Pulmonary nodule rendering method and device based on lung template image

Abstract

The invention provides a lung nodule rendering method and device based on a lung template image, and the method comprises the steps: screening out a target lung image from a plurality of groups of lung images, carrying out the image segmentation of the target lung image, and obtaining a lung segment feature image of each lung segment and a lung structure feature image of a target lung tissue; based on the corresponding spatial position of each lung segment in the target lung image, determining sphere parameters of a target inscribed sphere in each lung segment; based on the lung segment feature map, the lung structure feature map and the sphere parameters, generating a lung template image including a target inscribed sphere in each lung segment; and in response to the acquired nodule information of the lung of the patient, generating a nodule feature image corresponding to each pulmonary nodule in the lung template image based on the nodule information, so as to render the pulmonary nodules in the lung template image. Through the above method, the accuracy and efficiency of rendering the pulmonary nodule are improved, and the cost of processing the lung image and rendering the image is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of medical image rendering, and in particular to a method and device for rendering lung nodules based on lung template images. Background Art

[0002] When performing imaging examinations of lung nodules on patients, 3D rendering (three-dimensional reconstruction) of lung nodules in the images based on the imaging information of the lung nodules is of great significance in clinical medical practice. Traditional two-dimensional CT images can only provide planar information and it is difficult to fully display the shape, size, edge characteristics of lung nodules and their spatial relationship with surrounding blood vessels and bronchi. 3D rendering can clearly present the three-dimensional morphology of nodules (for example, whether they are regular, whether they have lobes, burrs, etc.) and their relationship with surrounding tissues (for example, whether they invade blood vessels or bronchi, etc.) through stereo modeling.

[0003] Currently, rendering of lung nodules mainly involves performing image segmentation of each lung nodule on the patient's lung images using deep learning segmentation technology. However, this method requires independent segmentation modeling and rendering for each patient's lung image. The rendering effect is affected by the quality of CT images, which reduces the accuracy and efficiency of rendering lung nodules. In addition, deep learning segmentation technology has high requirements for computing resources, which increases the cost of processing lung images and rendering images. Summary of the Invention

[0004] In view of this, the purpose of the present application is to provide a lung nodule rendering method and device based on a lung template image, by performing image segmentation on the screened target lung image, obtaining the lung segment feature image of each lung segment in the target lung image and the lung structure feature image of other lung tissues, and determining the target inscribed sphere and its sphere parameters in each lung segment feature image, and generating the target inscribed sphere in each lung segment according to the sphere parameters in the lung template image generated by the lung segment feature image and the lung structure feature image, and when the lung nodule information of the patient's lung is received, the nodule feature image corresponding to the lung nodule information is rendered in the lung template image, thereby improving the accuracy and efficiency of rendering the lung nodules, and thereby reducing the cost of processing lung images and rendering images.

[0005] The present application provides a method for rendering lung nodules based on a lung template image, the method comprising:

[0006] Screening a target lung image from a preset plurality of lung image groups according to a preset evaluation mechanism, and performing image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image, and a lung structure feature image corresponding to the target lung tissue in the target lung image;

[0007] Based on the spatial position corresponding to each lung segment in the target lung image, determining, in each lung segment feature image, spherical parameters representing the position and size corresponding to the target inscribed sphere in each lung segment;

[0008] generating a lung template image including the target inscribed sphere in each of the lung segments based on the lung segment feature map, the lung structure feature map, and the sphere parameters;

[0009] In response to obtaining nodule information corresponding to at least one lung nodule in the patient's lungs, a nodule feature image corresponding to each lung nodule is generated in the lung template image based on the nodule information corresponding to each lung nodule, so as to render the lung nodule in the lung template image.

[0010] Furthermore, screening out a target lung image from a preset plurality of lung image groups according to a preset evaluation mechanism includes:

[0011] Respectively obtaining image parameters corresponding to each set of lung images in a plurality of preset sets of lung images;

[0012] Screening out at least one group of pending lung images whose image parameters meet preset image conditions from a plurality of preset groups of lung images;

[0013] Performing a quality assessment on each group of pending lung images based on the image parameters corresponding to each group of pending lung images to obtain a quality score corresponding to each group of pending lung images;

[0014] A pending lung image corresponding to the first quality score is screened out from the pending lung images, and the pending lung image is determined as a target lung image.

[0015] Furthermore, the target lung tissue includes at least the trachea and the lung lobe; the lung structure characteristic image includes at least the trachea characteristic image and the lung lobe characteristic image;

[0016] The performing image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image and a lung structure feature image corresponding to the target lung tissue in the target lung image, includes:

[0017] performing image segmentation on the trachea and the lung lobe in the target lung image, respectively, to obtain a trachea feature image corresponding to the trachea and a lung lobe feature image corresponding to the lung lobe;

[0018] Acquiring lung and trachea information corresponding to the lung and trachea characteristic image;

[0019] Based on the lung and trachea information, image segmentation is performed on each lung segment in the lung lobe feature image to obtain a lung segment feature image corresponding to each lung segment.

[0020] Furthermore, determining, in each lung segment feature image, spherical parameters representing the position and size of a target inscribed sphere in each lung segment based on the spatial position corresponding to each lung segment in the target lung image includes:

[0021] Establishing a three-dimensional coordinate system in the target lung image, and determining a surface coordinate set corresponding to each lung segment in the three-dimensional coordinate system;

[0022] Within the preset constraint range, set the coordinates of the center of the target inscribed sphere and the radius of the sphere to be calculated;

[0023] Based on the distance between each surface coordinate in the surface coordinate set and the undetermined circle center coordinate, the circle center coordinate and the sphere radius corresponding to the target inscribed sphere that meets the preset conditions in each lung segment are calculated, and the circle center coordinate and the sphere radius are determined as sphere parameters that characterize the position size corresponding to the target inscribed sphere in each lung segment.

[0024] Furthermore, generating a lung template image including the target inscribed sphere in each lung segment based on the lung segment feature map, the lung structure feature map, and the sphere parameters includes:

[0025] Integrating the lung segment feature map and the lung structure feature map to generate a template image corresponding to the target lung image;

[0026] For each lung segment in the template image, a target inscribed sphere corresponding to the sphere parameters is generated in each lung segment to obtain a lung template image corresponding to the target lung image.

