Femoral head central point identification method, device, equipment and computer program product

Abstract

The embodiment of the invention is suitable for the technical field of computer-aided medical treatment and image processing, and provides a femoral head central point recognition method, device and equipment and a computer program product, and the method comprises the steps: obtaining to-be-recognized femoral head image data; rendering the femoral head image data to obtain a first femoral head image and a second femoral head image; a femoral head center recognizer is called to process the first femoral head image and the second femoral head image, two groups of two-dimensional coordinates output by the femoral head center recognizer are obtained, and the femoral head center recognizer is obtained through training based on a plurality of two-dimensional femoral head sample images marked with femoral head center points; and determining a femoral head center point in the femoral head image data according to the two groups of two-dimensional coordinates. By adopting the method, the caput femoris central point can be quickly and accurately determined for any type of caput femoris image data.

Description

TECHNICAL FIELD

[0001] The embodiment of the present application belongs to the technical field of computer-aided medical technology and image processing technology, and particularly relates to a femoral head center point identification method, device, equipment and computer program product. BACKGROUND

[0002] Identification of human joint landmark points has important applications in the fields of medical image analysis, orthopedic surgery planning and auxiliary diagnosis. For example, the femoral head center point is a key reference point for preoperative planning of hip arthroplasty, biomechanical analysis and related disease diagnosis, and has important significance in orthopedic clinical diagnosis and surgery planning. Traditional femoral head center point identification methods mainly rely on the experience and manual measurement of doctors. This method is not only time-consuming and labor-intensive, but also low in efficiency, and is affected by the subjective factors of doctors, making it difficult to ensure the accuracy of the identification results.

[0003] With the development of medical imaging technology, the industry has begun to use computer devices combined with image processing related means to assist in the automatic identification of human landmark points such as femoral head center points. However, the efficiency of automatic identification of femoral head center points depends on computer resources, and in particular for three-dimensional femoral head image data, the efficiency of computer automatic identification still needs to be improved. SUMMARY

[0004] Therefore, the embodiment of the present application provides a femoral head center point identification method, device, equipment and computer program product to improve the identification efficiency and accuracy of the femoral head center point.

[0005] The first aspect of the embodiment of the present application provides a femoral head center point identification method, comprising: obtaining femoral head image data to be identified, the femoral head image data being two-dimensional image data or three-dimensional image data; rendering the femoral head image data to obtain a first femoral head image and a second femoral head image, the first femoral head image and the second femoral head image both being two-dimensional images; calling a femoral head center identifier to process the first femoral head image and the second femoral head image, and obtaining two sets of two-dimensional coordinates output by the femoral head center identifier, the femoral head center identifier being trained based on a plurality of two-dimensional femoral head sample images labeled with femoral head center points; determining a femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates.

[0006] Optionally, the rendering of the femoral head image data to obtain a first femoral head image and a second femoral head image comprises: rendering the femoral head image data to a three-dimensional space; determining a first view angle and a second view angle; performing image acquisition on the femoral head image data in the three-dimensional space based on the first view angle and the second view angle respectively to obtain a first femoral head image and a second femoral head image.

[0007] Optionally, the determining the first view angle and the second view angle comprises: constructing a three-dimensional coordinate system corresponding to the three-dimensional space; determining a target surface of the three-dimensional space and determining the first view angle in a direction directly facing the target surface, the target surface being a plane perpendicular to any axis in the three-dimensional coordinate system; determining the second view angle orthogonal to the first view angle.

[0008] Optionally, the performing image acquisition on the femoral head image data in the three-dimensional space based on the first view angle and the second view angle respectively to obtain a first femoral head image and a second femoral head image comprises: based on the first view angle, simulating an operation of performing image acquisition on the femoral head image data by a first camera at the first view angle position to generate a first femoral head image acquired by the first camera; based on the second view angle, simulating an operation of performing image acquisition on the femoral head image data by a second camera at the second view angle position to generate a second femoral head image acquired by the second camera.

[0009] Optionally, the two sets of two-dimensional coordinates comprise a first coordinate and a second coordinate, the first coordinate being a first femoral head center point coordinate output by the femoral head center identifier for the first femoral head image, the second coordinate being a second femoral head center point coordinate output by the femoral head center identifier for the second femoral head image, the first view angle and the second view angle being orthogonal view angles, and the determining the femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates comprising: determining a first connecting line from the first view angle position to the first femoral head center point and a second connecting line from the second view angle position to the second femoral head center point, an extension line of the first connecting line intersecting an extension line of the second connecting line at a point in space; projecting the intersection point of the extension line of the first connecting line and the extension line of the second connecting line to a three-dimensional space corresponding to the femoral head image data; identifying the projection point in the three-dimensional space as the femoral head center point in the femoral head image data.

