Lung image data processing method and device, computer equipment, medium and product

Abstract

The invention relates to a lung image data processing method and device, computer equipment, a medium and a product. The method comprises the following steps: acquiring a lung scanning image of a to-be-detected object; performing tissue segmentation processing on the lung scanning image to obtain a tissue segmentation result corresponding to the lung scanning image; performing three-dimensional reconstruction on the tissue segmentation result to obtain a three-dimensional tissue image corresponding to the lung scanning image; performing region segmentation processing on a lung region of the to-be-detected object according to the three-dimensional tissue image to obtain a region segmentation result corresponding to the lung region; and positioning a focus area in the lung area based on an area segmentation result. The method comprises the following steps: firstly, segmenting and reconstructing tissues in a lung scanning image to obtain a three-dimensional multi-tissue image, and then carrying out multi-region division on a lung region based on the three-dimensional multi-tissue image to obtain a three-dimensional lung region segmentation result, so that the three-dimensional lung region segmentation result can stereoscopically present a plurality of regions of the lung, and therefore, the lung region segmentation efficiency is improved. Fine division of lung focuses can be realized.

Description

Technical Field

[0001] The present application relates to the field of medical image processing technology, and in particular to a lung image data processing method, device, computer equipment, medium and product. Background Art

[0002] With the significant increase in the incidence of lung cancer, the requirements for detection methods for lung lesions are becoming increasingly higher. The commonly used detection method is to analyze lung tomography images through computers to identify lesion information such as lung shadows and masses.

[0003] Traditionally, a lung segmentation and lesion extraction model is used to analyze the patient's tomographic images to identify and locate lung lesions and obtain the position of the lesions in the lung area.

[0004] However, using the traditional lung segmentation and lesion extraction model, the lesion location obtained is relatively rough and the lesion positioning accuracy is low. Summary of the invention

[0005] Based on this, it is necessary to provide a lung image data processing method, device, computer equipment, computer-readable storage medium and computer program product that can improve the accuracy of lung lesion positioning in response to the above-mentioned technical problems.

[0006] In a first aspect, the present application provides a method for processing lung image data. The method comprises:

[0007] Acquire a lung scan image of the subject to be tested;

[0008] Performing tissue segmentation processing on the lung scan image to obtain a tissue segmentation result corresponding to the lung scan image;

[0009] Perform three-dimensional reconstruction on the tissue segmentation result to obtain a three-dimensional tissue image corresponding to the lung scan image;

[0010] Performing regional segmentation processing on the lung region of the subject to be tested according to the three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung region;

[0011] The lesion area in the lung region is located based on the region segmentation result.

[0012] In one embodiment, the tissue segmentation result includes a multi-tissue mask image; performing tissue segmentation processing on the lung scan image to obtain the tissue segmentation result corresponding to the lung scan image includes:

[0013] Inputting the lung scan image into a preset tissue segmentation model for tissue segmentation processing to obtain tissue label information of each pixel in the lung scan image; the preset tissue segmentation model includes a model for segmenting multiple tissues;

[0014] According to the tissue label information of each pixel in the lung scan image, a multi-tissue mask image corresponding to the lung scan image is generated.

[0015] In one embodiment, if a pixel corresponds to multiple tissue label information, a multi-tissue mask image of the lung scan image is generated according to the tissue label information of each pixel in the lung scan image, including:

[0016] For each pixel corresponding to a plurality of tissue label information in the lung scan image, based on a preset corresponding relationship, determining the label level corresponding to each tissue label information of the pixel; the preset corresponding relationship includes a corresponding relationship between different tissue label information and corresponding label levels;

[0017] Determine target tissue label information of the pixel point from multiple tissue label information of the pixel point according to label levels corresponding to each tissue label information of the pixel point;

[0018] A multi-tissue mask image of the lung scan image is generated according to the target tissue label information of each pixel in the lung scan image.

[0019] In one embodiment, the method further comprises:

[0020] For each tissue label information, pixel information corresponding to the tissue label information is determined from the multi-tissue mask image; the pixel information includes the number of pixels corresponding to the tissue label information and position information of each pixel belonging to the tissue label information;

[0021] Based on the pixel information corresponding to each tissue label information, the multi-tissue mask image is compressed to generate compressed data;

[0022] The compressed data is stored; the compressed data is used to generate a multi-tissue mask image after decompression.