[0027] Furthermore, in response to acquiring nodule information corresponding to at least one pulmonary nodule in the patient's lung, generating a nodule feature image corresponding to each pulmonary nodule in the lung template image based on the nodule information corresponding to each pulmonary nodule includes:

[0028] In response to obtaining nodule information corresponding to at least one pulmonary nodule in the patient's lung, determining, based on the nodule information, a first target lung segment where each pulmonary nodule is located, and size information and category information corresponding to each pulmonary nodule;

[0029] Matching and aligning each of the first target lung segments with a lung segment in the lung template image, and determining a second target lung segment where each of the lung nodules is located in the lung template image;

[0030] determining target size information corresponding to each pulmonary nodule when rendered in the second target lung segment based on the spheroid parameters and the size information in the second target lung segment;

[0031] generating a smooth nodule feature image corresponding to each lung nodule in each second target lung segment in the lung template image according to the target size information;

[0032] Based on the category information, texture blurring processing is performed on each of the smooth nodule feature images to generate a nodule feature image corresponding to each of the lung nodules in the lung template image.

[0033] Furthermore, the step of performing texture blurring processing on each of the smooth nodule feature images based on the category information to generate a nodule feature image corresponding to each of the lung nodules in the lung template image includes:

[0034] Determining a texture parameter corresponding to each of the pulmonary nodules based on the category information;

[0035] Based on the texture parameters and the preset noise parameters, a nodule texture is generated at each surface coordinate corresponding to each of the smooth nodule feature images, and edge blurring is performed on the nodule texture to obtain a textured nodule feature image corresponding to each of the smooth nodule feature images;

[0036] The texture nodule feature image is determined as a nodule feature image corresponding to each lung nodule generated in the lung template image.

[0037] The embodiment of the present application further provides a lung nodule rendering device based on a lung template image, the lung nodule rendering device comprising:

[0038] an image segmentation module, configured to screen a target lung image from a plurality of preset lung image groups according to a preset evaluation mechanism, and perform image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image, and a lung structure feature image corresponding to the target lung tissue in the target lung image;

[0039] an inscribed sphere determination module, configured to determine, in each lung segment feature image, sphere parameters representing the position and size of the target inscribed sphere in each lung segment based on the spatial position of each lung segment in the target lung image;

[0040] a template generation module, configured to generate a lung template image including the target inscribed sphere in each lung segment based on the lung segment feature map, the lung structure feature map, and the sphere parameters;

[0041] A nodule rendering module is used to generate a nodule feature image corresponding to each lung nodule in the lung template image in response to obtaining nodule information corresponding to at least one lung nodule in the patient's lungs, based on the nodule information corresponding to each lung nodule, so as to render the lung nodule in the lung template image.

[0042] Furthermore, when the image segmentation module is used to screen out a target lung image from a preset plurality of lung image groups according to a preset evaluation mechanism, the image segmentation module is used to:

[0043] Respectively obtaining image parameters corresponding to each set of lung images in a plurality of preset sets of lung images;

[0044] Screening out at least one group of pending lung images whose image parameters meet preset image conditions from a plurality of preset groups of lung images;

[0045] Performing a quality assessment on each group of pending lung images based on the image parameters corresponding to each group of pending lung images to obtain a quality score corresponding to each group of pending lung images;

[0046] A pending lung image corresponding to the first quality score is screened out from the pending lung images, and the pending lung image is determined as a target lung image.

[0047] Furthermore, the target lung tissue includes at least the trachea and the lung lobe; the lung structure characteristic image includes at least the trachea characteristic image and the lung lobe characteristic image;

[0048] When the image segmentation module is used to perform image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image and a lung structure feature image corresponding to the target lung tissue in the target lung image, the image segmentation module is used to:

[0049] performing image segmentation on the trachea and the lung lobe in the target lung image, respectively, to obtain a trachea feature image corresponding to the trachea and a lung lobe feature image corresponding to the lung lobe;

[0050] Acquiring lung and trachea information corresponding to the lung and trachea characteristic image;

[0051] Based on the lung and trachea information, image segmentation is performed on each lung segment in the lung lobe feature image to obtain a lung segment feature image corresponding to each lung segment.

[0052] Furthermore, when the inscribed sphere determination module is used to determine, in each lung segment feature image, sphere parameters representing the position and size of the target inscribed sphere corresponding to each lung segment based on the spatial position corresponding to each lung segment in the target lung image, the inscribed sphere determination module is used to:

[0053] Establishing a three-dimensional coordinate system in the target lung image, and determining a surface coordinate set corresponding to each lung segment in the three-dimensional coordinate system;

[0054] Within the preset constraint range, set the coordinates of the center of the target inscribed sphere and the radius of the sphere to be calculated;

[0055] Based on the distance between each surface coordinate in the surface coordinate set and the undetermined circle center coordinate, the circle center coordinate and the sphere radius corresponding to the target inscribed sphere that meets the preset conditions in each lung segment are calculated, and the circle center coordinate and the sphere radius are determined as sphere parameters that characterize the position size corresponding to the target inscribed sphere in each lung segment.

[0056] Furthermore, when the template generation module is used to generate a lung template image including the target inscribed sphere in each lung segment based on the lung segment feature map, the lung structure feature map, and the sphere parameters, the template generation module is used to:

[0057] Integrating the lung segment feature map and the lung structure feature map to generate a template image corresponding to the target lung image;

[0058] For each lung segment in the template image, a target inscribed sphere corresponding to the sphere parameters is generated in each lung segment to obtain a lung template image corresponding to the target lung image.

[0059] Furthermore, when the nodule rendering module is used to generate a nodule feature image corresponding to each pulmonary nodule in the lung template image based on the nodule information corresponding to each pulmonary nodule in response to obtaining nodule information corresponding to at least one pulmonary nodule in the patient's lung, the nodule rendering module is used to:

[0060] In response to obtaining nodule information corresponding to at least one pulmonary nodule in the patient's lung, determining, based on the nodule information, a first target lung segment where each pulmonary nodule is located, and size information and category information corresponding to each pulmonary nodule;

[0061] Matching and aligning each of the first target lung segments with a lung segment in the lung template image, and determining a second target lung segment where each of the lung nodules is located in the lung template image;

[0062] determining target size information corresponding to each pulmonary nodule when rendered in the second target lung segment based on the spheroid parameters and the size information in the second target lung segment;

[0063] generating a smooth nodule feature image corresponding to each lung nodule in each second target lung segment in the lung template image according to the target size information;

[0064] Based on the category information, texture blurring processing is performed on each of the smooth nodule feature images to generate a nodule feature image corresponding to each of the lung nodules in the lung template image.

[0065] Furthermore, when the nodule rendering module is used to perform texture blurring processing on each of the smooth nodule feature images based on the category information to generate a nodule feature image corresponding to each of the lung nodules in the lung template image, the nodule rendering module is used to:

[0066] Determining a texture parameter corresponding to each of the pulmonary nodules based on the category information;

[0067] Based on the texture parameters and the preset noise parameters, a nodule texture is generated at each surface coordinate corresponding to each of the smooth nodule feature images, and edge blurring is performed on the nodule texture to obtain a textured nodule feature image corresponding to each of the smooth nodule feature images;

[0068] The texture nodule feature image is determined as a nodule feature image corresponding to each lung nodule generated in the lung template image.

