone where the bone tumor is located;

[0012] The bone tumor capsule boundary model is extended outward to obtain a preliminary bone tumor boundary model that includes the capsule and the reaction zone;

[0013] The preliminary model of the bone tumor boundary and the healthy bone surface model are subjected to surface intersection calculation. The part of the preliminary model of the bone tumor boundary located outside the healthy bone surface model is removed to obtain the bone tumor boundary model including the capsule and the reaction zone.

[0014] Optionally, the bone tumor capsule boundary model is extended outward to obtain a preliminary bone tumor boundary model including the capsule and the reaction zone, specifically including:

[0015] Calculate the normal vector of each vertex in the bone tumor capsule boundary model;

[0016] For each vertex, the vertex is extended outward by a preset offset along the direction of the vertex's normal vector to obtain a new vertex corresponding to the vertex;

[0017] Based on the connection relationships of each vertex in the bone tumor capsule boundary model, all the new vertices are connected to obtain a preliminary bone tumor boundary model that includes the capsule and reaction zone.

[0018] Optionally, the normal vector of each vertex in the bone tumor capsule boundary model is calculated, specifically including:

[0019] Calculate the normal vector of each plane in the bone tumor capsule boundary model;

[0020] For each vertex in the bone tumor capsule boundary model, the sum of the normal vectors of all planes containing the vertex is calculated to obtain the cumulative normal vector, and the ratio of the cumulative normal vector to the magnitude of the cumulative normal vector is calculated to obtain the normal vector of the vertex.

[0021] Optionally, before extending the bone tumor capsule boundary model outward to obtain a preliminary bone tumor boundary model including the capsule and reaction zone, the automatic extraction method for the bone tumor boundary including the capsule and reaction zone further includes:

[0022] The bone tumor capsule boundary model and the healthy bone surface model are simplified by using a quadratic edge folding extraction algorithm to obtain a simplified bone tumor capsule boundary model and a simplified healthy bone surface model. The simplified bone tumor capsule boundary model is used as the new bone tumor capsule boundary model, and the simplified healthy bone surface model is used as the new healthy bone surface model.

[0023] Optionally, the preliminary model of the bone tumor boundary and the healthy bone surface model are subjected to surface intersection calculation to remove the portion of the preliminary model of the bone tumor boundary located outside the healthy bone surface model, thereby obtaining a bone tumor boundary model including the capsule and reaction zone, specifically including:

[0024] Calculate the minimum distance from each vertex in the preliminary model of the bone tumor boundary to the healthy bone surface model;

[0025] For each vertex, determine whether the minimum distance is less than a distance threshold; if not, then record the vertex as the first vertex.

[0026] Connect all the first vertices to form the first model;

[0027] The first model is removed from the preliminary model of the bone tumor boundary to obtain the second model. The second model and the healthy bone surface model are then intersected by surface calculation. The part of the second model located outside the healthy bone surface model is removed to obtain the third model.

[0028] The first model and the third model are spliced ​​together to obtain a bone tumor boundary model that includes the capsule and the reaction zone.

[0029] Optionally, before generating a bone tumor capsule boundary model using the trained first segmentation model with the CT image as input, the automatic bone tumor boundary extraction method including the capsule and reaction zone further includes:

[0030] Obtain a dataset; the dataset includes sample CT images and a first label and a second label corresponding to each sample CT image, the first label being a sample bone tumor capsule boundary model, and the second label being a sample healthy bone surface model; both the first label and the second label are obtained by combining sample CT images and sample MRI images to perform image annotation on the sample CT images; both sample CT images and sample MRI images are obtained by scanning the bone where the bone tumor is located;

[0031] Construct the initial first segmentation model and the initial second segmentation model;

[0032] Using the sample CT image as input and the first label as the label, the initial first segmentation model is trained to obtain the trained first segmentation model.

[0033] Using the sample CT image as input and the second label as the label, the initial second segmentation model is trained to obtain the trained second segmentation model.

[0034] Secondly, this application provides an automated bone tumor boundary extraction device comprising a capsule and a reaction zone, the automated bone tumor boundary extraction device comprising:

[0035] The image acquisition module is used to acquire CT images obtained from scanning the bone where the bone tumor is located;

[0036] The first segmentation module is used to generate a bone tumor capsule boundary model using the CT image as input and a trained first segmentation model; the bone tumor capsule boundary model includes the primary lesion and capsule area of ​​the bone tumor.

[0037] The second segmentation module is used to generate a healthy bone surface model using the CT image as input and a trained second segmentation model; the healthy bone surface model includes the surface of the bone where the bone tumor is located.

[0038] The extension module is used to extend the bone tumor capsule boundary model outward to obtain a preliminary bone tumor boundary model that includes the capsule and the reaction zone;

[0039] The surface intersection module is used to perform surface intersection on the preliminary model of the bone tumor boundary and the healthy bone surface model, and remove the part of the model located outside the healthy bone surface model in the preliminary model of the bone tumor boundary to obtain a bone tumor boundary model that includes the capsule and the reaction zone.

