Lung segment model reconstruction method, device and electronic device

Abstract

The present disclosure relates to a lung segment model reconstruction method, device and electronic device, including: obtaining an initial chest computer tomography CT image to be processed, preprocessing the initial chest CT image, and obtaining a preprocessed target chest CT image; performing feature extraction on the target chest CT image to obtain a double lung lobe mask and a bronchial tree; classifying the bronchial tree to obtain a classification result of the bronchial tree; and generating a three-dimensional lung segment model according to the double lung lobe mask and the classification result of the bronchial tree. Therefore, the present disclosure obtains a double lung lobe mask and a bronchial tree by performing feature extraction on the target chest CT image, and generates a three-dimensional lung segment model according to the classification result of the double lung lobe mask and the bronchial tree, thereby ensuring the anatomical logic of the three-dimensional lung segment model, improving the accuracy and reliability of obtaining the three-dimensional lung segment model, and improving the efficiency in the process of lung segment model reconstruction.

Description

Technical Field

[0001] The present disclosure relates to the field of image processing technology, and in particular to a lung segment model reconstruction method, device, and electronic device. Background Art

[0002] Three-dimensional reconstruction of lung segments is of great significance in the medical field. In related technologies, three-dimensional reconstruction of lung segments can be achieved through manual labeling and segmentation, but this method is often time-consuming, labor-intensive and inefficient. Three-dimensional reconstruction of lung segments can also be achieved through deep learning, but this method is often less accurate. Therefore, how to improve the accuracy and efficiency of lung segment model reconstruction has become an urgent problem to be solved. Summary of the Invention

[0003] The present disclosure provides a lung segment model reconstruction method, device, and electronic device to at least solve the problem of low accuracy and efficiency of segment model reconstruction in the related art.

[0004] The technical solutions disclosed in this disclosure are as follows:

[0005] According to a first aspect of an embodiment of the present disclosure, a method for reconstructing a lung segment model is provided, the method comprising: obtaining an initial chest computed tomography (CT) image to be processed, preprocessing the initial chest CT image, and obtaining a preprocessed target chest CT image; performing feature extraction on the target chest CT image to obtain a bilateral lung lobe mask and a bronchial tree; classifying the bronchial tree to obtain a classification result of the bronchial tree; and generating a three-dimensional lung segment model based on the bilateral lung lobe mask and the classification result of the bronchial tree.

[0006] According to a second aspect of an embodiment of the present disclosure, a lung segment model reconstruction device is provided, comprising: an acquisition module for acquiring an initial chest electronic computed tomography (CT) image to be processed, preprocessing the initial chest CT image, and acquiring a preprocessed target chest CT image; an extraction module for performing feature extraction on the target chest CT image to acquire a bilateral lung lobe mask and a bronchial tree; a classification module for classifying the bronchial tree and acquiring a classification result of the bronchial tree; and a reconstruction module for generating a three-dimensional lung segment model based on the bilateral lung lobe mask and the classification result of the bronchial tree.

[0007] According to a third aspect of an embodiment of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the instructions to implement the lung segment model reconstruction method as described in the first aspect of the embodiment of the present disclosure.

[0008] According to the fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided. When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the lung segment model reconstruction method as described in the first aspect of the embodiment of the present disclosure.

[0009] The technical solutions provided by the embodiments of the present disclosure bring at least the following beneficial effects:

[0010] The embodiment of the present disclosure obtains an initial chest computed tomography CT image to be processed, preprocesses the initial chest CT image, obtains a preprocessed target chest CT image, performs feature extraction on the target chest CT image, obtains a mask of both lung lobes and a bronchial tree, classifies the bronchial tree, obtains the classification result of the bronchial tree, and generates a three-dimensional lung segment model based on the classification result of the mask of both lung lobes and the bronchial tree. Thus, the present disclosure obtains a mask of both lung lobes and a bronchial tree, and generates a three-dimensional lung segment model based on the classification result of the mask of both lung lobes and the bronchial tree by performing feature extraction on the target chest CT image, thereby ensuring the anatomical logic of the three-dimensional lung segment model, improving the accuracy and reliability of obtaining the three-dimensional lung segment model, reducing the cost and time in the process of reconstructing the lung segment model, and improving the efficiency in the process of reconstructing the lung segment model.

[0011] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The accompanying drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute an improper limitation of the present disclosure.

[0013] Figure 1 The figure is a flowchart of a lung segment model reconstruction method according to an exemplary embodiment.