[0069] An embodiment of the present application also provides an electronic device, comprising: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps of the lung nodule rendering method based on the lung template image as described above are performed.

[0070] An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the computer program executes the steps of the above-mentioned method for rendering lung nodules based on lung template images.

[0071] The embodiments of the present application provide a lung nodule rendering method and device based on a lung template image, the method comprising: screening out a target lung image from a preset plurality of lung images according to a preset evaluation mechanism, and performing image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image, and a lung structure feature image corresponding to the target lung tissue in the target lung image; determining, in each lung segment feature image, spherical parameters representing the position and size of a target inscribed sphere in each lung segment based on the spatial position corresponding to each lung segment in the target lung image; generating, based on the lung segment feature map, the lung structure feature map and the spherical parameters, a lung template image including the target inscribed sphere in each lung segment; and generating, in response to obtaining nodule information corresponding to at least one lung nodule in the patient's lungs, a nodule feature image corresponding to each lung nodule in the lung template image based on the nodule information corresponding to each lung nodule, so as to render the lung nodule in the lung template image.

[0072] Compared with the existing method of rendering lung nodules mainly through deep learning segmentation technology to perform image segmentation on the patient's lung image for each lung nodule, by performing image segmentation on the screened target lung image, the lung segment feature image of each lung segment in the target lung image and the lung structure feature image of other lung tissues are obtained, and the target inscribed sphere and its sphere parameters are determined in each lung segment feature image. In the lung template image generated by the lung segment feature image and the lung structure feature image, the target inscribed sphere in each lung segment is generated according to the sphere parameters. When the lung nodule information of the patient's lung is received, the nodule feature image corresponding to the lung nodule information is rendered in the lung template image, which improves the accuracy and efficiency of rendering the lung nodules, thereby reducing the cost of processing lung images and rendering images.

[0073] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0075] Figure 1 A flowchart of a lung nodule rendering method based on a lung template image provided in an embodiment of the present application;

[0076] Figure 2 This is a schematic diagram of the effect of a target inscribed sphere corresponding to a lung segment in a lung template image provided by an embodiment of the present application;

[0077] Figure 3 A schematic diagram of the effect of generating a nodule feature image from a lung template image provided in an embodiment of the present application;

[0078] Figure 4 A comparison chart of the results of a lung nodule rendering test provided in an embodiment of the present application;

[0079] Figure 5 A schematic structural diagram of a lung nodule rendering device based on a lung template image provided in an embodiment of the present application;

[0080] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0081] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for which protection is claimed, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work falls within the scope of protection of the present application.

[0082] Research has found that currently, rendering of lung nodules mainly relies on deep learning segmentation technology to perform image segmentation of each lung nodule on the patient's lung images. However, this method requires independent segmentation modeling and rendering for each patient's lung image. The rendering effect will be affected by the quality of CT images, reducing the accuracy and efficiency of rendering of lung nodules. In addition, deep learning segmentation technology has high requirements for computing resources, which increases the cost of processing lung images and rendering images.

[0083] Based on this, an embodiment of the present application provides a lung nodule rendering method based on a lung template image. By performing image segmentation on the screened target lung image, a lung segment feature image of each lung segment in the target lung image and lung structure feature images of other lung tissues are obtained, and a target inscribed sphere and its spherical parameters are determined in each lung segment feature image. In the lung template image generated by the lung segment feature image and the lung structure feature image, a target inscribed sphere in each lung segment is generated according to the spherical parameters. When the lung nodule information of the patient's lung is received, a nodule feature image corresponding to the lung nodule information is rendered in the lung template image, thereby improving the accuracy and efficiency of rendering the lung nodules, and thereby reducing the cost of processing lung images and rendering images.

[0084] See also Figure 1 , Figure 1 This is a flow chart of a lung nodule rendering method based on a lung template image provided in an embodiment of the present application. Figure 1 As shown in , the lung nodule rendering method based on the lung template image provided in the embodiment of the present application includes:

[0085] S101. Filter out a target lung image from a preset plurality of lung image groups according to a preset evaluation mechanism, and perform image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image, and a lung structure feature image corresponding to the target lung tissue in the target lung image.

[0086] In an embodiment of the present application, the lung image represents a CT image collected by a CT examination of a human lung. By pre-evaluating the corresponding lung images collected from the lungs of multiple patients, the target lung image with the best evaluation result and the highest quality is screened out, and the target lung image is further templated, and finally a lung template image for rendering the patient's lung nodules is obtained.

[0087] The lung CT images include a set of continuous tomographic images, and each set includes multiple tomographic CT images, that is, one patient corresponds to a set of lung images, and each set of lung images includes multiple lung images.

[0088] Here, the preset evaluation mechanism refers to a method for evaluating the image quality of lung images and the lung health quality in the lung images, so as to screen out the target lung images with the best image quality and lung health quality from multiple groups of lung images through the evaluation mechanism.

[0089] In an embodiment of the present application, the target lung tissue includes at least the trachea and the lung lobes; the lung structure feature image includes at least the trachea feature image and the lung lobe feature image; the lung segment feature image can be obtained in the lung lobe feature image based on the trachea feature image.

[0090] In one embodiment of the present application, in a specific implementation, the step of screening out a target lung image from a preset plurality of lung image groups according to a preset evaluation mechanism in step S101 may include:

[0091] S1011 , respectively obtaining image parameters corresponding to each set of lung images in a plurality of preset sets of lung images.

[0092] In the embodiments of the present application, the imaging parameters include but are not limited to image resolution, signal-to-noise ratio, contrast-to-noise ratio, artifact parameters, image integrity, lung parenchymal area integrity, pathological structure interference, etc.

[0093] S1012. Filter out at least one group of pending lung images whose image parameters meet preset image conditions from the preset multiple groups of lung images.

[0094] In an embodiment of the present application, the preset imaging conditions include at least a resolution greater than a first preset value, a signal-to-noise ratio greater than a second preset value, a contrast signal-to-noise ratio greater than a third preset value, an artifact parameter of no preset artifacts, and the integrity of the lung parenchyma area is complete, and there is no interference from pathological structures.

[0095] Among them, the first preset value is generally set to 1mm 3 The second preset value is generally set to 30db, the third preset value is generally set to 5db, the preset artifacts may include motion artifacts, metal artifacts and breathing artifacts, and the above preset values ​​and preset artifacts may also be set to other values ​​and artifacts, which are no longer limited in this application.