[0023] In one embodiment, performing regional segmentation processing on the lung region of the subject to be tested according to the three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung region includes:

[0024] Determine a target region segmentation method; the target region segmentation method includes at least one of a lung unit segmentation method and a watershed segmentation method;

[0025] Based on the target region segmentation method, a target tissue corresponding to the target region segmentation method is acquired from the three-dimensional tissue image;

[0026] Based on the target tissue, a regional segmentation process is performed on the lung region of the object to be tested in the three-dimensional tissue image to obtain a regional segmentation result of the lung region.

[0027] In one embodiment, if the target region segmentation method is a lung unit segmentation method, the target tissue includes a bronchus, and based on the target tissue, a region segmentation process is performed on the lung region of the object to be measured in the three-dimensional tissue image to obtain a region segmentation result of the lung region, including:

[0028] Segmenting the bronchus in the three-dimensional tissue image to obtain a plurality of bronchus segments;

[0029] Based on the multiple bronchial segments, the lung region of the object to be tested in the three-dimensional tissue image is segmented into lobes to obtain multiple lobes corresponding to the lung region;

[0030] For each lung lobe, segment the lung lobe to obtain multiple lung segments corresponding to the lung lobe;

[0031] For each lung segment, lung subsegment segmentation is performed on the lung segment to obtain multiple lung subsegments corresponding to the lung segment, and the multiple lung subsegments corresponding to each lung segment are used as the regional segmentation result corresponding to the lung region.

[0032] In one embodiment, if the target region segmentation method is a watershed segmentation method, the target tissue includes at least one of a bronchus, a pulmonary artery, and a pulmonary vein. Based on the target tissue, a region segmentation process is performed on the lung region of the subject to be measured in the three-dimensional tissue image to obtain a region segmentation result of the lung region, including:

[0033] Based on the target tissue, determining a target flow domain corresponding to the target tissue in the lung region of the subject to be tested from the three-dimensional tissue image;

[0034] Segmenting the target tissue in the three-dimensional tissue image to obtain multiple tissue segments;

[0035] Based on each tissue segment, the target watershed is segmented to obtain sub-watersheds corresponding to each tissue segment, and each sub-watershed is used as the regional segmentation result corresponding to the lung region.

[0036] In one embodiment, the target area segmentation method further includes a free segmentation method; the method further includes:

[0037] In response to a user's marking operation on a three-dimensional tissue image, a marking anchor point corresponding to the marking operation is acquired, and at least one segmentation surface is generated according to the marking anchor point; the user's marking operation corresponds to a free segmentation method;

[0038] Based on each segmentation curved surface, a regional segmentation process is performed on the lung region of the object to be tested in the three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung region.

[0039] In a second aspect, the present application also provides a lung image data processing device. The device comprises:

[0040] An acquisition module, used for acquiring a lung scan image of a subject to be tested;

[0041] A tissue segmentation module is used to perform tissue segmentation processing on the lung scan image and obtain a tissue segmentation result corresponding to the lung scan image;

[0042] A three-dimensional reconstruction module is used to perform three-dimensional reconstruction on the tissue segmentation result to obtain a three-dimensional tissue image corresponding to the lung scan image;

[0043] A region segmentation module is used to perform region segmentation processing on the lung region of the object to be tested according to the three-dimensional tissue image to obtain a region segmentation result corresponding to the lung region;

[0044] The positioning module is used to locate the lesion area in the lung area based on the region segmentation result.

[0045] In a third aspect, the present application further provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the lung image data processing method in the first aspect when executing the computer program.

[0046] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the lung image data processing method in the first aspect.

[0047] In a fifth aspect, the present application further provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the lung image data processing method in the first aspect are implemented.

[0048] The above-mentioned lung image data processing method, device, computer equipment, storage medium and computer program product, the computer equipment obtains a lung scan image of the object to be tested, and performs tissue segmentation processing on the lung scan image to obtain a tissue segmentation result corresponding to the lung scan image; then, three-dimensionally reconstructs the tissue segmentation result to obtain a three-dimensional tissue image corresponding to the lung scan image; performs regional segmentation processing on the lung area of ​​the object to be tested according to the three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung area; and locates the lesion area in the lung area based on the regional segmentation result. That is to say, in the present application, a three-dimensional tissue image including at least one tissue of the object to be tested is obtained by first performing tissue segmentation processing on the lung scan image, and a three-dimensional reconstruction is performed based on the tissue segmentation result, and the lung area is divided into small-grained areas based on the three-dimensional tissue image to obtain a regional segmentation result including multiple lung areas; then, based on the divided multiple areas, the lesion tissue in the lung is located, and the lesion area where the lesion is located is determined; that is, in the present application, a three-dimensional multi-tissue image is obtained by first segmenting and reconstructing the tissue in the lung scan image, and then the lung area is divided into multiple areas based on the three-dimensional multi-tissue image to obtain a three-dimensional lung area segmentation result. Since the three-dimensional lung area segmentation result can present multiple areas of the lung in three dimensions, it can achieve a fine division of lung lesions, and is conducive to subsequent lesion removal analysis based on the located lesion area, thereby improving the efficiency of lesion removal. In addition, when the tissue segmentation results include multiple tissues inside the lungs and multiple tissues around the lungs, the mutual correlation between the tissues can be comprehensively considered to achieve accurate and reasonable division of the lung area, thereby further improving the accuracy and rationality of the lesion area and improving the accuracy of lesion positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A diagram of an application environment of a lung image data processing method in one embodiment;