[0014] FIG2( a ) is a schematic diagram showing a double lung lobe mask according to an exemplary embodiment.

[0015] FIG2( b ) is a schematic diagram of a bronchial tree according to an exemplary embodiment.

[0016] Figure 3 is a schematic diagram showing a classification result of a bronchial tree according to an exemplary embodiment.

[0017] Figure 4 The figure is a flowchart of a lung segment model reconstruction method according to an exemplary embodiment.

[0018] Figure 5The figure is a flowchart of a lung segment model reconstruction method according to an exemplary embodiment.

[0019] Figure 6 is a schematic diagram of a three-dimensional double lung model according to an exemplary embodiment.

[0020] Figure 7 The figure is a schematic diagram of a three-dimensional double lung lobe model according to an exemplary embodiment.

[0021] Figure 8 is a schematic diagram of a three-dimensional lung segment model according to an exemplary embodiment.

[0022] Figure 9 is a block diagram of a lung segment model reconstruction device according to another exemplary embodiment.

[0023] Figure 10 is a block diagram of an electronic device according to another exemplary embodiment. DETAILED DESCRIPTION

[0024] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0025] It should be noted that the terms "first," "second," and the like in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the numbers used in this manner are interchangeable where appropriate so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure as detailed in the appended claims.

[0026] The lung segment model reconstruction method, device, and electronic device proposed in the present disclosure are described in detail using embodiments below.

[0027] Figure 1 A flowchart of a lung segment model reconstruction method provided in an embodiment of the present disclosure.

[0028] like Figure 1 As shown, the lung segment model reconstruction method proposed in this embodiment includes the following steps:

[0029] S101, obtaining an initial chest computed tomography (CT) image to be processed, preprocessing the initial chest CT image, and obtaining a preprocessed target chest CT image.

[0030] The initial chest computed tomography (CT) image to be processed is any chest CT image that requires lung segment model reconstruction.

[0031] It should be noted that the present disclosure does not limit the specific method for obtaining the initial chest computed tomography (CT) image to be processed, and the method can be selected according to actual circumstances.

[0032] Optionally, a chest CT image may be acquired by a computed tomography device, and the acquired chest CT image may be used as the initial chest CT image.

[0033] Optionally, a chest CT image may be selected from a chest CT image database, and the selected chest CT image may be used as the initial chest CT image.

[0034] It should be noted that, in order to improve the quality of chest CT images and obtain a more accurate lung segment model, after obtaining the initial chest CT image, the initial chest CT image can be preprocessed to obtain a preprocessed target chest CT image.

[0035] In an embodiment of the present disclosure, an initial chest CT image can be binarized according to a preset grayscale threshold to obtain a first binarized image, the first binarized image can be eroded to obtain a second binarized image, the second binarized image can be closed to obtain a third binarized image, the third binarized image can be filled and inverted to obtain a first mask image, target connected areas in the first mask image that are smaller than a first preset pixel value can be eliminated to obtain a second mask image, and the background pixel values ​​in the second mask image can be set to the second preset pixel values ​​to obtain a preprocessed target chest CT image.

[0036] For example, the pixel values ​​of the initial chest CT image greater than -850 (a preset grayscale threshold) can be set to 1, and the other pixel values ​​can be set to 0 to obtain a first binary image. The first binary image can be eroded using a circular structuring element with a radius of 3 to remove some small noise points to obtain a second binary image. The second binary image can be closed using a circular structuring element with a radius of 4 to remove holes at the edges to obtain a third binary image. The third binary image can be filled in both the upper and lower directions (flood filling) and the periphery of the third binary image can be filled and inverted to obtain a first mask image. The connected domain of pixel values ​​less than 15,000 (a first preset pixel value) in the first mask image can be eliminated to remove the lowest foreign matter in the first mask image to obtain a second mask image. The background pixel values ​​in the second mask image can be set to -1,000 (a second preset pixel value) to retain the lung area to obtain a target chest CT image.

[0037] S102: Extract features from the target chest CT image to obtain the masks of both lung lobes and the bronchial tree.

[0038] In the disclosed embodiment, after acquiring the target chest CT image, feature extraction may be performed on the target chest CT image to acquire the double lung lobe masks and the bronchial tree.

[0039] Optionally, the target chest CT image may be input into a pre-trained lung feature extraction model to obtain the bilateral lung lobe masks and bronchial trees output by the lung feature extraction model.