[0040] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for automatic extraction of bone tumor boundaries including a capsule and a reaction zone as described above.

[0041] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for automatically extracting the boundary of a bone tumor including a capsule and a reaction zone as described above.

[0042] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the method for automatically extracting the boundary of a bone tumor containing a capsule and a reaction zone as described above.

[0043] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0044] This application provides an automatic method and related apparatus for extracting the boundary of a bone tumor, including a capsule and a reaction zone. Using CT images obtained from scanning the bone containing the bone tumor as input, a bone tumor capsule boundary model and a healthy bone surface model are generated using a trained first segmentation model and a trained second segmentation model, respectively. The bone tumor capsule boundary model is expanded outward to obtain a preliminary bone tumor boundary model including the capsule and reaction zone. The preliminary bone tumor boundary model and the healthy bone surface model are subjected to surface intersection calculation. The portion of the preliminary bone tumor boundary model located outside the healthy bone surface model is removed, resulting in a bone tumor boundary model including the capsule and reaction zone. This method includes the reaction zone while avoiding damage to surrounding normal tissues, automatically extracting the bone tumor boundary including the capsule and reaction zone, and obtaining the minimally damaging surgical boundary for complete tumor resection, facilitating precise R0 resection of the bone tumor. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is an application environment diagram of an automatic extraction method for bone tumor boundaries, including a capsule and a reaction zone, provided in Embodiment 1 of this application.

[0047] Figure 2 This is a flowchart illustrating an automatic extraction method for bone tumor boundaries, including a capsule and a reaction zone, provided in Embodiment 1 of this application.

[0048] Figure 3 This is a detailed flowchart illustrating an automatic extraction method for bone tumor boundaries, including a capsule and a reaction zone, provided in Embodiment 1 of this application.

[0049] Figure 4 This is a schematic diagram of the features and boundaries of a bone tumor provided in Embodiment 1 of this application.

[0050] Figure 5 This is a schematic diagram of the automatic extraction process of the femoral bone tumor boundary model provided in Embodiment 1 of this application.

[0051] Figure 6 This is a schematic diagram of a preliminary model of the boundary of a bone tumor in the femur, including the capsule and the reaction zone, provided in Embodiment 1 of this application.

[0052] Figure 7 This is a schematic diagram of the boundary model of a bone tumor in the femur, including the capsule and the reaction zone, provided in Embodiment 1 of this application.

[0053] Figure 8 This is a schematic diagram of the functional modules of an automatic bone tumor boundary extraction device including a capsule and a reaction zone, provided in Embodiment 2 of this application.

[0054] Figure 9 This is a schematic diagram of the structure of a computer device provided in Embodiment 3 of this application. Detailed Implementation

[0055] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0056] Example 1

[0057] The automatic extraction method for bone tumor boundaries, including the capsule and reaction zone, provided in this application embodiment can be applied to, for example... Figure 1 The application environment shown illustrates this. The terminal communicates with the server via a network. A data storage system stores the data the server needs to process. This data storage system can be set up independently, integrated into the server, or located in the cloud or on another server. The terminal can send the CT image to be processed to the server. Upon receiving the CT image, the server uses the CT image as input to generate a bone tumor capsule boundary model using a trained first segmentation model; it then uses the CT image as input to generate a healthy bone surface model using a trained second segmentation model; the bone tumor capsule boundary model is expanded outwards to obtain a preliminary bone tumor boundary model including the capsule and reaction zone; the preliminary bone tumor boundary model and the healthy bone surface model are then intersected using surface mapping, and the portion of the preliminary bone tumor boundary model located outside the healthy bone surface model is removed, resulting in a final bone tumor boundary model including the capsule and reaction zone. The server can then feed the obtained bone tumor boundary model back to the terminal.

[0058] In addition, in some embodiments, the automatic extraction method for bone tumor boundaries including the capsule and reaction zone can also be implemented by a server or a terminal alone. For example, the terminal can directly process the CT image to be processed, or the server can obtain the CT image to be processed from the data storage system and process the CT image to be processed.

[0059] The terminal can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. The server can be a standalone server or a server cluster composed of multiple servers, or it can be a cloud server.

[0060] like Figure 2 and Figure 3 As shown, an automated method for extracting the boundary of a bone tumor, including its capsule and reaction zone, is provided. This method is executed by a computer device, specifically a terminal or server, or both. In this embodiment, the method is applied to... Figure 1 Taking a server as an example, the automatic extraction method for bone tumor boundaries, including the capsule and reaction zone, includes the following steps:

[0061] Step S1: Obtain CT images of the bone containing the bone tumor.

[0062] Step S2: Using the CT image as input, generate a bone tumor capsule boundary model using the trained first segmentation model; the bone tumor capsule boundary model includes the primary lesion and capsule area of ​​the bone tumor.