[0096] In this step, each of the plurality of preset lung image groups is judged according to the preset image conditions, so as to screen out at least one group of pending lung images whose image parameters meet the preset image conditions from the plurality of lung image groups.

[0097] S1013. Based on the image parameters corresponding to each group of the pending lung images, perform a quality assessment on each group of the pending lung images to obtain a quality score corresponding to each group of the pending lung images.

[0098] In the embodiment of the present application, the formula for evaluating the quality of each set of pending lung images is as follows.

[0099] S=ω1*SNR+ω2*CNR+ω3*k.

[0100] Among them, S represents the quality score corresponding to each pending lung image in each group of pending lung images; SNR represents the signal-to-noise ratio in the image parameters corresponding to each pending lung image in each group of pending lung images; CNR represents the contrast-to-noise ratio in the image parameters corresponding to each pending lung image in each group of pending lung images; k represents the image completeness (percentage) in the image parameters corresponding to each pending lung image in each group of pending lung images; ω1, ω2, and ω3 represent the weight coefficients corresponding to the signal-to-noise ratio, contrast-to-noise ratio, and image completeness, respectively.

[0101] In this step, the image parameters corresponding to each group of pending lung images are input into the quality assessment formula expression, and the quality score corresponding to each pending lung image in each group of pending lung images is calculated.

[0102] S1014: Filter out a pending lung image corresponding to the first quality score from the pending lung images, and determine the pending lung image as a target lung image.

[0103] In this step, the quality scores corresponding to each pending lung image in each group of pending lung images are sorted in descending order of quantity, the quality score ranked first in the sorting order is determined as the first quality score, and the pending lung image corresponding to the first quality score is determined as the target lung image.

[0104] In one embodiment of the present application, in a specific implementation, the step of performing image segmentation on the target lung image in step S101 to obtain a lung segment feature image corresponding to each lung segment in the target lung image and a lung structure feature image corresponding to the target lung tissue in the target lung image may include:

[0105] S1015. Perform image segmentation on the trachea and the lung lobe in the target lung image, respectively, to obtain a trachea feature image corresponding to the trachea and a lung lobe feature image corresponding to the lung lobe.

[0106] In this step, during specific implementation, first, based on the preset lung and trachea segmentation model, the target lung image is segmented for lung and trachea to obtain the trachea segmentation result; then, based on the trachea segmentation result, the tracheal skeleton segmentation result is generated; finally, based on the tracheal skeleton segmentation result, the lung and trachea feature image corresponding to the lung and trachea is generated.

[0107] In which, the lung trachea feature image can be represented as a tracheal connectivity map, including a three-dimensional tangent vector from each tracheal skeleton voxel along the tracheal skeleton to the tracheal root skeleton voxel as quickly as possible; the tracheal segmentation result includes tracheal voxels and non-tracheal voxels; the tracheal skeleton segmentation result includes tracheal skeleton voxels and non-tracheal skeleton voxels.

[0108] Furthermore, in the specific implementation, the lung trachea feature image and the target lung image are merged to obtain a four-channel lung image; then, based on the preset lung lobe feature extraction model, the four-channel lung image is feature extracted to obtain the lung lobe feature image corresponding to the lung lobe.

[0109] The lung lobe feature image is represented as a three-dimensional lung lobe feature map, and each voxel in the lung lobe feature image has a corresponding lung lobe feature and corresponds to a voxel in the target lung image.

[0110] S1016: Obtain lung and trachea information corresponding to the lung and trachea characteristic image.

[0111] In the embodiment of the present application, the pulmonary trachea information includes but is not limited to the grading results of the pulmonary trachea and the extension direction of the pulmonary trachea.

[0112] S1017. Based on the lung and trachea information, perform image segmentation on each lung segment in the lung lobe feature image to obtain a lung segment feature image corresponding to each lung segment.

[0113] In this step, based on the tracheal grading results and tracheal extension direction and other tracheal information, the position and size of each lung segment are determined in the lung lobe feature image, and each lung segment is segmented in the lung lobe feature image to obtain the lung segment feature image corresponding to each lung segment.

[0114] S102. Based on the spatial position corresponding to each lung segment in the target lung image, determine, in each lung segment feature image, spherical parameters representing the position and size corresponding to the target inscribed sphere in each lung segment.

[0115] In an embodiment of the present application, since the target lung image is a three-dimensional image, the spatial position corresponding to each lung segment in the target lung image can be determined based on the lung segment feature image, that is, the surface coordinates and boundary coordinates of each lung segment in the spatial coordinate system established in the target lung image can be determined.

[0116] Here, each lung segment corresponds to a target endosphere, and the target endosphere represents the endosphere with the largest volume among multiple endospheres existing in the lung segment.

[0117] Among them, in three-dimensional space, all points on the surface of the target inscribed sphere (the inscribed sphere with the largest volume) must be located inside the lung segment, and there is no larger sphere that satisfies the condition that all points on the surface are located inside the lung segment. The distance from the center of the target inscribed sphere to the boundary of the lung segment is the farthest distance (i.e., the maximum value point of the Euclidean distance transform), and the radius of the target inscribed sphere is the distance value from the center of the sphere to the boundary of the lung segment.

[0118] In one embodiment of the present application, during specific implementation, step S102 may include:

[0119] S1021. Establish a three-dimensional coordinate system in the target lung image, and determine a surface coordinate set corresponding to each lung segment in the three-dimensional coordinate system.

[0120] In this step, the preset point in the target lung image is determined as the origin of the three-dimensional coordinate system, and based on the origin, the positive direction of the horizontal axis, the positive direction of the longitudinal axis, and the positive direction of the vertical axis are determined respectively to establish the three-dimensional coordinate system in the target lung image.

[0121] Furthermore, based on the lung segment feature image in the three-dimensional coordinate system of the target lung image, a surface coordinate set corresponding to the surface of each lung segment in the target lung image in the three-dimensional coordinate system is determined.

[0122] S1022: setting the undetermined center coordinates and the undetermined sphere radius corresponding to the target inscribed sphere to be calculated within a preset constraint range.

[0123] In an embodiment of the present application, the preset constraint range includes at least that the Euclidean distance from each coordinate in the surface coordinate set to the coordinate of the center of the circle to be determined is less than or equal to the radius of the sphere to be determined, and the boundary range corresponding to the surface of each lung segment to which the coordinate of the center of the circle to be determined belongs in the three-dimensional coordinate system.

[0124] Specifically, the expression of the preset constraint range is as follows.

[0125]

[0126] x min ≤x≤x max ;

[0127] y min ≤y≤y max ;

[0128] z min ≤z≤z max .