[0010] Optionally, the projecting the intersection point of the extension line of the first connecting line and the extension line of the second connecting line to the space corresponding to the femoral head image data comprises: determine a spatial matrix and a calibration matrix corresponding to the first view angle, the first view angle being directly opposite to a target surface of the femoral head image data, the target surface being a plane perpendicular to any axis in a coordinate system in which the three-dimensional space is located; perform matrix calculation on the intersection point based on the spatial matrix and the calibration matrix, to obtain a femoral head center point coordinate in the three-dimensional space corresponding to the femoral head image data.

[0011] Optionally, the two-dimensional femoral head sample images used to train the femoral head center identifier include a plurality of two-dimensional images obtained by performing view reconstruction on a plurality of femoral head sample image data under a plurality of view angles, and the number of two-dimensional femoral head sample images obtained based on any femoral head sample image data is equal to the number of the plurality of view angles.

[0012] A second aspect of the embodiment of the present application provides a femoral head center point identification device, which comprises: An acquisition module is configured to acquire femoral head image data to be identified, the femoral head image data being two-dimensional image data or three-dimensional image data. A rendering module is configured to render the femoral head image data to obtain a first femoral head image and a second femoral head image, the first femoral head image and the second femoral head image both being two-dimensional images. A processing module is configured to call a femoral head center identifier to process the first femoral head image and the second femoral head image, and obtain two sets of two-dimensional coordinates output by the femoral head center identifier, the femoral head center identifier being trained based on a plurality of two-dimensional femoral head sample images labeled with femoral head center points. A determination module is configured to determine a femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates.

[0013] A third aspect of the embodiment of the present application provides a computer device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the computer device implements the method according to any one of the first aspect.

[0014] A fourth aspect of the embodiment of the present application provides a computer readable storage medium, which stores a computer program, and when the computer program is executed by a computer, the method according to any one of the first aspect is implemented.

[0015] A fifth aspect of the embodiment of the present application provides a computer program product, which comprises a computer program, and when the computer program is executed, the method according to any one of the first aspect is executed.

[0016] Compared with the prior art, the embodiments of the present application have the following beneficial effects: The femoral head center identifier can render the femoral head image data to a three-dimensional space after obtaining the femoral head image data to be identified. In this way, the first femoral head image and the second femoral head image under different viewing angles can be collected. The trained femoral head center identifier can be called to process the first femoral head image and the second femoral head image, and two sets of two-dimensional coordinates output by the femoral head center identifier can be obtained. The computer device can calculate the femoral head center point in the femoral head image data based on the two sets of two-dimensional coordinates. The femoral head center identifier is trained to automatically process the femoral head image data in the embodiments of the present application, and the femoral head center point is identified, which improves the identification efficiency and accuracy of the femoral head center point. In addition, the input data of the femoral head center identifier in the embodiments of the present application is a two-dimensional image. For the femoral head image data to be identified, which is three-dimensional image data, only two-dimensional images under different viewing angles are collected for identification. Compared with the way of directly processing three-dimensional data by the model in the prior art, the model processing speed can be improved by more than 100 times, which can significantly improve the efficiency of femoral head center point identification. In addition, the embodiments of the present application can process any type of image data. Whether it is two-dimensional femoral head image data or three-dimensional femoral head image data, the identification method provided by the embodiments of the present application can obtain accurate identification results. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] Figure 1 is a schematic diagram of a femoral head center point identification method provided by the embodiments of the present application; Figure 2 is a schematic diagram of another femoral head center point identification method provided by the embodiments of the present application; Figure 3 is a schematic diagram of a possible implementation of S202 in the femoral head center point identification method provided by the embodiments of the present application; Figure 4 is a schematic diagram of a femoral head center point identification process provided by the embodiments of the present application; Figure 5is a schematic diagram of a femoral head center point recognition device provided by an embodiment of the present application. Figure 6 is a schematic diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0019] In the following description, for the purpose of explanation and not limitation, specific details are set forth, such as particular system configurations, techniques, etc., in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

[0020] The technical solutions of the present application are described below through specific embodiments.

[0021] Referring to Figure 1 , a schematic diagram of a femoral head center point recognition method provided by an embodiment of the present application is shown, which can specifically include the following steps: S101, acquiring femoral head image data to be recognized, the femoral head image data being two-dimensional image data or three-dimensional image data.

[0022] It should be noted that the present method can be applied to a computer device, i.e., the execution subject of the embodiments of the present application is a computer device, which can automatically recognize the femoral head center point in the femoral head image data by executing each step of the present method, thereby improving the recognition efficiency and accuracy of the femoral head center point.

[0023] The femoral head image data to be recognized can be obtained by performing image scanning on the relevant region of the femoral head of a patient, and the scanning process can be performed before surgery or can include intraoperative scanning. The femoral head image data obtained by preoperative scanning can be used for preoperative plan planning, and the femoral head image data obtained by intraoperative scanning can provide a reference for the surgeon during the operation.