[0063] Step S3: Using the CT image as input, a healthy bone surface model is generated using the trained second segmentation model; the healthy bone surface model includes the surface of the bone where the bone tumor is located.

[0064] Step S4: Expand the bone tumor capsule boundary model outward to obtain a preliminary bone tumor boundary model including the capsule and reaction zone.

[0065] Step S5: Perform surface intersection on the preliminary model of the bone tumor boundary and the healthy bone surface model, and remove the part of the preliminary model of the bone tumor boundary located outside the healthy bone surface model to obtain a bone tumor boundary model including the capsule and reaction zone.

[0066] By implementing steps S1 to S5 above, this embodiment can automatically extract the boundary of bone tumors including the capsule and reaction zone, which is beneficial for achieving precise R0 resection.

[0067] like Figure 4As shown, the inner edge of the reaction zone of a bone tumor is attached to the surface of the capsule, and its thickness is smooth and relatively uniform. Therefore, the approach of this embodiment is to expand the bone tumor capsule boundary model by a certain offset distance along the normal direction of the bone tumor capsule boundary model, encompassing both the capsule and the reaction zone, to form a new preliminary bone tumor boundary model. Further processing of this preliminary model yields a complete bone tumor boundary model, providing a safe and effective path planning for precise R0 resection of bone tumors using robot-assisted or computer-assisted methods. Based on this approach, this embodiment proposes an automatic bone tumor boundary extraction method that includes both the capsule and the reaction zone. By extending outwards along the normal direction of the bone tumor capsule boundary model by a certain distance, the reaction zone is included within the bone tumor identification range, thus obtaining a bone tumor boundary model that includes both the primary lesion and the capsule area, as well as the reaction zone. This method comprises three steps: Step 1, a deep learning network is used to automatically generate a bone tumor capsule boundary model and a healthy bone surface model for surface intersection calculation, followed by mesh simplification; Step 2, the model is extended outwards along the normals of each point on the bone tumor capsule boundary model to obtain a preliminary bone tumor boundary model including the capsule and reaction zone; Step 3, along the directions where the preliminary bone tumor boundary model may interfere with surrounding bone tissue, surface intersection operations are performed on the preliminary bone tumor boundary model and the healthy bone surface model of the bone containing the bone tumor to obtain the final bone tumor boundary model. This method can provide a safe tumor boundary for path planning in precise bone tumor resection surgery.

[0068] In S1, CT (Computed Tomography) images of the bone containing the bone tumor are acquired. The bone containing the bone tumor can be any bone in the human body, such as the sternum, coccyx, phalanges, femur, etc. The CT images are obtained by scanning the bone containing the bone tumor. The CT images are three-dimensional images and include multiple slices.

[0069] In S2, CT images are used as input, and a pre-trained first segmentation model is used to generate a bone tumor capsule boundary model. The bone tumor capsule boundary model includes the primary lesion and capsule region of the bone tumor. The bone tumor capsule boundary model is a three-dimensional model.

[0070] In S3, CT images are used as input, and a trained second segmentation model is used to generate a healthy bone surface model, which includes the surface of the bone where the bone tumor is located. The healthy bone surface model is a three-dimensional model.

[0071] Before generating a bone tumor capsule boundary model using a trained first segmentation model with CT images as input, the automatic bone tumor boundary extraction method including the capsule and reaction zone in this embodiment further includes:

[0072] (1) Obtain the dataset, which includes sample CT images and a first label and a second label corresponding to each sample CT image. The first label is the sample bone tumor capsule boundary model, and the second label is the sample healthy bone surface model. Both the first label and the second label are obtained by combining the sample CT images and sample MRI images to perform image annotation on the sample CT images. Both the sample CT images and sample MRI images are obtained by scanning the bone where the bone tumor is located.

[0073] The bone containing the bone tumor was scanned to obtain sample CT and sample MRI (Magnetic Resonance Imaging) images. The bone containing the bone tumor in this case is the same as the bone containing the bone tumor in S1. The sample CT and sample MRI images were merged to obtain a merged image. Based on the merged image, the sample CT image was manually labeled to obtain the first and second labels corresponding to the sample CT image.

[0074] (2) Construct the initial first segmentation model and the initial second segmentation model.

[0075] The initial first segmentation model and the initial second segmentation model can be any existing model with segmentation capabilities, such as a deep learning neural network model; specifically, the nnU-Net model can be used.

[0076] (3) Using the sample CT image as input and the first label as the label, the initial first segmentation model is trained to obtain the trained first segmentation model.

[0077] (4) Using the sample CT image as input and the second label as the label, train the initial second segmentation model to obtain the trained second segmentation model.