[0040] It should be noted that the present disclosure can pre-construct an initial lung feature extraction model and construct labeled training samples to ensure that the lung feature extraction model can accurately identify the lung boundaries and the details and branches of the bronchial tree, and iteratively train the initial lung feature extraction model based on the training samples to obtain a pre-trained lung feature extraction model.

[0041] Optionally, the initial lung feature extraction model may be a U-Net model; optionally, the initial lung feature extraction model may be an improved U-Net.

[0042] For example, if the pre-trained lung feature extraction model is a U-Net model, the target chest CT image is input into the pre-trained lung feature extraction model, and the obtained lung lobe masks are shown in Figure 2 (a) and the bronchial tree is shown in Figure 2 (b).

[0043] S103: Classify the bronchial tree and obtain the classification result of the bronchial tree.

[0044] In the embodiment of the present disclosure, after the bronchial tree is obtained, the bronchial tree may be classified to obtain a classification result of the bronchial tree.

[0045] Optionally, the bronchial tree may be input into a pre-trained classification model to obtain a classification result of the bronchial tree output by the classification model, wherein the classification result is used to allocate bronchial branches in the bronchial tree to different lung segments.

[0046] It should be noted that the present disclosure can pre-construct an initial classification model and construct labeled training samples to ensure that the classification results conform to anatomical logic, and iteratively train the initial classification model based on the training samples to obtain a pre-trained classification model.

[0047] Optionally, the initial classification model may be a U-Net model; optionally, the initial classification model may be an improved U-Net model.

[0048] It should be noted that by classifying the bronchial tree, the bronchial branches of the bronchial tree can be assigned to different lung segments according to the course and anatomical position of the bronchial tree.

[0049] It should be noted that, in order to ensure the accuracy and reliability of the bronchial tree classification result, the bronchial tree classification result may be verified to obtain the final bronchial tree classification result.

[0050] For example, for the bronchial tree shown in Figure 2 (b), morphological operations and connected domain analysis can be used to remove misclassified branches to obtain the final bronchial tree classification results as shown in Figure 2 (b). Figure 3 shown.

[0051] S104: Generate a three-dimensional lung segment model based on the classification results of the lung lobe masks and the bronchial tree.

[0052] In the embodiment of the present disclosure, after obtaining the bilateral lung lobe mask and the bronchial tree, the bilateral lung lobes can be three-dimensionally reconstructed based on the bilateral lung mask to generate a three-dimensional bilateral lung lobe model. Based on the classification results, the first interface between the lung segments in each lung lobe is determined, and the three-dimensional bilateral lung lobe model is cut according to the first interface to generate a three-dimensional lung segment model.

[0053] In summary, the lung segment model reconstruction method provided by the embodiment of the present disclosure obtains an initial chest electronic computed tomography CT image to be processed, preprocesses the initial chest CT image, obtains a preprocessed target chest CT image, performs feature extraction on the target chest CT image, obtains a mask of both lung lobes and a bronchial tree, classifies the bronchial tree, obtains the classification result of the bronchial tree, and generates a three-dimensional lung segment model based on the classification result of the mask of both lung lobes and the bronchial tree. Therefore, the present disclosure obtains a mask of both lung lobes and a bronchial tree based on the classification result of the mask of both lung lobes and the bronchial tree, thereby ensuring the anatomical logic of the three-dimensional lung segment model, improving the accuracy and reliability of obtaining the three-dimensional lung segment model, reducing the cost and time in the process of lung segment model reconstruction, and improving the efficiency in the process of lung segment model reconstruction.

[0054] The following explains the specific process of generating a three-dimensional lung segment model based on the classification results of the lung lobe masks and the bronchial tree.

[0055] As a possible implementation, Figure 4 As shown, based on the above embodiment, the specific process of generating a three-dimensional lung segment model according to the classification results of the lung lobe masks and the bronchial tree includes the following steps:

[0056] S401, performing three-dimensional reconstruction of the lobes of both lungs based on the lung masks to generate a three-dimensional lung lobe model.

[0057] As a possible implementation, Figure 5 As shown, based on the above embodiment, the specific process of performing three-dimensional reconstruction of the two lung lobes according to the two lung masks to generate a three-dimensional two lung lobe model includes the following steps:

[0058] S501: Perform a three-dimensional transformation on the lung mask according to the marching cube MC algorithm to generate a three-dimensional lung model.