[0024] The femoral head image data to be recognized in the embodiments of the present application can be any type of data. For example, the femoral head image data can be two-dimensional image data or three-dimensional image data. Exemplarily, the two-dimensional image data can include X-ray images, etc., and the three-dimensional image data can include computed tomography (CT) data, magnetic resonance imaging (MRI) data, etc., and the embodiments of the present application do not limit the type of femoral head image data.

[0025] The femoral head image data of any type can be acquired or scanned by using a corresponding image device. For example, the X-ray image can be acquired by using an X-ray machine, and the CT data or MRI data can be acquired by using a CT device or an MRI device, respectively. The image data acquired by the image device and containing the femoral head region is the femoral head image data in the embodiments of the present application, which can be transmitted to the computer device for processing to automatically identify the femoral head center point.

[0026] In S102, the femoral head image data is rendered to obtain a first femoral head image and a second femoral head image, both of which are two-dimensional images.

[0027] In the embodiments of the present application, the rendering of the femoral head image data can refer to the process of rendering the femoral head image data to a three-dimensional space by the computer device after receiving the femoral head image data. Whether it is two-dimensional image data or three-dimensional image data, the computer device can render it to a three-dimensional space.

[0028] The purpose of rendering the femoral head image data to a three-dimensional space is to facilitate image acquisition based on the rendered image data in the three-dimensional space to obtain a two-dimensional image. That is, through rendering, the subsequent identification process can be converted to a related processing process for a two-dimensional image to reduce the computer resources required in the identification process and facilitate the implementation of the method in a wider range of devices. At the same time, since the subsequent identification process is for a two-dimensional image, compared with the processing of three-dimensional data by the computer device, the identification efficiency of the femoral head center point can also be improved.

[0029] In the embodiments of the present application, for the femoral head image data that is originally three-dimensional image data, it can be directly rendered to a three-dimensional space by using a standard processing method. For the femoral head image data that is two-dimensional image data, such as a two-dimensional image, it can be rendered to a three-dimensional space by using different window widths and window levels, which are not limited in the embodiments of the present application.

[0030] After rendering the femoral head image data to a three-dimensional space, the computer device can acquire a first femoral head image and a second femoral head image based on the image data in the three-dimensional space, both of which are two-dimensional images.

[0031] In an example, the femoral head image data rendered to a three-dimensional space can be photographed by using two cameras to obtain a first femoral head image and a second femoral head image. Alternatively, the computer device can simulate the operation of the camera for image acquisition of the femoral head image data at different angles to generate a first femoral head image and a second femoral head image. The embodiments of the present application are not limited in this regard.

[0032] S103, calling a femoral head center identifier to process the first femoral head image and the second femoral head image, and obtaining two sets of two-dimensional coordinates output by the femoral head center identifier, the femoral head center identifier being trained based on a plurality of two-dimensional femoral head sample images labeled with femoral head center points.

[0033] In the embodiments of the present application, the femoral head center identifier can be a model that is pre-trained and can process two-dimensional images and automatically identify the femoral head center points in the images. The training of the femoral head center identifier can be achieved by using a plurality of two-dimensional femoral head sample images labeled with femoral head center points.

[0034] In a possible implementation manner of the embodiments of the present application, the two-dimensional femoral head sample images used to train the femoral head center identifier can include a plurality of image sets, each image set containing two sample images from the same data at different angles, and the positions of the femoral head center points in the two sample images have been pre-labeled. For example, the coordinates of the femoral head center points in the sample images. In this way, a plurality of paired data {sample image, femoral head center point coordinates} can be formed.

[0035] The computer device can train the plurality of paired data {sample image, femoral head center point coordinates} formed above to obtain the femoral head center identifier.

[0036] In a possible implementation manner of the embodiments of the present application, the two-dimensional femoral head sample images can be model trained based on related target detection algorithms or models in the field of computer vision to obtain the femoral head center identifier. For example, the femoral head center identifier can be trained based on Gaussian Heatmap, various versions of YOLO algorithm. Taking Gaussian Heatmap as an example, a heat map can be trained for each point in the sample image, and the femoral head center point in the image is determined through image segmentation and regression processing, and the convergence of the model is determined in combination with the labeled actual femoral head center point in the image. In the case of meeting the corresponding convergence conditions, the model is output as the femoral head center identifier for subsequent automatic identification of the femoral head center point.

[0037] The femoral head center identifier can be deployed in the computer device, or in other devices or servers connected to the computer device, for example, the femoral head center identifier can be deployed in the cloud. After the computer device completes the rendering of the femoral head image data to obtain the first femoral head image and the second femoral head image, the femoral head center identifier can be called to process the two two-dimensional images, and output the two-dimensional coordinates (u, v) of the femoral head center point.

[0038] In the embodiments of the present application, the femoral head center identifier can process the input first femoral head image and the second femoral head image respectively, and output the two-dimensional coordinates of the femoral head center point. That is, for the first femoral head image, the femoral head center identifier can output a set of two-dimensional coordinates (u1, v1); for the second femoral head image, the femoral head center identifier can also output a set of two-dimensional coordinates (u2, v2). In this way, two sets of two-dimensional coordinates are obtained.