[0129] Wherein, c(x, y, z) represents the coordinates of the center of the target inscribed sphere of each lung segment; r represents the radius of the center of the target inscribed sphere of each lung segment; P = {p1, p2, ..., p} represents the surface coordinate set corresponding to each lung segment in the three-dimensional coordinate system; p i represents the coordinates in the surface coordinate set; (x min ,x max )、(y min ,y max )、(z min ,z max ) represents the boundary range of the surface of each lung segment corresponding to the horizontal axis, vertical axis and vertical axis in the three-dimensional coordinate system.

[0130] Furthermore, ||p i -c|| represents the Euclidean distance from each coordinate in the surface coordinate set of each lung segment in the three-dimensional coordinate system to the coordinate of the center of the circle to be determined. The expression of the Euclidean distance is as follows.

[0131]

[0132] Among them, (x i ,y i , z i ) represents each coordinate of the surface coordinate set of each lung segment in the three-dimensional coordinate system; c(x, y, z) represents the coordinates of the center of the target inscribed sphere of each lung segment to be determined; r represents the radius of the center of the target inscribed sphere of each lung segment to be determined; p i Represents a coordinate in the surface coordinate set.

[0133] S1023. Based on the distance between each surface coordinate in the surface coordinate set and the to-be-determined circle center coordinate, calculate the circle center coordinate and sphere radius corresponding to the target inscribed sphere that meets the preset conditions in each lung segment, and determine the circle center coordinate and sphere radius as sphere parameters that characterize the position size corresponding to the target inscribed sphere in each lung segment.

[0134] In the embodiment of the present application, in order to find the center coordinates and sphere radius of each target inscribed sphere, it is necessary to set the preset conditions corresponding to the target inscribed sphere, that is, to set the objective function corresponding to the target inscribed sphere.

[0135] Here, since the radius of the target inscribed sphere (the inscribed sphere with the largest volume) is equal to the shortest distance from the center of the target inscribed sphere to the boundary of each lung segment, maximizing the radius of the target inscribed sphere is equivalent to finding the coordinate point on the surface of each lung segment that maximizes the shortest distance from the center of the target inscribed sphere to the boundary of each lung segment, so as to convert the problem of solving the radius of the target inscribed sphere into a minimization-maximization problem.

[0136] Among them, the expression of the objective function corresponding to the target inscribed ball is as follows.

[0137] minimize(-min{||pi-c||}).

[0138] Furthermore, the objective function is transformed into a minimization-maximization problem, and the expression of the minimization-maximization problem is obtained as follows.

[0139] d(c)=min{||pi-c||};

[0140] r = maxd(c);

[0141] c=argmaxd(c).

[0142] Wherein, d(c) represents the minimum distance between each surface coordinate in the surface coordinate set and the coordinates of the undetermined circle center; r represents the maximum value among the minimum distances, that is, the radius of the undetermined circle center of the target inscribed sphere of each lung segment; c(x, y, z) represents the coordinates of the undetermined circle center of the target inscribed sphere of each lung segment; p i Represents a coordinate in the surface coordinate set.

[0143] Furthermore, the preset solution method is used to solve the objective function, and the center coordinates and sphere radius corresponding to the target inscribed sphere in each lung segment that meets the preset conditions (objective function) are calculated, and then the center coordinates and the sphere radius are determined as sphere parameters that characterize the position size corresponding to the target inscribed sphere in each lung segment.

[0144] The solution methods include but are not limited to particle swarm optimization algorithm and gradient ascent method.

[0145] S103 : Based on the lung segment feature map, the lung structure feature map, and the sphere parameters, generate a lung template image including the target inscribed sphere in each of the lung segments.

[0146] In an embodiment of the present application, the lung template image can reflect the structural characteristics of each lung segment and the structural characteristics of lung tissues such as the trachea and lung lobes, and each lung segment in the lung template image includes the target inscribed sphere corresponding to the lung segment.

[0147] In one embodiment of the present application, during specific implementation, step S103 may include:

[0148] S1031. Integrate the lung segment feature map and the lung structure feature map to generate a template image corresponding to the target lung image.

[0149] In this step, during specific implementation, first, the lung segment feature map, the pulmonary trachea feature map in the lung structure feature map, and the lung lobe feature map are assigned to different color channels (red, green, and blue) to generate an RGB image; then, different structures in the RGB image are distinguished by transparency to generate a multi-layer overlay image; finally, the multi-layer overlay image is subjected to feature averaging and structural enhancement processing to generate a template image corresponding to the target lung image.

[0150] S1032. For each lung segment in the template image, generate a target inscribed sphere corresponding to the sphere parameters in each lung segment to obtain a lung template image corresponding to the target lung image.

[0151] In this step, based on the spherical parameters corresponding to the target inscribed sphere of each lung segment, a corresponding target inscribed sphere is generated in each lung segment in the template image according to the spherical parameters to obtain a lung template image corresponding to the target lung image.

[0152] For example, see Figure 2 , Figure 2 This is a schematic diagram of the effect of the target inscribed sphere corresponding to the lung segment in the lung template image provided by the embodiment of the present application. Figure 2 As shown in , in the lung template image, a target inscribed sphere of each lung segment is generated corresponding to the lung segment.

[0153] S104. In response to obtaining nodule information corresponding to at least one lung nodule in the patient's lungs, based on the nodule information corresponding to each lung nodule, a nodule feature image corresponding to each lung nodule is generated in the lung template image to render the lung nodule in the lung template image.

[0154] In an embodiment of the present application, the nodule information includes at least the lung segment where the nodule is located, size information, and category information.

[0155] Here, when obtaining nodule information corresponding to at least one lung nodule in the patient's lungs, there are multiple ways of obtaining it. One is to obtain a lung image of the patient's lungs, and perform image segmentation and analysis on the lung image to determine the nodule information corresponding to at least one lung nodule in the patient's lungs; the other is to directly obtain the nodule information corresponding to at least one lung nodule in the patient's lungs.

[0156] The lung images obtained from the patient's lungs may include but are not limited to DR / X-ray chest films, lung CT images, lung PET-CT images, lung MRI images, and lung ultrasound images.

[0157] In one embodiment of the present application, during specific implementation, step S104 may include:

[0158] S1041. In response to obtaining nodule information corresponding to at least one pulmonary nodule in the patient's lung, based on the nodule information, determine the first target lung segment where each pulmonary nodule is located, as well as the size information and category information corresponding to each pulmonary nodule.

[0159] In this step, when nodule information corresponding to at least one lung nodule in the patient's lungs is obtained through lung imaging of the patient's lungs or directly, the first target lung segment where each lung nodule is located, as well as the size information and category information corresponding to each lung nodule are determined in the nodule information.