[0078] After obtaining the trained first segmentation model and the trained second segmentation model, this embodiment can automatically generate a bone tumor capsule boundary model including the primary lesion and capsule region, and a healthy bone surface model of the bone containing the bone tumor for surface intersection calculation using a deep learning neural network model. Specifically, the bone tumor capsule boundary model and the healthy bone surface model are automatically extracted based on image segmentation networks such as nnU-Net. Both the bone tumor capsule boundary model and the healthy bone surface model are three-dimensional models, and the surfaces of both can use triangular meshes. At this point, this embodiment can further use a triangular mesh simplification algorithm to compress the number of triangular meshes in the two models to improve the subsequent calculation and display speed.

[0079] Before extending the bone tumor capsule boundary model outward to obtain a preliminary bone tumor boundary model including the capsule and reaction zone, the automatic extraction method of the bone tumor boundary including the capsule and reaction zone in this embodiment further includes: using the Quadric Edge Collapse Decimation algorithm (QECD) to simplify the mesh of the bone tumor capsule boundary model and the healthy bone surface model respectively, to obtain a simplified bone tumor capsule boundary model and a simplified healthy bone surface model, and using the simplified bone tumor capsule boundary model as the new bone tumor capsule boundary model and the simplified healthy bone surface model as the new healthy bone surface model, and then performing subsequent S4 and S5.

[0080] Specifically, this embodiment uses a quadratic edge folding extraction algorithm to simplify the mesh. It reduces the number of vertices and faces by merging edges in the mesh while preserving the original mesh's geometric characteristics as much as possible. This algorithm is widely used in computer graphics and computer vision, especially in 3D model processing and optimization. Its core lies in calculating the cost function for each edge, which considers the edge's geometric properties such as length, angle, and the distribution of adjacent vertices. This cost function determines which edges should be merged, thereby simplifying the model. This quadratic edge folding extraction algorithm optimizes the vertex selection strategy based on a mesh simplification method using quadratic error metrics. It uses the shrinkage cost of vertex pairs as the sorting criterion, providing a more flexible simplified path selection algorithm. Furthermore, it updates the error matrix during the folding process, making the simplification process more accurately preserve the original model's geometry and reducing geometric errors. In this embodiment, the maximum error, average error, and mean squared error can be used to evaluate the simplification error of the model's geometry.

[0081] Compared to existing quadratic edge folding extraction algorithms, this embodiment can jointly use multiple local geometric features and corresponding weights in the calculation of the error matrix. That is, most existing quadratic edge folding extraction algorithms use fixed geometric features (distance metric or triangle importance) to calculate the error. This embodiment further introduces multiple local geometric features of vertices and surfaces on the basis of distance metric, and performs joint weighting based on weights to obtain the error matrix. The type and weight of the local geometric features are determined according to user needs. Furthermore, an adaptive error matrix is ​​used to calculate the shrinkage cost, thereby better preserving key geometric features while ensuring simplification, thus providing an improved quadratic edge folding extraction algorithm.

[0082] In S4, the bone tumor capsule boundary model is expanded outward to obtain a preliminary bone tumor boundary model that includes the capsule and the reaction zone. Specifically, the direction from the primary lesion to the capsule is defined as "outward." The normals of each vertex in the bone tumor capsule boundary model are automatically calculated, and each vertex is expanded outward along the normal direction by a certain distance. This expansion includes the reaction zone within the bone tumor boundary, thus incorporating it into the resection boundary. This results in a preliminary bone tumor boundary model that includes both the capsule and the reaction zone. This process extends the bone tumor capsule boundary model to meet the minimum damage boundary for complete bone tumor resection. The minimum damage boundary includes both the primary lesion and capsule of the bone tumor, as well as the reaction zone.

[0083] The preliminary model of the bone tumor boundary is an extension of the bone tumor capsule boundary model automatically segmented from the CT image. In this embodiment, the normal direction is represented by the normal vector. The normal vectors of adjacent triangular planes in the bone tumor capsule boundary model are averaged and normalized to obtain the normal vectors of each vertex in the bone tumor capsule boundary model. The normal vectors of each vertex in the bone tumor capsule boundary model are then added outward along the direction of the normal vector by a set offset to obtain the preliminary model of the bone tumor boundary. At this time, the preliminary model of the bone tumor boundary further includes the reaction zone, realizing the full coverage of the primary lesion, capsule area and reaction zone of the bone tumor, and realizing the extension of the boundary for complete tumor resection.

[0084] In S4, the bone tumor capsule boundary model is expanded outward to obtain a preliminary bone tumor boundary model that includes the capsule and reaction zone, specifically including:

[0085] (1) Calculate the normal vector of each vertex in the bone tumor capsule boundary model.

[0086] The calculation of the normal vector of each vertex in the bone tumor capsule boundary model specifically includes: calculating the normal vector of each plane in the bone tumor capsule boundary model; for each vertex in the bone tumor capsule boundary model, calculating the sum of the normal vectors of all planes containing that vertex to obtain the cumulative normal vector, and calculating the ratio of the cumulative normal vector to the magnitude of the cumulative normal vector to obtain the normal vector of that vertex.