[0039] S104, determining the femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates.

[0040] The computer device can calculate based on the two sets of two-dimensional coordinates output by the femoral head center identifier to obtain the coordinates of the femoral head center point in the to-be-identified femoral head image data.

[0041] In the embodiments of the present application, since the first femoral head image and the second femoral head image are two-dimensional images collected at different angles for the same femoral head image data, the two sets of two-dimensional coordinates output by the femoral head center identifier actually represent the coordinates of the same point at different angles. Based on the above principle, a position point can be determined in the rendered three-dimensional space according to the two sets of two-dimensional coordinates, which represents the femoral head center point in the to-be-identified femoral head image data.

[0042] In the embodiment of the present application, the femoral head center identifier capable of automatically processing and identifying two-dimensional images and outputting the femoral head center point coordinates in the image is pre-trained. After obtaining the femoral head image data to be identified, the femoral head image data is rendered to a three-dimensional space. In this way, the first femoral head image and the second femoral head image under different viewing angles can be acquired. The trained femoral head center identifier is called to process the first femoral head image and the second femoral head image, and two sets of two-dimensional coordinates output by the femoral head center identifier can be obtained. The computer device can calculate the femoral head center point in the femoral head image data based on the two sets of two-dimensional coordinates. The femoral head center identifier is trained in the embodiment of the present application to automatically process the femoral head image data, and the femoral head center point is identified, which improves the identification efficiency and accuracy of the femoral head center point. Moreover, the input data of the femoral head center identifier in the embodiment of the present application is a two-dimensional image. For the femoral head image data to be identified, which is three-dimensional image data, only two-dimensional images under different viewing angles are acquired for identification. Compared with the way of directly processing three-dimensional data by using other machine learning models in the prior art, the model processing speed can be improved by more than 100 times, and the efficiency of femoral head center point identification can be significantly improved. In addition, the embodiment of the present application can process any type of image data. Whether it is two-dimensional femoral head image data or three-dimensional femoral head image data, the identification method provided in the embodiment of the present application can obtain accurate identification results.

[0043] Referring to Figure 2 , another schematic diagram of a femoral head center point identification method provided in the embodiment of the present application is shown, which can specifically include the following steps: S201, obtaining femoral head image data to be identified, and rendering the femoral head image data to a three-dimensional space.

[0044] The execution subject of the embodiment of the present application is a computer device. The computer device can be in communication connection with an image device. After the image device scans the femoral head region of a patient and obtains femoral head image data, the image device can transmit the femoral head image data to the computer device. The femoral head image data in the embodiment of the present application can refer to medical image data containing the femoral head region of a patient. For example, an X-ray image, CT data or MRI data.

[0045] For the femoral head image data to be identified, the computer device can render it to a three-dimensional space. That is: {X-Ray, CT, MRI}→{S3}, where X-Ray represents an X-ray image, CT represents CT data, MRI represents MRI data, and S3 represents a three-dimensional space.

[0046] It should be noted that the femoral head center point identification method provided in the embodiments of the present application can process and identify any type of femoral head image data, and output the coordinates of the femoral head center point. Before processing, whether it is two-dimensional image data or three-dimensional image data, the computer device can render it to a three-dimensional space.

[0047] S202, image acquisition is performed on the femoral head image data in the three-dimensional space based on the first view angle and the second view angle respectively, to obtain a first femoral head image and a second femoral head image.

[0048] In the embodiments of the present application, after rendering the femoral head image data to be identified to a three-dimensional space, the computer device can determine the first view angle and the second view angle, and obtain the first femoral head image and the second femoral head image based on the first view angle and the second view angle. The above process can be similar to the process of image acquisition on the femoral head image data rendered to the three-dimensional space under the first view angle and the second view angle respectively. Therefore, the first femoral head image can refer to the image acquired under the first view angle, and the second femoral head image can refer to the image acquired under the second view angle. The first femoral head image and the second femoral head image are both two-dimensional images.

[0049] In a possible implementation manner of the embodiments of the present application, as shown in FIG. 2, S202 can specifically include the following sub-steps S2021-S2023: Figure 3 S2021, a three-dimensional coordinate system corresponding to the three-dimensional space is constructed.

[0050] S2022, a target surface of the three-dimensional space is determined, and the first view angle is determined in a direction facing the target surface, the target surface being a plane perpendicular to any axis in the three-dimensional coordinate system.

[0051] S2023, the second view angle orthogonal to the first view angle is determined.

[0052] Specifically, the computer device can first construct a three-dimensional coordinate system corresponding to the three-dimensional space based on a certain point in the three-dimensional space, taking the point as the origin, which can be represented as XYZ, i.e., the three-dimensional coordinate system XYZ. Then, a target surface is determined in the three-dimensional space, which can be a plane perpendicular to any axis in the three-dimensional coordinate system XYZ. Exemplarily, the XY plane in the three-dimensional coordinate system XYZ can be taken as the target surface, and the XY plane is a plane perpendicular to the Z axis in the three-dimensional coordinate system XYZ; or, the YZ plane in the three-dimensional coordinate system XYZ can be taken as the target surface, and the YZ plane is perpendicular to the X axis in the three-dimensional coordinate system XYZ; or, the XZ plane in the three-dimensional coordinate system XYZ can be taken as the target surface, and the present application does not limit this.