[0059] In the disclosed embodiment, the lung mask may be transformed into three dimensions according to a Marching Cubes (MC) algorithm to generate a three-dimensional surface mesh of the lungs, i.e., a three-dimensional lung model.

[0060] For example, for the double lung mask shown in 2(a), the specific process of generating a three-dimensional double lung model by performing three-dimensional transformation on the double lung mask includes the following steps:

[0061] S1, converting the data format: Flatten the 3D Numpy array of the lung masks into one-dimensional data and call the format conversion function "numpy_support.numpy_to_vtk" of the Visualization Toolkit (VTK) to convert the data into VTK format data;

[0062] S2, create a VTK image data object: Initialize the image data (vtkImageData) object in the VTK library, set the dimension, pixel spacing and origin of the image data, and set the VTK format data in S1 as a scalar value into the image data object;

[0063] S3, generate discrete isosurfaces: create a discrete marching cube (vtkDiscreteMarchingCubes) object in the VTK library to extract the discrete isosurfaces of the image data object, set the input image data object, and generate 20 isosurfaces from 1 to 20;

[0064] S4, smoothing: Create a vtkWindowedSincPolyDataFilter object in the VTK library to smooth the generated polygon data, and set the number of smoothing iterations and the pass bandwidth;

[0065] S5, simplification processing: Create a resampling decimation (vtkDecimatePro) object in the VTK library to simplify the smoothed polygon data;

[0066] S6, output: Figure 6 The three-dimensional model data after smoothing and simplification is returned, that is, a three-dimensional double-lung model is obtained.

[0067] S502: Determine a second interface between adjacent lobes in both lungs based on the double lung mask.

[0068] It should be noted that the double lung mask is a three-value mask containing 0, 1, and 2 elements, where 1: target lobe 1, 2: target lobe 2, and 0: other areas. The polygon data to be cut is the whole of the two lobes. There are three adjacent lobes in the right lung. The upper lobe and middle lobe of the right lung can be separated as a whole.

[0069] For example, you can create a zero array mask1 with the same shape as the double lung mask mask, and set the pixel position of category 1 in the double lung mask mask to 1. Use the ndimage.sobel function in the advanced scientific computing library scipy to determine the gradient of mask1 in three directions (x, y, z). The gradient amplitude gradient1 is obtained by determining the sum of the squares of the gradients, and the gradient amplitude is binarized. The mask2 and gradient amplitude gradient2 of the other lung lobe are determined through the above process. The boundary position boundary between the gradient amplitude gradient1 and mask2 and gradient2 and mask1 is obtained, that is, the boundary boundary = gradient1&mask2 | gradient2&mask1, get the point list points at the intersection position, and initialize the point object vtkPoints in the VTK library, traverse the input point list points, insert each point into vtkPoints, create a polygon data object vtkPolyData in the VTK library, set the point object vtkPoints to the point data of the polygon data object, create a surface reconstruction filter (vtkSurfaceReconstructionFilter) object in the VTK library, used to reconstruct the surface from scattered point data, set the input data to the polygon data object, and set the neighborhood size and sampling interval, create an isosurface filter (vtkContourFilter) object, used to extract the isosurface, set the input connection to the output port of the vtkSurfaceReconstructionFilter object, set the isovalue to 0.0, return the output data of the vtkContourFilter object, and repeat the above steps to determine the second intersection interface (reconstructed three-dimensional surface) of adjacent lobes in both lungs.

[0070] S503: Cut the three-dimensional double-lung model according to the second interface to generate a three-dimensional double-lung lobe model.

[0071] For example, the first step: use the implicit polygon data distance (vtkImplicitPolyDataDistance) in the VTK library to convert the second interface obtained in the above steps into an implicit function, create the cutting polygon data (vtkClipPolyData) in the VTK library, and set the input data to the polygon data to be cut, set the cutting function to the implicit function, and start generating the output of the clipped part; the second step: return the clipped polygon data and the clipped polygon data in the first step, the clipped polygon data and the clipped polygon data represent the outer surfaces of the two lung lobes respectively; the third step: and use the second interface as the polygon data to be cut, and the polygon data to be cut as the cutting function. The specific process can be found in the first step and will not be repeated here; the fourth step: return the clipped polygon data in the third step, that is, the inner surface of the two lung lobes; the fifth step: use the merged polygon data (vtkAppendPolyData) object in the VTK library to merge the inner and outer surfaces of the lung lobes, switch to the second interface of different adjacent lobes in the two lungs, and repeat the steps to generate the following Figure 7 The three-dimensional double lung lobe model shown.