[0087] (2) For each vertex, extend the vertex outward by a preset offset along the direction of the vertex's normal vector to obtain the new vertex corresponding to the vertex.

[0088] The preset offset can be 5mm.

[0089] (3) Based on the connection relationship of each vertex in the bone tumor capsule boundary model, connect all new vertices to obtain a preliminary bone tumor boundary model containing the capsule and reaction zone.

[0090] Since there is a correspondence between vertices and new vertices, all new vertices can be connected according to the connection relationship of each vertex to obtain a preliminary model of the bone tumor boundary.

[0091] In step S5, a surface intersection operation is performed on the preliminary model of the bone tumor boundary and the healthy bone surface model. The portion of the preliminary model located outside the healthy bone surface model is removed, resulting in a bone tumor boundary model that includes the capsule and reaction zone. This embodiment performs a surface intersection operation on the preliminary model of the bone tumor boundary and the healthy bone surface model of the bone containing the bone tumor, along the direction in which interference may occur between the preliminary model of the bone tumor boundary and the surrounding bone tissue. This yields the final bone tumor boundary model that includes the capsule and reaction zone, resolving the issue that the preliminary model of the bone tumor boundary may damage other surrounding bone tissue and ensuring the safety of the precise surgical planning for bone tumor resection.

[0092] During the surface intersection process, the distance between the surface of the preliminary model of the bone tumor boundary and the bone surface that may interfere is calculated. Models exceeding the set distance threshold are filtered out. The intersection of the remaining models and the healthy bone surface models is calculated. The filtered models are then merged with the intersection to obtain the final bone tumor boundary model. This ensures that the final bone tumor boundary model does not interfere with other surrounding bone tissues, minimizes tissue damage during surgical resection, and ensures the effectiveness of the resection range.

[0093] In S5, the preliminary model of the bone tumor boundary and the healthy bone surface model are intersected by surface calculation. The part of the preliminary model of the bone tumor boundary located outside the healthy bone surface model is removed, resulting in a bone tumor boundary model that includes the capsule and reaction zone. Specifically, this includes:

[0094] (1) Calculate the minimum distance from each vertex of the preliminary model of the bone tumor boundary to the healthy bone surface model.

[0095] Since both the preliminary model of the bone tumor boundary and the healthy bone surface model are known, and the coordinates of each point within them are also known, this embodiment can directly calculate the minimum distance from each vertex in the preliminary model of the bone tumor boundary to the healthy bone surface model. To further improve computational efficiency, this embodiment can utilize a multidimensional binary search tree to generate a first spatial structure for the preliminary model of the bone tumor boundary. The first spatial structure includes the number and position of each vertex in the preliminary model of the bone tumor boundary, the number of each plane, the association between each vertex and each plane, and the connection relationships between each vertex. The association relationship refers to which plane a vertex belongs to. A second spatial structure for the healthy bone surface model is generated using a multidimensional binary search tree. The second spatial structure includes the number and position of each vertex in the healthy bone surface model, the number of each plane, the association between each vertex and each plane, and the connection relationships between each vertex. Based on the first and second spatial structures, the minimum distance from each vertex in the preliminary model of the bone tumor boundary to the healthy bone surface model is calculated.

[0096] (2) For each vertex, determine whether the minimum distance is less than the distance threshold; if not, record the vertex as the first vertex.

[0097] (3) Connect all the first vertices to form the first model.

[0098] (4) Remove the first model from the preliminary model of the bone tumor boundary to obtain the second model, and perform surface intersection between the second model and the healthy bone surface model. Remove the part of the second model located outside the healthy bone surface model to obtain the third model.

[0099] This embodiment utilizes the `intersection` function in Python to find the intersection of surfaces. The third model is the intersection of the second model and the healthy bone surface model. After removing the portion of the model located outside the healthy bone surface model from the second model, the second model becomes the third model.

[0100] (5) The first model and the third model are spliced ​​together to obtain a bone tumor boundary model that includes the capsule and the reaction zone.

[0101] This embodiment proposes an automatic method for extracting the boundary of a bone tumor, including its capsule and reaction zone. First, the bone tumor capsule boundary model is extended outward along the normal direction of the capsule surface by a certain distance, thereby incorporating the reaction zone into the bone tumor boundary range, forming a preliminary bone tumor boundary model that includes the primary lesion, capsule, and reaction zone. Then, to avoid potential interference between the preliminary bone tumor boundary model and other surrounding bone tissues, a surface intersection operation is performed on the preliminary bone tumor boundary model and the healthy bone surface model of the bone containing the bone tumor along the direction where interference may occur, forming the final bone tumor boundary model. This final bone tumor boundary model includes the primary lesion, capsule, and reaction zone of the bone tumor within the boundary recognition range, achieving complete resection of the bone tumor. This method is mainly used for surgical path planning for precise R0 resection of bone tumors with robot assistance or computer assistance, ensuring the effectiveness and safety of the surgical path planning for precise R0 resection of bone tumors.