[0053] ​On this basis, the first view angle and the second view angle can be determined. Exemplarily, the first view angle can be determined in a direction facing the target plane. For example, in the case of taking the XY plane in the three-dimensional coordinate system XYZ as the target plane, the direction of the first view angle can be facing the XY plane.

[0054] In the embodiments of the present application, the first view angle and the second view angle can be orthogonal view angles. Therefore, in the case of determining the first view angle based on the target plane in the three-dimensional space, the direction of the second view angle can be determined according to the orthogonal relationship.

[0055] Then, the femoral head image data in the three-dimensional space can be respectively image collected based on the first view angle and the second view angle, to obtain a first femoral head image and a second femoral head image.

[0056] Specifically, the computer device can simulate the operation of image collecting the femoral head image data by the first camera at the first view angle position based on the first view angle, to generate the first femoral head image collected by the first camera. Similarly, the computer device can simulate the operation of image collecting the femoral head image data by the second camera at the second view angle position based on the second view angle, to generate the second femoral head image collected by the second camera.

[0057] In this way, for any type of femoral head image data to be recognized, two femoral head images, i.e., the first femoral head image and the second femoral head image, can be obtained through the foregoing steps. The above process can be represented as: {X-Ray, CT, MRI}→{S3}→{P1, P2}, wherein P1 represents the first femoral head image, and P2 represents the second femoral head image.

[0058] S203, calling a femoral head center identifier to process the first femoral head image and the second femoral head image, and obtaining two sets of two-dimensional coordinates output by the femoral head center identifier.

[0059] The femoral head center identifier can be pre-trained. For how to train the femoral head center identifier, reference can be made to the introduction in S103 of the previous embodiment, which will not be repeated here.

[0060] The femoral head center identifier can process and identify the first femoral head image and the second femoral head image respectively, and output the corresponding femoral head center point coordinates (u, v). That is: (u, v) = model(P), wherein model represents the trained femoral head center identifier, and P represents the first femoral head image or the second femoral head image. The whole process can also be represented as: {X-Ray, CT, MRI}→{S3}→{P1, P2}→{(u, v)}.

[0061] Further, the femoral head center identifier can output a set of two-dimensional coordinates, i.e., first coordinates (u1, v1) for the first femoral head image; and output a set of two-dimensional coordinates, i.e., second coordinates (u2, v2) for the second femoral head image. That is, the femoral head center identifier can output a set of two-dimensional coordinates for any femoral head image. In this way, a corresponding relationship between the image and the femoral head center point coordinates can be established: {P: (u, v)}, where P represents any femoral head image, (u, v) represents the two-dimensional coordinates output by the femoral head center identifier based on the image P, that is, the coordinates of the femoral head center point in the image P.

[0062] In this way, for the femoral head image data to be identified, the computer device can obtain two sets of two-dimensional coordinates after the foregoing steps. The computer device can determine the femoral head center point based on the two sets of two-dimensional coordinates.

[0063] S204, determining a first connecting line from the first perspective position to the first femoral head center point, and a second connecting line from the second perspective position to the second femoral head center point.

[0064] In the embodiments of the present application, when determining the femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates, the first connecting line can be obtained by connecting the identified first femoral head center point with the first perspective position, and the second connecting line can be obtained by connecting the identified second femoral head center point with the second perspective position. The first femoral head center point can refer to the position point corresponding to the first coordinates (u1, v1) in the foregoing example, the second femoral head center point can refer to the position point corresponding to the second coordinates (u2, v2) in the foregoing example, the first perspective position can refer to the position point where the first camera is located, and the second perspective position can refer to the position point where the second camera is located.

[0065] S205, projecting the intersection point of the extension line of the first connecting line and the extension line of the second connecting line to the three-dimensional space corresponding to the femoral head image data.

[0066] In the embodiments of the present application, the first perspective and the second perspective are orthogonal perspectives. Therefore, the extension line of the first connecting line and the extension line of the second connecting line will intersect at a point, and by re-projecting the point to the three-dimensional space, the femoral head center point in the femoral head image data can be obtained.

[0067] Specifically, the intersection point can be projected to the three-dimensional space by using the first perspective or the second perspective. For example, taking the first perspective as an example, the space matrix and the calibration matrix corresponding to the first perspective can be determined, and then the intersection point can be calculated based on the space matrix and the calibration matrix to obtain the coordinates of the femoral head center point in the three-dimensional space corresponding to the femoral head image data.

[0068] S206, identify the projection point in the three-dimensional space as the femoral head center point in the femoral head image data.