[0072] S402: Determine a first interface between each lung segment in each lung lobe according to the classification result.

[0073] It should be noted that the bronchial branches in the bronchial tree can be assigned to different lung segments based on the classification results. A target lung segment can be arbitrarily selected to obtain the point set of bronchial branches on the target lung segment and the point set of bronchial branches on other lung segments on the same lung lobe. The point set is classified into two categories using the support vector machine (SVM) algorithm. The trajectory formed by the points on the decision plane is the interface between the target lung segments. The above process is repeated to determine the first interface between the lung segments in each lung lobe.

[0074] S403: Cut the three-dimensional double lung lobe model according to the first interface to generate a three-dimensional lung segment model.

[0075] For example, after obtaining the first interface, the specific method in "S503, cutting the three-dimensional double lung model according to the second interface" can be used to cut the three-dimensional double lung lobe model, wherein the reconstructed bronchial branch point set of the lung segment does not need to be considered for calculation, thereby generating the following Figure 8 The three-dimensional lung segment model is shown.

[0076] In summary, the lung segment model reconstruction method provided by the embodiment of the present disclosure performs three-dimensional transformation on the double lung mask according to the moving cube MC algorithm to generate a three-dimensional double lung model, determines the second interface of adjacent lobes in the double lungs according to the double lung mask, cuts the three-dimensional double lung model according to the second interface to generate a three-dimensional double lung lobe model, and determines the first interface between each lung segment in each lobe according to the classification result, cuts the three-dimensional double lung lobe model according to the first interface to generate a three-dimensional lung segment model, thereby, by sequentially acquiring the three-dimensional double lung model and the three-dimensional double lung lobe model, and cutting the three-dimensional double lung lobe model according to the first interface between each lung segment in each lobe to generate a three-dimensional lung segment model, the three-dimensional lung segment model can be automatically generated, which reduces the cost and time in the process of lung segment model reconstruction, improves the efficiency in the process of lung segment model reconstruction, and improves the accuracy and reliability of obtaining the three-dimensional lung segment model.

[0077] Figure 9 FIG. 1 is a block diagram of a lung segment model reconstruction device according to an exemplary embodiment. Figure 9 As shown, the lung segment model reconstruction device 900 of the embodiment of the present disclosure may specifically include: an acquisition module 901 , an extraction module 902 , a classification module 903 and a reconstruction module 904 .

[0078] An acquisition module 901 is configured to acquire an initial chest computed tomography (CT) image to be processed, preprocess the initial chest CT image, and acquire a preprocessed target chest CT image.

[0079] Extraction module 902, configured to perform feature extraction on the target chest CT image to obtain bilateral lung lobe masks and bronchial trees;

[0080] A classification module 903 is configured to classify the bronchial tree and obtain a classification result of the bronchial tree;

[0081] The reconstruction module 904 is configured to generate a three-dimensional lung segment model based on the double lung lobe masks and the classification results of the bronchial tree.

[0082] In one embodiment of the present disclosure, the acquisition module 901 is further used to: binarize the initial chest CT image according to a preset grayscale threshold to obtain a first binarized image; perform corrosion processing on the first binarized image to obtain a second binarized image; perform closing operation processing on the second binarized image to obtain a third binarized image; perform filling processing and inversion processing on the third binarized image to obtain a first mask image; eliminate the target connected areas in the first mask image that are smaller than a first preset pixel value to obtain a second mask image, and set the background pixel values ​​in the second mask image to the second preset pixel value to obtain the preprocessed target chest CT image.

[0083] In one embodiment of the present disclosure, the extraction module 902 is further used to: input the target chest CT image into a pre-trained lung feature extraction model to obtain the bilateral lung lobe masks and the bronchial tree output by the lung feature extraction model.

[0084] In one embodiment of the present disclosure, the extraction module 902 is further used to: input the bronchial tree into a pre-trained classification model to obtain a classification result of the bronchial tree output by the classification model, wherein the classification result is used to allocate the bronchial branches in the bronchial tree to different lung segments.

[0085] In one embodiment of the present disclosure, the apparatus 900 is further configured to verify the classification result of the bronchial tree to obtain a final classification result of the bronchial tree.