[0102] This embodiment is suitable for the automatic extraction of precise and safe boundaries of bone tumors, and has been tested on clinical CT scan images. The following example illustrates the automatic extraction of a femoral bone tumor boundary model.

[0103] Considering that the initial model of the bone tumor boundary may interfere with other surrounding bone tissues in the distal axial direction, and the object of surface intersection is the femoral condyle surface at the distal end of the femur, this example does not build a healthy bone surface model of the entire femur, but rather a healthy bone surface model of the femoral condyle surface at the distal end of the femur. In practical applications, if the user can confirm the direction of possible interference with other surrounding bone tissues, a healthy surface model of a portion of the bone containing the bone tumor can be built; otherwise, a healthy surface model of the entire bone containing the bone tumor can be built.

[0104] The automated method for extracting the boundaries of bone tumors in the femur, including the capsule and reaction zone, comprises the following six steps:

[0105] (1) Design an automated extraction scheme for bone tumor boundaries including capsule and reaction zone.

[0106] like Figure 5 As shown, with the goal of obtaining the minimum damage boundary that conforms to the complete resection of bone tumors in the femur, firstly, the extent of the bone tumor capsule in the input raw CT image data is identified, and a bone tumor capsule boundary model is automatically segmented and generated, and the mesh is simplified. The same segmentation and mesh simplification operation is performed on the distal femur as on the bone tumor capsule boundary model to obtain the femoral condyle surface model. Then, the bone tumor capsule boundary model is extended outward by a certain distance to include the reaction zone. The minimum reference value for this distance can be 5 mm. Finally, the intersection of the extended preliminary bone tumor boundary model and the femoral condyle surface model is obtained to obtain the final bone tumor boundary model.

[0107] (2) Automatically generate a femoral bone tumor capsule boundary model

[0108] This example uses a neural network model to segment images and generate a femoral tumor capsule boundary model. First, sample CT and MRI images of the femur are merged using ITK-SNAP software (version 4.0.2). Then, clinicians perform semi-automatic segmentation and image annotation to form a dataset. Finally, nnU-Net is trained using the annotated dataset to obtain a trained first segmentation model, which is then used for the automatic segmentation of the bone tumor capsule boundary model. The automatically segmented bone tumor capsule boundary model is stored as a binary image in NIFTI format and as a 3D geometric model in STL format.

[0109] (3) Mesh-simplified femoral bone tumor capsule boundary model

[0110] For the aforementioned STL-formatted bone tumor capsule boundary model, a quadratic edge folding extraction algorithm is used for mesh simplification. First, the error matrix of each vertex is calculated, and the error tetrahedron of each vertex is further calculated. The squared distance between the vertex and its adjacent surfaces is accumulated to quantify the influence of vertex motion on the model shape. Then, for each vertex pair, the optimal shrinkage target is calculated to minimize the sum of the error tetrahedrons of the two vertices. All vertex pairs are inserted into a priority queue in ascending order of shrinkage cost. The vertex pair with the smallest shrinkage cost is selected from the priority queue for shrinkage until the set simplification target is reached, resulting in the simplified bone tumor capsule boundary model, which is then stored in STL format.

[0111] (4) Extend the bone tumor capsule boundary model in the normal direction to obtain a preliminary model of the bone tumor boundary.

[0112] For the simplified bone tumor capsule boundary model obtained in STL format above, firstly according to the formula The normal vectors of all adjacent triangular planes in the simplified bone tumor capsule boundary model are averaged and normalized to obtain the normal vector of each vertex, n. vertax (v i ) represents the i-th vertex v i The normal vector, n facet (j) is the value containing the i-th vertex v. i The normal vector of the j-th plane. Specifically, firstly, by traversing each plane in the simplified bone tumor capsule boundary model, the normal vector of each plane is added to each vertex it contains, thereby calculating the cumulative normal vector of the vertices; then, the cumulative normal vector of the vertices is normalized to obtain the normal vector of each vertex. Based on the calculation of the vertex normal vectors of the simplified bone tumor capsule boundary model, the new position of each vertex after extending 5mm along its normal vector direction is calculated, thereby obtaining a preliminary model of the bone tumor boundary that can further cover the reaction area, such as... Figure 6 As shown. In this example, the 5mm offset is a reference value set based on clinician experience, which can effectively cover most of the reaction zone of bone tumors.

[0113] (5) Automatically generate and simplify the femoral condyle surface model

[0114] First, sample CT and MRI images of the femur were merged using ITK-SNAP software (version 4.0.2). Then, clinicians performed semi-automatic segmentation and image annotation to create a dataset. The annotated dataset was used to train nnU-Net, resulting in a trained second segmentation model. This trained second segmentation model was then used for automatic segmentation of the femoral condyle surface model. Finally, a binarized image in NIFTI format and a 3D geometric model in STL format were output. A quadratic edge-folding extraction algorithm was used for mesh simplification, resulting in a simplified femoral condyle surface model.