[0160] The first target lung segment represents the lung segment where each lung nodule in the patient's lung is located; the size information at least includes the radius corresponding to the lung nodule; and the category information at least includes solid and ground glass.

[0161] S1042: Match and align each of the first target lung segments with the lung segments in the lung template image to determine the second target lung segment where each of the lung nodules is located in the lung template image.

[0162] In this step, the first target lung segment where each lung nodule in the patient's lung is located is matched and aligned with the lung segment in the lung template image by radioactive transformation registration, and the second target lung segment where each lung nodule in the patient's lung is located in the lung template image is determined.

[0163] The second target lung segment represents the lung segment where each lung nodule in the patient's lungs corresponds to in the lung template image.

[0164] S1043. Based on the sphere parameters and the size information in the second target lung segment, determine target size information corresponding to each lung nodule when rendered in the second target lung segment.

[0165] In this step, during specific implementation, first, the sphere radius in the sphere parameters of the target inscribed sphere in the second target lung segment is determined; then, the sphere radius is numerically compared with the nodule radius in the size information of the lung nodules in the patient's lung in the second target lung segment to obtain a comparison result; finally, the smaller radius between the sphere radius and the nodule radius, which is the comparison result, is determined as the target radius corresponding to each lung nodule when rendered in the second target lung segment, so as to determine the target radius as the target size information corresponding to each lung nodule when rendered in the second target lung segment.

[0166] S1044. Generate a smooth nodule feature image corresponding to each lung nodule in each second target lung segment in the lung template image according to the target size information.

[0167] In this step, based on the target size information corresponding to each lung nodule when rendered in the second target lung segment, that is, the target radius corresponding to each lung nodule when rendered in the second target lung segment, a smooth nodule feature image corresponding to each lung nodule in the patient's lung is generated in each second target lung segment in the lung template image according to the target radius.

[0168] S1045. Based on the category information, perform texture blurring processing on each of the smooth nodule feature images to generate a nodule feature image corresponding to each of the lung nodules in the lung template image.

[0169] In an embodiment of the present application, based on the category information, a preset noise function is used to generate a corresponding texture in the smooth nodule feature image, and the texture is blurred to obtain a nodule feature image corresponding to each lung nodule.

[0170] In one embodiment of the present application, during specific implementation, step S1045 may include:

[0171] S10451. Based on the category information, determine the texture parameters corresponding to each of the lung nodules.

[0172] In an embodiment of the present application, the category information includes at least solidity and ground glass, etc. In a preset noise function, the amplitude value corresponding to each category information in the noise function is determined, and the amplitude value is determined as the texture parameter corresponding to each lung nodule.

[0173] For example, when the category information of the lung nodule is solid, the texture parameter corresponding to the lung nodule (the corresponding amplitude value in the noise function) can be set to 0.1; when the category information of the lung nodule is ground glass, the texture parameter corresponding to the lung nodule (the corresponding amplitude value in the noise function) can be set to 0.3.

[0174] S10452. Based on the texture parameters and the preset noise parameters, a nodule texture is generated at each surface coordinate corresponding to each of the smooth nodule feature images, and edge blurring is performed on the nodule texture to obtain a textured nodule feature image corresponding to each of the smooth nodule feature images.

[0175] In this step, based on the texture parameters and the preset noise parameters, a preset noise function is used to generate a nodule texture at each surface coordinate corresponding to each smooth nodule feature image. The expression of the noise function is shown below.

[0176] density(x k ,y k ,z k )=Dbase+A·N(f·x k ,f·y k ,f·z k ).

[0177] Among them, density(x k ,y k ,z k ) represents the nodule texture generated at each surface coordinate corresponding to each smooth nodule feature image; (x k ,y k ,z k) represents each surface coordinate corresponding to each smooth nodule feature image; Dbase represents the base parameter in the preset noise parameter; N represents the amplification parameter in the preset noise parameter; f represents the frequency parameter in the preset noise parameter; A represents the texture parameter corresponding to each lung nodule (that is, the amplitude value corresponding to each category information in the noise function).

[0178] Furthermore, the nodule boundary of the smooth nodule feature image is Gaussian smoothed to ensure a natural transition between the nodule boundary and the surrounding tissue, and the nodule texture is edge blurred to obtain a texture nodule feature image corresponding to each smooth nodule feature image.

[0179] S10453. Determine the texture nodule feature image as a nodule feature image corresponding to each lung nodule generated in the lung template image.

[0180] In this step, the obtained texture nodule feature image is determined to generate a nodule feature image corresponding to each lung nodule in the lung template image, so as to achieve the effect of rendering each lung nodule in the patient's lung in the lung template image.

[0181] For example, see Figure 3 , Figure 3 This is a schematic diagram of the effect of generating a nodule feature image from a lung template image provided in an embodiment of the present application. Figure 3 As shown in , the pulmonary nodules in the patient's lungs are generated in each lung segment in the lung template image in the form of nodule feature images, thereby achieving the effect of rendering each pulmonary nodule in the patient's lungs in the lung template image.

[0182] The lung nodule rendering method based on the lung template image provided in the embodiment of the present application performs image segmentation on the screened target lung image to obtain the lung segment feature image of each lung segment in the target lung image and the lung structure feature image of other lung tissues, and determines the target inscribed sphere and its sphere parameters in each lung segment feature image. In the lung template image generated by the lung segment feature image and the lung structure feature image, the target inscribed sphere in each lung segment is generated according to the sphere parameters. When the lung nodule information of the patient's lung is received, the nodule feature image corresponding to the lung nodule information is rendered in the lung template image, thereby improving the accuracy and efficiency of rendering the lung nodules, and thereby reducing the cost of processing lung images and rendering images.

[0183] For example, see Figure 4 , Figure 4 This is a comparison chart of the results of a lung nodule rendering test provided in an embodiment of the present application. Figure 4As shown in , when a lung nodule rendering test was performed on the nodule information corresponding to the lung nodules in the lungs of 83 patients, the average processing time for independently rendering the nodule information corresponding to the lung nodules in each patient's lung was 116 seconds, while the average processing time for rendering the lung nodules based on the lung template image for the nodule information corresponding to the lung nodules in each patient's lung was 52 seconds. It can be seen that the lung nodule rendering method based on the lung template image provided in the application embodiment has higher rendering efficiency.