[0086] In one embodiment of the present disclosure, the reconstruction module 904 is further used to: perform three-dimensional reconstruction of the bilateral lung lobes according to the bilateral lung mask to generate a three-dimensional bilateral lung lobe model; determine the first interface between the lung segments in each lung lobe according to the classification result; and cut the three-dimensional bilateral lung lobe model according to the first interface to generate the three-dimensional lung segment model.

[0087] In one embodiment of the present disclosure, the reconstruction module 904 is further used to: perform three-dimensional transformation on the double-lung mask according to the marching cube MC algorithm to generate a three-dimensional double-lung model; determine the second interface between adjacent lobes in the double lungs according to the double-lung mask; and cut the three-dimensional double-lung model according to the second interface to generate the three-dimensional double-lung lobe model.

[0088] In the embodiment of the present disclosure, the specific manner in which each module in the lung segment model reconstruction device of the above embodiment performs operations has been described in detail in the embodiment of the lung segment model reconstruction method and will not be repeated here.

[0089] In summary, the lung segment model reconstruction device provided by the embodiment of the present disclosure obtains an initial chest electronic computed tomography CT image to be processed, preprocesses the initial chest CT image, obtains a preprocessed target chest CT image, performs feature extraction on the target chest CT image, obtains a double lung lobe mask and a bronchial tree, classifies the bronchial tree, obtains the classification result of the bronchial tree, and generates a three-dimensional lung segment model based on the classification result of the double lung lobe mask and the bronchial tree. Therefore, the present disclosure obtains a double lung lobe mask and a bronchial tree based on the classification result of the double lung lobe mask and the bronchial tree by performing feature extraction on the target chest CT image, and generates a three-dimensional lung segment model based on the classification result of the double lung lobe mask and the bronchial tree, thereby ensuring the anatomical logic of the three-dimensional lung segment model, improving the accuracy and reliability of obtaining the three-dimensional lung segment model, reducing the cost and time in the process of lung segment model reconstruction, and improving the efficiency in the process of lung segment model reconstruction.

[0090] Figure 10 is a block diagram of an electronic device 1000 according to an exemplary embodiment.

[0091] like Figure 10 As shown, the electronic device 1000 includes:

[0092] The memory 1001 and the processor 1002 , and the bus 1003 connecting different components (including the memory 1001 and the processor 1002 ), the memory 1001 stores a computer program, and when the processor 1002 executes the program, the lung segment model reconstruction method of the embodiment of the present disclosure is implemented.

[0093] Bus 1003 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0094] The electronic device 1000 typically includes a variety of electronic device-readable media, which can be any available media that can be accessed by the electronic device 1000, including volatile and non-volatile media, removable and non-removable media.

[0095] The memory 1001 may also include computer system readable media in the form of volatile memory, such as random access memory (RAM) 1004 and / or cache memory 1005. The electronic device 1000 may further include other removable / non-removable, volatile / non-volatile computer system storage media. For example only, the storage system 1006 may be used to read and write non-removable, non-volatile magnetic media ( Figure 10 Not shown, usually called a "hard drive"). Although Figure 10 Although not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), as well as an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) can be provided. In these cases, each drive can be connected to bus 1003 via one or more data media interfaces. Memory 1001 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of various embodiments of the present disclosure.

[0096] A program / utility 1008 having a set (at least one) of program modules 1007 may be stored, for example, in memory 1001. Such program modules 1007 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may include an implementation of a network environment. Program modules 1007 generally implement the functions and / or methods described in the embodiments of the present disclosure.

[0097] The electronic device 1000 may also communicate with one or more external devices 1009 (e.g., a keyboard, a pointing device, a display 1011, etc.), one or more devices that enable a user to interact with the electronic device 1000, and / or any device that enables the electronic device 1000 to communicate with one or more other computing devices (e.g., a network card, a modem, etc.). Such communication may be performed through an input / output (I / O) interface 1012. Furthermore, the electronic device 1000 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 1013. Figure 10 As shown, the network adapter 1013 communicates with other modules of the electronic device 1000 via the bus 1003. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 1000, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0098] The processor 1002 executes various functional applications and data processing by running programs stored in the memory 1001 .

[0099] It should be noted that the implementation process and technical principles of the electronic device of this embodiment can be found in the aforementioned explanation of the lung segment model reconstruction method of the embodiment of the present disclosure, and will not be repeated here.