[0115] (6) The preliminary model of the femoral condyle boundary and the simplified femoral condyle surface model are obtained by using curved surfaces to obtain the final femoral bone tumor boundary model.

[0116] Read the STL files of the models from steps (4) and (5), use a multidimensional binary search tree to implement spatial indexing, calculate the minimum distance from each vertex in the preliminary model of the bone tumor boundary to the femoral condyle surface, and set a distance threshold to filter the preliminary model of the bone tumor boundary, removing models that exceed the set distance threshold. Further, find the intersection between the remaining model and the simplified femoral condyle surface model, merge the removed models and the intersection, ensuring that the surface of the preliminary model of the bone tumor boundary does not exceed the femoral condyle surface, thus obtaining the final bone tumor boundary model that includes the capsule and reaction zone and meets the requirement of minimum tissue damage for safe resection, as shown below. Figure 7 As shown.

[0117] This application also provides an application scenario where the above-described automatic extraction method for bone tumor boundaries, including the capsule and reaction zone, is applied. Specifically, the automatic extraction method for bone tumor boundaries, including the capsule and reaction zone, provided in this embodiment can be applied in a bone tumor boundary display scenario. The bone tumor boundary display scenario includes an image acquisition stage, an image processing stage, and a result display stage. The image acquisition stage is used to acquire CT images, the data processing stage is used to process the CT images to obtain a bone tumor boundary model, and the result display stage is used to display the bone tumor boundary model to the user. The automatic extraction method for bone tumor boundaries, including the capsule and reaction zone, provided in this embodiment belongs to the image processing stage.

[0118] Example 2

[0119] Based on the same inventive concept, this application also provides an automatic bone tumor boundary extraction device including a capsule and a reaction zone for implementing the above-described automatic bone tumor boundary extraction method including a capsule and a reaction zone. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations in the embodiments of the automatic bone tumor boundary extraction device including a capsule and a reaction zone provided below can be found in the limitations of the automatic bone tumor boundary extraction method including a capsule and a reaction zone described above, and will not be repeated here.

[0120] like Figure 8 As shown, an automated bone tumor boundary extraction device comprising a capsule and a reaction zone is provided. The automated bone tumor boundary extraction device comprising a capsule and a reaction zone includes:

[0121] The image acquisition module M1 is used to acquire CT images obtained by scanning the bone where the bone tumor is located.

[0122] The first segmentation module M2 is used to generate a bone tumor capsule boundary model using the CT image as input and a trained first segmentation model; the bone tumor capsule boundary model includes the primary lesion and capsule area of ​​the bone tumor.

[0123] The second segmentation module M3 is used to generate a healthy bone surface model using the CT image as input and a trained second segmentation model; the healthy bone surface model includes the surface of the bone where the bone tumor is located.

[0124] The extension module M4 is used to extend the bone tumor capsule boundary model outward to obtain a preliminary bone tumor boundary model that includes the capsule and the reaction zone.

[0125] The surface intersection module M5 is used to perform surface intersection on the preliminary model of the bone tumor boundary and the healthy bone surface model, and remove the part of the model located outside the healthy bone surface model in the preliminary model of the bone tumor boundary to obtain a bone tumor boundary model that includes the capsule and the reaction zone.

[0126] Example 3

[0127] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 9 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores image data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements an automatic extraction method for bone tumor boundaries, including the capsule and reaction zone.

[0128] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0129] In one exemplary embodiment, a computer device is also provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to implement the automatic extraction method for bone tumor boundaries including a capsule and a reaction zone as described in Embodiment 1.

[0130] Example 4

[0131] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the automatic extraction method for bone tumor boundaries including a capsule and a reaction zone as described in Embodiment 1.

[0132] Example 5

[0133] This application provides a computer program product, including a computer program that, when executed by a processor, implements the automatic extraction method for bone tumor boundaries including a capsule and a reaction zone as described in Embodiment 1.

[0134] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0135] Any references to memory, databases, or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0136] The processors involved in the various embodiments provided in this application may be general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited thereto.