[0184] See also Figure 5 , Figure 5 This is a schematic diagram of the structure of a lung nodule rendering device based on a lung template image provided in an embodiment of the present application. Figure 5 As shown in , the pulmonary nodule rendering device 500 includes:

[0185] An image segmentation module 510 is configured to screen a target lung image from a plurality of preset lung image groups according to a preset evaluation mechanism, and perform image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image, and a lung structure feature image corresponding to the target lung tissue in the target lung image;

[0186] an inscribed sphere determining module 520 for determining, in each lung segment feature image, sphere parameters representing the position and size of the target inscribed sphere in each lung segment based on the spatial position of each lung segment in the target lung image;

[0187] a template generation module 530 for generating a lung template image including the target inscribed sphere in each lung segment based on the lung segment feature map, the lung structure feature map, and the sphere parameters;

[0188] The nodule rendering module 540 is used to generate a nodule feature image corresponding to each lung nodule in the lung template image based on the nodule information corresponding to each lung nodule in response to obtaining nodule information corresponding to at least one lung nodule in the patient's lung, so as to render the lung nodule in the lung template image.

[0189] Furthermore, when the image segmentation module 510 is used to screen out a target lung image from a preset plurality of lung image groups according to a preset evaluation mechanism, the image segmentation module 510 is used to:

[0190] Respectively obtaining image parameters corresponding to each set of lung images in a plurality of preset sets of lung images;

[0191] Screening out at least one group of pending lung images whose image parameters meet preset image conditions from a plurality of preset groups of lung images;

[0192] Performing a quality assessment on each group of pending lung images based on the image parameters corresponding to each group of pending lung images to obtain a quality score corresponding to each group of pending lung images;

[0193] A pending lung image corresponding to the first quality score is screened out from the pending lung images, and the pending lung image is determined as a target lung image.

[0194] Furthermore, the target lung tissue includes at least the trachea and the lung lobe; the lung structure characteristic image includes at least the trachea characteristic image and the lung lobe characteristic image;

[0195] When the image segmentation module 510 is used to perform image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image and a lung structure feature image corresponding to the target lung tissue in the target lung image, the image segmentation module 510 is used to:

[0196] performing image segmentation on the trachea and the lung lobe in the target lung image, respectively, to obtain a trachea feature image corresponding to the trachea and a lung lobe feature image corresponding to the lung lobe;

[0197] Acquiring lung and trachea information corresponding to the lung and trachea characteristic image;

[0198] Based on the lung and trachea information, image segmentation is performed on each lung segment in the lung lobe feature image to obtain a lung segment feature image corresponding to each lung segment.

[0199] Furthermore, when the inscribed sphere determination module 520 is used to determine, in each lung segment feature image, sphere parameters representing the position and size of the target inscribed sphere corresponding to each lung segment based on the spatial position corresponding to each lung segment in the target lung image, the inscribed sphere determination module 520 is used to:

[0200] Establishing a three-dimensional coordinate system in the target lung image, and determining a surface coordinate set corresponding to each lung segment in the three-dimensional coordinate system;

[0201] Within the preset constraint range, set the coordinates of the center of the target inscribed sphere and the radius of the sphere to be calculated;

[0202] Based on the distance between each surface coordinate in the surface coordinate set and the undetermined circle center coordinate, the circle center coordinate and the sphere radius corresponding to the target inscribed sphere that meets the preset conditions in each lung segment are calculated, and the circle center coordinate and the sphere radius are determined as sphere parameters that characterize the position size corresponding to the target inscribed sphere in each lung segment.

[0203] Furthermore, when the template generation module 530 is used to generate a lung template image including the target inscribed sphere in each lung segment based on the lung segment feature map, the lung structure feature map, and the sphere parameters, the template generation module 530 is used to:

[0204] Integrating the lung segment feature map and the lung structure feature map to generate a template image corresponding to the target lung image;

[0205] For each lung segment in the template image, a target inscribed sphere corresponding to the sphere parameters is generated in each lung segment to obtain a lung template image corresponding to the target lung image.

[0206] Furthermore, the nodule rendering module 540 is configured to, in response to obtaining nodule information corresponding to at least one pulmonary nodule in the patient's lung, generate a nodule feature image corresponding to each pulmonary nodule in the lung template image based on the nodule information corresponding to each pulmonary nodule, and the nodule rendering module 540 is configured to:

[0207] In response to obtaining nodule information corresponding to at least one pulmonary nodule in the patient's lung, determining, based on the nodule information, a first target lung segment where each pulmonary nodule is located, and size information and category information corresponding to each pulmonary nodule;

[0208] Matching and aligning each of the first target lung segments with a lung segment in the lung template image, and determining a second target lung segment where each of the lung nodules is located in the lung template image;

[0209] determining target size information corresponding to each pulmonary nodule when rendered in the second target lung segment based on the spheroid parameters and the size information in the second target lung segment;

[0210] generating a smooth nodule feature image corresponding to each lung nodule in each second target lung segment in the lung template image according to the target size information;

[0211] Based on the category information, texture blurring processing is performed on each of the smooth nodule feature images to generate a nodule feature image corresponding to each of the lung nodules in the lung template image.

[0212] Furthermore, when the nodule rendering module 540 is used to perform texture blurring processing on each of the smooth nodule feature images based on the category information to generate a nodule feature image corresponding to each of the lung nodules in the lung template image, the nodule rendering module 540 is used to:

[0213] Determining a texture parameter corresponding to each of the pulmonary nodules based on the category information;

[0214] Based on the texture parameters and the preset noise parameters, a nodule texture is generated at each surface coordinate corresponding to each of the smooth nodule feature images, and edge blurring is performed on the nodule texture to obtain a textured nodule feature image corresponding to each of the smooth nodule feature images;

[0215] The texture nodule feature image is determined as a nodule feature image corresponding to each lung nodule generated in the lung template image.

[0216] The lung nodule rendering device based on the lung template image provided in the embodiment of the present application performs image segmentation on the screened target lung image to obtain the lung segment feature image of each lung segment in the target lung image and the lung structure feature image of other lung tissues, and determines the target inscribed sphere and its sphere parameters in each lung segment feature image. In the lung template image generated by the lung segment feature image and the lung structure feature image, the target inscribed sphere in each lung segment is generated according to the sphere parameters. When the lung nodule information of the patient's lung is received, the nodule feature image corresponding to the lung nodule information is rendered in the lung template image, thereby improving the accuracy and efficiency of rendering the lung nodules, and thereby reducing the cost of processing lung images and rendering images.

[0217] See also Figure 6 , Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 6 As shown in FIG, the electronic device 600 includes a processor 610 , a memory 620 and a bus 630 .

[0218] The memory 620 stores machine-readable instructions executable by the processor 610. When the electronic device 600 is running, the processor 610 communicates with the memory 620 via the bus 630. When the machine-readable instructions are executed by the processor 610, the above-mentioned Figure 1 The steps of the lung nodule rendering method based on the lung template image in the method embodiment shown are specifically implemented in the method embodiment and will not be repeated here.