[0100] The electronic device provided by the embodiment of the present disclosure obtains an initial chest computed tomography CT image to be processed, preprocesses the initial chest CT image, obtains a preprocessed target chest CT image, performs feature extraction on the target chest CT image, obtains a mask of both lung lobes and a bronchial tree, classifies the bronchial tree, obtains the classification result of the bronchial tree, and generates a three-dimensional lung segment model based on the classification result of the mask of both lung lobes and the bronchial tree. Therefore, the present disclosure obtains a mask of both lung lobes and a bronchial tree, and generates a three-dimensional lung segment model based on the classification result of the mask of both lung lobes and the bronchial tree, thereby ensuring the anatomical logic of the three-dimensional lung segment model, improving the accuracy and reliability of obtaining the three-dimensional lung segment model, reducing the cost and time in the process of reconstructing the lung segment model, and improving the efficiency in the process of reconstructing the lung segment model.

[0101] In order to implement the above embodiments, the present disclosure also proposes a computer-readable storage medium.

[0102] When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the lung segment model reconstruction method as described above. Optionally, the computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, or optical data storage device.

[0103] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0104] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A lung segment model reconstruction method, characterized in that: The method comprises: Acquire an initial chest computer tomography (CT) image to be processed, preprocess the initial chest CT image, and acquire a preprocessed target chest CT image; Extracting features from the target chest CT image to obtain double lung lobe masks and bronchial trees; Classifying the bronchial tree to obtain a classification result of the bronchial tree; A three-dimensional lung segment model is generated according to the classification results of the double lung lobe masks and the bronchial tree.

2. The method according to claim 1, characterized in that The preprocessing of the initial chest CT image to obtain a preprocessed target chest CT image includes: Binarize the initial chest CT image according to a preset grayscale threshold to obtain a first binary image; Performing corrosion processing on the first binary image to obtain a second binary image; Performing a closing operation on the second binary image to obtain a third binary image; Performing filling processing and inversion processing on the third binary image to obtain a first mask image; The target connected areas in the first mask image that are smaller than a first preset pixel value are eliminated to obtain a second mask image, and the background pixel values ​​in the second mask image are set to second preset pixel values ​​to obtain the preprocessed target chest CT image.

3. The method according to claim 1, characterized in that The step of extracting features from the target chest CT image to obtain double lung lobe masks and bronchial trees includes: The target chest CT image is input into a pre-trained lung feature extraction model to obtain the bilateral lung lobe masks and the bronchial tree output by the lung feature extraction model.

4. The method according to claim 1, characterized in that: The step of classifying the bronchial tree and obtaining the classification result of the bronchial tree further includes: The bronchial tree is input into a pre-trained classification model to obtain a classification result of the bronchial tree output by the classification model, wherein the classification result is used to allocate bronchial branches in the bronchial tree to different lung segments.

5. The method according to claim 4, characterized in that The method further comprises: The classification result of the bronchial tree is verified to obtain a final classification result of the bronchial tree.

6. The method according to claim 5, characterized in that The generating of a three-dimensional lung segment model according to the double lung lobe masks and the classification results of the bronchial tree also includes: Performing three-dimensional reconstruction of the bilateral lung lobes according to the bilateral lung lobes mask to generate a three-dimensional bilateral lung lobe model; Determining a first interface between each lung segment in each lung lobe according to the classification result; According to the first interface, the three-dimensional double-lung lobe model is cut to generate the three-dimensional lung segment model.

7. The method according to claim 6, characterized in that The step of performing three-dimensional reconstruction of the two lung lobes according to the two lung lobes mask to generate a three-dimensional two lung lobes model further includes: According to the marching cube MC algorithm, the double lung lobe masks are transformed into three dimensions to generate a three-dimensional double lung model; Determining a second interface between adjacent lobes in both lungs according to the masks of both lung lobes; The three-dimensional double-lung model is cut according to the second interface to generate the three-dimensional double-lung lobe model.

8. A lung segment model reconstruction device, characterized in that: The device comprises: An acquisition module is used to acquire an initial chest computer tomography (CT) image to be processed, preprocess the initial chest CT image, and acquire a preprocessed target chest CT image; An extraction module, used for performing feature extraction on the target chest CT image to obtain double lung lobe masks and bronchial trees; A classification module, used to classify the bronchial tree and obtain the classification result of the bronchial tree; A reconstruction module is used to generate a three-dimensional lung segment model according to the double lung lobe masks and the classification results of the bronchial tree.

9. An electronic device, characterized in that: include: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: in, The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 7.

Images