[0069] It should be noted that the size of the serial number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0070] For ease of understanding, the femoral head center point identification method provided by the embodiment of the present application will be introduced below in conjunction with a complete example.

[0071] As Figure 4 shown is a schematic diagram of a femoral head center point identification process provided by the embodiment of the present application, Figure 4 The process mainly includes the steps of identifier training, image rendering, and center point identification, which are described in detail as follows: (1) Identifier training That is Figure 4 the step S401. In this step, the femoral head center identifier can be trained using sample images.

[0072] In the embodiment of the present application, the sample images used to train the femoral head center identifier are labeled two-dimensional femoral head sample images. The above-mentioned two-dimensional femoral head sample images include a plurality of two-dimensional images obtained by performing view reconstruction on a plurality of femoral head sample image data under a plurality of viewing angles. Among them, the number of two-dimensional femoral head sample images reconstructed based on any femoral head sample image data is equal to the number of multiple viewing angles. In this way, in the case of less sample image data, the number of sample images can be increased by increasing the viewing angle during reconstruction, so as to make the model training converge faster and enhance the stability of the model.

[0073] In the process of collecting sample images, sample images can be collected for any data type, such as two-dimensional image data and three-dimensional image data, which can be applied to the identifier training of the present method.

[0074] Specifically, for any type of sample image data, it can be first rendered to a three-dimensional space, and two-dimensional images under different viewing angles can be collected. By labeling the femoral head center point of each two-dimensional image, a pair of data can be obtained, each pair of data including a two-dimensional image and the coordinate information of the femoral head center point in the two-dimensional image.

[0075] A large number of paired data can be divided into a training set, a test set and a validation set for training the femoral head center identifier. The above-mentioned femoral head center identifier can be trained based on a Gaussian heat map or a YOLO model. Subsequently, the femoral head center identifier can be used to identify the femoral head image data to obtain the corresponding femoral head center point.

[0076] (2) Image rendering That is Figure 4 the step S402. In this step, the first femoral head image and the second femoral head image for subsequent identification can be rendered for the femoral head image data to be identified.

[0077] In the embodiment of the present application, since the input data of the femoral head center identifier is a two-dimensional image, it is necessary to render any type of femoral head image data to be identified to a three-dimensional space and to collect two-dimensional images under two orthogonal viewing angles. For example, the first femoral head image under the first viewing angle and the second femoral head image under the second viewing angle are collected.

[0078] (3) Center point identification That is Figure 4 the step S403. In this step, the trained femoral head center identifier can be called to process the two-dimensional images and output two sets of two-dimensional coordinates; and the femoral head center point is determined based on the two sets of two-dimensional coordinates.

[0079] In the embodiment of the present application, for the first femoral head image and the second femoral head image that have been collected, the computer device can call the femoral head center identifier for processing and output the corresponding femoral head center point coordinates. In this way, the two-dimensional images input to the femoral head center identifier include two, and the two-dimensional coordinates output by the femoral head center identifier include two sets. Based on the two sets of two-dimensional coordinates, the viewing angle during image rendering is used to obtain a point in the three-dimensional space through projection processing, and the coordinates of the point are the femoral head center point in the femoral head image data.

[0080] Referring to Figure 5 , a schematic diagram of a femoral head center point identification device provided by an embodiment of the present application is shown, which can specifically include an acquisition module 501, a rendering module 502, a processing module 503, and a determination module 504, wherein: The acquisition module 501 is configured to acquire femoral head image data to be identified, the femoral head image data being two-dimensional image data or three-dimensional image data; The rendering module 502 is configured to render the femoral head image data to obtain a first femoral head image and a second femoral head image, the first femoral head image and the second femoral head image both being two-dimensional images; The processing module 503 is configured to call a femoral head center identifier to process the first femoral head image and the second femoral head image, and obtain two sets of two-dimensional coordinates output by the femoral head center identifier, the femoral head center identifier being trained based on a plurality of two-dimensional femoral head sample images labeled with femoral head center points; The determining module 504 is configured to determine a femoral head center point in the femoral head image data according to the two groups of two-dimensional coordinates.

[0081] In the embodiments of the present application, the rendering module 502 can be specifically configured to: render the femoral head image data to a three-dimensional space; determine a first view angle and a second view angle; collect images of the femoral head image data in the three-dimensional space based on the first view angle and the second view angle respectively, to obtain a first femoral head image and a second femoral head image.

[0082] In a possible implementation manner of the embodiments of the present application, the rendering module 502 can be further configured to: construct a three-dimensional coordinate system corresponding to the three-dimensional space; determine a target surface of the three-dimensional space and determine the first view angle in a direction facing the target surface, the target surface being a plane perpendicular to any axis in the three-dimensional coordinate system; determine the second view angle orthogonal to the first view angle.