[0219] The embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the computer program can execute the above-mentioned Figure 1 The steps of the lung nodule rendering method based on the lung template image in the method embodiment shown are specifically implemented in the method embodiment and will not be repeated here.

[0220] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0221] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. There may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interface, indirect coupling or communication connection of devices or units, which may be electrical, mechanical or other forms.

[0222] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0223] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0224] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0225] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The scope of protection of the present application is not limited thereto. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-mentioned embodiments within the technical scope disclosed in the present application, or perform equivalent replacements for some of the technical features thereof. These modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A lung nodule rendering method based on a lung template image, characterized in that: The method comprises: Screening a target lung image from a preset plurality of lung image groups according to a preset evaluation mechanism, and performing image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image, and a lung structure feature image corresponding to the target lung tissue in the target lung image; Based on the spatial position corresponding to each lung segment in the target lung image, determining, in each lung segment feature image, spherical parameters representing the position and size corresponding to the target inscribed sphere in each lung segment; generating a lung template image including the target inscribed sphere in each of the lung segments based on the lung segment feature map, the lung structure feature map, and the sphere parameters; In response to obtaining nodule information corresponding to at least one lung nodule in the patient's lungs, a nodule feature image corresponding to each lung nodule is generated in the lung template image based on the nodule information corresponding to each lung nodule, so as to render the lung nodule in the lung template image.

2. The method according to claim 1, characterized in that The step of screening out a target lung image from a preset plurality of lung image groups according to a preset evaluation mechanism includes: Respectively obtaining image parameters corresponding to each set of lung images in a plurality of preset sets of lung images; Screening out at least one group of pending lung images whose image parameters meet preset image conditions from a plurality of preset groups of lung images; Performing a quality assessment on each group of pending lung images based on the image parameters corresponding to each group of pending lung images to obtain a quality score corresponding to each group of pending lung images; A pending lung image corresponding to the first quality score is screened out from the pending lung images, and the pending lung image is determined as a target lung image.

3. The method according to claim 1, characterized in that The target lung tissue includes at least the trachea and the lung lobe; the lung structure characteristic image includes at least the trachea characteristic image and the lung lobe characteristic image; The performing image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image and a lung structure feature image corresponding to the target lung tissue in the target lung image, includes: performing image segmentation on the trachea and the lung lobe in the target lung image, respectively, to obtain a trachea feature image corresponding to the trachea and a lung lobe feature image corresponding to the lung lobe; Acquiring lung and trachea information corresponding to the lung and trachea characteristic image; Based on the lung and trachea information, image segmentation is performed on each lung segment in the lung lobe feature image to obtain a lung segment feature image corresponding to each lung segment.

4. The method according to claim 1, wherein The determining, based on the spatial position corresponding to each lung segment in the target lung image, in each lung segment feature image, spherical parameters representing the position and size corresponding to the target inscribed sphere in each lung segment includes: Establishing a three-dimensional coordinate system in the target lung image, and determining a surface coordinate set corresponding to each lung segment in the three-dimensional coordinate system; Within the preset constraint range, set the coordinates of the center of the target inscribed sphere and the radius of the sphere to be calculated; Based on the distance between each surface coordinate in the surface coordinate set and the undetermined circle center coordinate, the circle center coordinate and the sphere radius corresponding to the target inscribed sphere that meets the preset conditions in each lung segment are calculated, and the circle center coordinate and the sphere radius are determined as sphere parameters that characterize the position size corresponding to the target inscribed sphere in each lung segment.

5. The method according to claim 1, characterized in that Generating a lung template image including the target inscribed sphere in each lung segment based on the lung segment feature map, the lung structure feature map, and the sphere parameters comprises: Integrating the lung segment feature map and the lung structure feature map to generate a template image corresponding to the target lung image; For each lung segment in the template image, a target inscribed sphere corresponding to the sphere parameters is generated in each lung segment to obtain a lung template image corresponding to the target lung image.

6. The method according to claim 1, characterized in that In response to acquiring nodule information corresponding to at least one pulmonary nodule in the patient's lung, generating a nodule feature image corresponding to each pulmonary nodule in the lung template image based on the nodule information corresponding to each pulmonary nodule, including: In response to obtaining nodule information corresponding to at least one pulmonary nodule in the patient's lung, determining, based on the nodule information, a first target lung segment where each pulmonary nodule is located, and size information and category information corresponding to each pulmonary nodule; Matching and aligning each of the first target lung segments with a lung segment in the lung template image, and determining a second target lung segment where each of the lung nodules is located in the lung template image; determining target size information corresponding to each pulmonary nodule when rendered in the second target lung segment based on the spheroid parameters and the size information in the second target lung segment; generating a smooth nodule feature image corresponding to each lung nodule in each second target lung segment in the lung template image according to the target size information; Based on the category information, texture blurring processing is performed on each of the smooth nodule feature images to generate a nodule feature image corresponding to each of the lung nodules in the lung template image.

7. The method according to claim 6, characterized in that The step of performing texture blurring processing on each of the smooth nodule feature images based on the category information to generate a nodule feature image corresponding to each of the lung nodules in the lung template image includes: Determining a texture parameter corresponding to each of the pulmonary nodules based on the category information; Based on the texture parameters and the preset noise parameters, a nodule texture is generated at each surface coordinate corresponding to each of the smooth nodule feature images, and edge blurring is performed on the nodule texture to obtain a textured nodule feature image corresponding to each of the smooth nodule feature images; The texture nodule feature image is determined as a nodule feature image corresponding to each lung nodule generated in the lung template image.

8. A lung nodule rendering device based on a lung template image, characterized in that: The pulmonary nodule rendering device comprises: an image segmentation module, configured to screen a target lung image from a plurality of preset lung image groups according to a preset evaluation mechanism, and perform image segmentation on the target lung image to obtain a lung segment feature image corresponding to each lung segment in the target lung image, and a lung structure feature image corresponding to the target lung tissue in the target lung image; an inscribed sphere determination module, configured to determine, in each lung segment feature image, sphere parameters representing the position and size of the target inscribed sphere in each lung segment based on the spatial position of each lung segment in the target lung image; a template generation module, configured to generate a lung template image including the target inscribed sphere in each lung segment based on the lung segment feature map, the lung structure feature map, and the sphere parameters; A nodule rendering module is used to generate a nodule feature image corresponding to each lung nodule in the lung template image in response to obtaining nodule information corresponding to at least one lung nodule in the patient's lungs, based on the nodule information corresponding to each lung nodule, so as to render the lung nodule in the lung template image.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus. When the processor is running, the machine-readable instructions execute the steps of the lung nodule rendering method based on the lung template image as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the computer program executes the steps of the lung nodule rendering method based on a lung template image according to any one of claims 1 to 7.

Images