[0137] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0138] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for automatically extracting the boundary of a bone tumor, comprising a capsule and a reaction zone, characterized in that, The automatic extraction method for bone tumor boundaries, including the capsule and reaction zone, includes: Obtain CT images of the bone containing the bone tumor; Using the CT image as input, a bone tumor capsule boundary model is generated using a trained first segmentation model; the bone tumor capsule boundary model includes the primary lesion and capsule area of ​​the bone tumor; Using the CT image as input, a healthy bone surface model is generated using a trained second segmentation model; the healthy bone surface model includes the surface of the bone where the bone tumor is located; Calculate the normal vector of each plane in the bone tumor capsule boundary model; for each vertex in the bone tumor capsule boundary model, calculate the sum of the normal vectors of all planes containing the vertex to obtain the cumulative normal vector, and calculate the ratio of the cumulative normal vector to its magnitude to obtain the normal vector of the vertex; for each vertex, extend the vertex outward by a preset offset along the direction of the vertex's normal vector to obtain a new vertex corresponding to the vertex; based on the connection relationship of each vertex in the bone tumor capsule boundary model, connect all the new vertices to obtain a preliminary bone tumor boundary model including the capsule and reaction zone; Calculate the minimum distance from each vertex in the preliminary model of the bone tumor boundary to the healthy bone surface model; for each vertex, determine whether the minimum distance is less than a distance threshold; if not, record the vertex as the first vertex; connect all the first vertices to form a first model; remove the first model from the preliminary model of the bone tumor boundary to obtain a second model, and perform surface intersection on the second model and the healthy bone surface model to remove the part of the second model located outside the healthy bone surface model to obtain a third model; stitch the first model and the third model together to obtain a bone tumor boundary model including the capsule and reaction zone.

2. The method for automatically extracting the boundary of a bone tumor, including a capsule and a reaction zone, according to claim 1, is characterized in that, Before extending the bone tumor capsule boundary model outward to obtain a preliminary bone tumor boundary model including the capsule and reaction zone, the automatic extraction method for the bone tumor boundary including the capsule and reaction zone further includes: The bone tumor capsule boundary model and the healthy bone surface model are simplified by using a quadratic edge folding extraction algorithm to obtain a simplified bone tumor capsule boundary model and a simplified healthy bone surface model. The simplified bone tumor capsule boundary model is used as the new bone tumor capsule boundary model, and the simplified healthy bone surface model is used as the new healthy bone surface model.

3. The method for automatically extracting the boundary of a bone tumor, including a capsule and a reaction zone, according to claim 1, is characterized in that... Before generating a bone tumor capsule boundary model using the trained first segmentation model with the CT image as input, the automatic bone tumor boundary extraction method including the capsule and reaction zone further includes: Obtain a dataset; the dataset includes sample CT images and a first label and a second label corresponding to each sample CT image, the first label being a sample bone tumor capsule boundary model, and the second label being a sample healthy bone surface model; both the first label and the second label are obtained by combining sample CT images and sample MRI images to perform image annotation on the sample CT images; both sample CT images and sample MRI images are obtained by scanning the bone where the bone tumor is located; Construct the initial first segmentation model and the initial second segmentation model; Using the sample CT image as input and the first label as the label, the initial first segmentation model is trained to obtain the trained first segmentation model. Using the sample CT image as input and the second label as the label, the initial second segmentation model is trained to obtain the trained second segmentation model.

4. An automated extraction device for bone tumor boundaries comprising a capsule and a reaction zone, characterized in that, The automated bone tumor boundary extraction device, which includes a capsule and a reaction zone, comprises: The image acquisition module is used to acquire CT images obtained from scanning the bone where the bone tumor is located; The first segmentation module is used to generate a bone tumor capsule boundary model using the CT image as input and a trained first segmentation model; the bone tumor capsule boundary model includes the primary lesion and capsule area of ​​the bone tumor. The second segmentation module is used to generate a healthy bone surface model using the CT image as input and a trained second segmentation model; the healthy bone surface model includes the surface of the bone where the bone tumor is located. An extension module is used to calculate the normal vector of each plane in the bone tumor capsule boundary model; for each vertex in the bone tumor capsule boundary model, the sum of the normal vectors of all planes containing the vertex is calculated to obtain the cumulative normal vector, and the ratio of the cumulative normal vector to its magnitude is calculated to obtain the normal vector of the vertex; for each vertex, the vertex is extended outward by a preset offset along the direction of the vertex's normal vector to obtain a new vertex corresponding to the vertex; based on the connection relationship of each vertex in the bone tumor capsule boundary model, all the new vertices are connected to obtain a preliminary bone tumor boundary model including the capsule and the reaction zone; The surface intersection module is used to calculate the minimum distance from each vertex in the preliminary model of the bone tumor boundary to the healthy bone surface model; for each vertex, it is determined whether the minimum distance is less than a distance threshold; if not, the vertex is recorded as the first vertex; all the first vertices are connected to form a first model; the first model is removed from the preliminary model of the bone tumor boundary to obtain a second model, and the second model and the healthy bone surface model are subjected to surface intersection, and the part of the second model located outside the healthy bone surface model is removed to obtain a third model; the first model and the third model are spliced ​​together to obtain a bone tumor boundary model including the capsule and the reaction zone.

5. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the automatic extraction method for bone tumor boundaries including a capsule and a reaction zone as described in any one of claims 1-3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the automatic extraction method for bone tumor boundaries including the capsule and reaction zone as described in any one of claims 1-3.

7. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the automatic extraction method for bone tumor boundaries including the capsule and reaction zone as described in any one of claims 1-3.