[0083] In another possible implementation manner of the embodiments of the present application, the rendering module 502 can be further configured to: simulate an operation of collecting images of the femoral head image data by a first camera at the first view angle position based on the first view angle, to generate a first femoral head image collected by the first camera; simulate an operation of collecting images of the femoral head image data by a second camera at the second view angle position based on the second view angle, to generate a second femoral head image collected by the second camera.

[0084] In the embodiments of the present application, the two groups of two-dimensional coordinates include a first coordinate and a second coordinate, the first coordinate being a first femoral head center point coordinate output by the femoral head center identifier for the first femoral head image, the second coordinate being a second femoral head center point coordinate output by the femoral head center identifier for the second femoral head image, the first view angle and the second view angle being orthogonal view angles, and the determining module 504 can be specifically configured to: determine a first connecting line from the first view angle position to the first femoral head center point and a second connecting line from the second view angle position to the second femoral head center point, and the extension line of the first connecting line intersects the extension line of the second connecting line at a point in space; project the intersection point of the extension line of the first connecting line and the extension line of the second connecting line to a three-dimensional space corresponding to the femoral head image data; Identify the projection point in the three-dimensional space as the femoral head center point in the femoral head image data.

[0085] In a possible implementation of the embodiment of the application, the determining module 504 can also be configured to: determine a spatial matrix and a calibration matrix corresponding to the first perspective, the first perspective being directly opposite a target surface of the femoral head image data, the target surface being a plane perpendicular to any axis in a coordinate system in which the three-dimensional space is located; perform matrix calculation on the intersection point based on the spatial matrix and the calibration matrix to obtain the femoral head center point coordinates in the three-dimensional space corresponding to the femoral head image data.

[0086] In the embodiment of the application, the two-dimensional femoral head sample images used to train the femoral head center identifier include a plurality of two-dimensional images obtained by performing perspective reconstruction on a plurality of femoral head sample image data at a plurality of perspectives, and the number of two-dimensional femoral head sample images reconstructed based on any femoral head sample image data is equal to the number of the plurality of perspectives.

[0087] The femoral head center point identification device provided in the embodiment of the application can be a device or a corresponding functional module, unit component, etc. that can realize the corresponding functions. For example, the femoral head center point identification device can be a computer device or a module in the computer device introduced in the foregoing embodiments. By applying the device, each step in the foregoing method embodiments can be realized.

[0088] For the device embodiment, it is basically similar to the method embodiment, so it is described more simply, and the related parts refer to the description in the method embodiment part.

[0089] Referring to Figure 6 , a schematic diagram of a computer device provided in the embodiment of the application is shown. As shown in Figure 6 , the computer device 600 in the embodiment of the application includes a processor 610, a memory 620, and a computer program 621 stored in the memory 620 and executable on the processor 610. The processor 610 implements the steps in each of the foregoing embodiments of the femoral head center point identification method when executing the computer program 621, such as the steps S101 to S104 shown in Figure 1 . Alternatively, the processor 610 implements the functions of each module / unit in the foregoing device embodiments when executing the computer program 621, such as the functions of the modules 501 to 504 shown in Figure 5 .

[0090] Exemplarily, the computer program 621 can be divided into one or more modules / units stored in the memory 620 and executed by the processor 610 to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which can be used to describe the execution process of the computer program 621 in the computer device 600. For example, the computer program 621 can be divided into an acquisition module, a rendering module, a processing module, and a determination module, and specific functions of each module are as follows: The acquisition module is configured to acquire femoral head image data to be recognized, the femoral head image data being two-dimensional image data or three-dimensional image data. The rendering module is configured to render the femoral head image data to obtain a first femoral head image and a second femoral head image, the first femoral head image and the second femoral head image both being two-dimensional images. The processing module is configured to call a femoral head center identifier to process the first femoral head image and the second femoral head image, and obtain two sets of two-dimensional coordinates output by the femoral head center identifier, the femoral head center identifier being trained based on a plurality of two-dimensional femoral head sample images labeled with femoral head center points. The determination module is configured to determine a femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates.

[0091] The computer device 600 can be a device capable of realizing relevant steps or functions in the foregoing various method embodiments. Exemplarily, the computer device 600 can be a desktop computer, a cloud server, or the like. The computer device 600 can include, but is not limited to, a processor 610 and a memory 620. Those skilled in the art can understand that the computer device 600 can further include a bus, an input / output device, a network access device, and the like. Figure 6 The computer device 600 is only an example and does not constitute a limitation on the computer device 600, and can include more or fewer components than those shown, or combine certain components, or different components, for example, the computer device 600 can further include an input / output device, a network access device, a bus, and the like.

[0092] The processor 610 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0093] The memory 620 can be an internal storage unit of the computer device 600, for example, a hard disk or a memory of the computer device 600. The memory 620 can also be an external storage device of the computer device 600, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 620 can include both the internal storage unit and the external storage device of the computer device 600. The memory 620 is used to store the computer program 621 and other programs and data required by the computer device 600. The memory 620 can also be used to temporarily store data that has been output or will be output.

[0094] The embodiments of the present application further disclose a computer device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the method described in the foregoing various embodiments is implemented.

[0095] The embodiments of the present application further disclose a computer readable storage medium, which stores a computer program, and when the computer program is executed by a computer, the method described in the foregoing various embodiments is implemented.

[0096] The embodiments of the present application further disclose a computer program product, which comprises a computer program, and when the computer program is executed on a computer, the computer is caused to execute the method described in the foregoing various embodiments.

[0097] The above-described embodiments are merely used to illustrate the technical solutions of the present application, but not to limit the same. Although the present application is described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some of the technical features can be replaced equivalently, and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A femoral head center point identification method, characterized by, The method comprises the following steps: obtaining femoral head image data to be identified, the femoral head image data being two-dimensional image data or three-dimensional image data; rendering the femoral head image data to obtain a first femoral head image and a second femoral head image, the first femoral head image and the second femoral head image both being two-dimensional images; calling a femoral head center identifier to process the first femoral head image and the second femoral head image, and obtaining two sets of two-dimensional coordinates output by the femoral head center identifier, the femoral head center identifier being trained based on a plurality of two-dimensional femoral head sample images labeled with femoral head center points; determining a femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates.

2. The method of claim 1, wherein, The rendering of the femoral head image data to obtain a first femoral head image and a second femoral head image comprises the following steps: rendering the femoral head image data to a three-dimensional space; determining a first view angle and a second view angle; respectively collecting images of the femoral head image data in the three-dimensional space based on the first view angle and the second view angle to obtain a first femoral head image and a second femoral head image.

3. The method of claim 2, wherein, The determination of the first view angle and the second view angle comprises the following steps: constructing a three-dimensional coordinate system corresponding to the three-dimensional space; determining a target surface of the three-dimensional space and determining a first view angle in a direction directly opposite to the target surface, the target surface being a plane perpendicular to any axis in the three-dimensional coordinate system; determining a second view angle orthogonal to the first view angle.

4. The method of claim 2, wherein, The image collection of the femoral head image data in the three-dimensional space based on the first view angle and the second view angle to obtain a first femoral head image and a second femoral head image comprises the following steps: based on the first view angle, simulating the operation of collecting images of the femoral head image data by a first camera at the first view angle position to generate a first femoral head image collected by the first camera; based on the second view angle, simulating the operation of collecting images of the femoral head image data by a second camera at the second view angle position to generate a second femoral head image collected by the second camera.

5. The method according to any one of claims 2 to 4, characterized in that, The two sets of two-dimensional coordinates comprise a first coordinate and a second coordinate, the first coordinate being a first femoral head center point coordinate output by the femoral head center identifier for the first femoral head image, the second coordinate being a second femoral head center point coordinate output by the femoral head center identifier for the second femoral head image, the first view angle and the second view angle being orthogonal view angles, and the determination of the femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates comprises the following steps: determining a first connecting line from the first view angle position to the first femoral head center point and a second connecting line from the second view angle position to the second femoral head center point, the extension line of the first connecting line intersecting the extension line of the second connecting line at a point in space; projecting the intersection point of the extension line of the first connecting line and the extension line of the second connecting line to a three-dimensional space corresponding to the femoral head image data; identifying the projection point in the three-dimensional space as the femoral head center point in the femoral head image data.

6. The method of claim 5, wherein, The intersection point of the extension line of the first connecting line and the extension line of the second connecting line is projected to a space corresponding to the femoral head image data, including: determining a space matrix and a calibration matrix corresponding to the first perspective; based on the space matrix and the calibration matrix, the intersection point is calculated by matrix to obtain the femoral head center point coordinates in the three-dimensional space corresponding to the femoral head image data.

7. The method according to any one of claims 1 to 4 or 6, characterized in that, The two-dimensional femoral head sample images used to train the femoral head center identifier include a plurality of two-dimensional images obtained by perspective reconstruction of a plurality of femoral head sample image data at a plurality of perspectives, and the number of two-dimensional femoral head sample images reconstructed based on any femoral head sample image data is equal to the number of perspectives.

8. A femoral head center point identification device, comprising: comprising: an acquisition module configured to acquire femoral head image data to be identified, the femoral head image data being two-dimensional image data or three-dimensional image data; a rendering module configured to render the femoral head image data to obtain a first femoral head image and a second femoral head image, the first femoral head image and the second femoral head image being two-dimensional images; a processing module configured to call a femoral head center identifier to process the first femoral head image and the second femoral head image, and obtain two sets of two-dimensional coordinates output by the femoral head center identifier, the femoral head center identifier being trained based on a plurality of two-dimensional femoral head sample images labeled with femoral head center points; a determination module configured to determine a femoral head center point in the femoral head image data according to the two sets of two-dimensional coordinates.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, the computer device implements the method of any one of claims 1 to 7.

10. A computer program product comprising a computer program, characterized in that, When the computer program is running, the method of any one of claims 1 to 7 is executed.

Images