[0050] Figure 2 is a schematic flow chart of a lung image data processing method in one embodiment;

[0051] Figure 3 It is a diagram showing the display effect of the tissue segmentation result and the lung scan image in various directions in one embodiment;

[0052] Figure 4 is a flow chart of a lung image data processing method in another embodiment;

[0053] Figure 5 is a flow chart of a lung image data processing method in another embodiment;

[0054] Figure 6is a schematic diagram of a multi-tissue mask image in one embodiment;

[0055] Figure 7 is a flow chart of a lung image data processing method in another embodiment;

[0056] Figure 8 is a flow chart of a lung image data processing method in another embodiment;

[0057] Fig. 9 is a flow chart of a lung image data processing method in another embodiment;

[0058] FIG10( a ) is a schematic diagram of a regional segmentation result of left lung tissue in one embodiment;

[0059] FIG10( b ) is a schematic diagram of another region segmentation result of left lung tissue in one embodiment;

[0060] Fig.11 is a flow chart of a lung image data processing method in another embodiment;

[0061] Fig.12 is a flow chart of a lung image data processing method in another embodiment;

[0062] Fig.13 is a schematic diagram of the organizational structure of a lung image data processing method in one embodiment;

[0063] Fig.14 is a schematic structural diagram of lung tissue segmentation in one embodiment;

[0064] Fig.15 is a structural block diagram of a lung image data processing device in one embodiment;

[0065] Fig.16 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0067] The incidence of lung cancer has shown a clear upward trend in recent years. In some countries, the incidence of lung cancer ranks first among malignant tumors. The commonly used detection method is to use computers to analyze chest and lung tomography images to identify lung shadows, mass morphology, lung density and other information for diagnosis. With the continuous advancement of computer technology and image analysis algorithms, there are currently many software and algorithms that can automatically identify lung tissue and lung lesions in images, such as 3D Slicer, Mimics and other segmentation algorithms, which can segment single tissues such as the heart, lungs, arteries, and bronchi in the chest and lungs. However, current analysis tools and methods mainly focus on single lung segmentation and lesion extraction methods, while ignoring a systematic positioning and functional analysis of the lungs in the entire chest cavity; as an important tissue organ in the chest cavity, the lungs are also closely related to the bronchi, arteries and veins, aorta and other tissues in the chest and abdomen. At the same time, the tissue structure of the lungs, such as the direction of the lobes, segments, oblique fissures, and horizontal fissures of the lungs, is also crucial for the positioning and analysis of lung lesions.

[0068] Currently, there are many studies on single lung tissue, but there is no method that combines the tissue inside the lung with the tissue around the lung as a functional whole for analysis. At the same time, most of the research focuses on the segmentation algorithm of lung tissue and the visualization of tissue, and there are few studies related to the management of multiple tissues. Therefore, a single lung tissue and lesion identification method can no longer meet the needs of systematic analysis of the lungs. A multi-dimensional chest and lung tissue management and analysis method is needed to achieve accurate positioning of lung lesions.

[0069] Based on this, the embodiment of the present application proposes a lung image data processing method, which aims to focus on a systematic analysis method of the lung and its surrounding tissues by performing multi-dimensional tissue segmentation management and analysis of the lung tissue and the tissues surrounding the lungs. By efficiently managing and displaying masks of multiple tissues and performing multi-dimensional segmentation of lung units, a more comprehensive lung lesion analysis method is provided to achieve accurate positioning of lung lesions.

[0070] The lung image data processing method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the computer device 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. Among them, the computer device 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers and other medical terminal devices. The server 104 can be implemented with an independent server or a server cluster consisting of multiple servers, including but not limited to an image storage server.

[0071] In one embodiment, Figure 2 As shown, a lung image data processing method is provided, and the method is applied to Figure 1 The computer device in the example is used to illustrate, including the following steps:

[0072] Step 201, obtaining a lung scan image of a subject to be tested.

[0073] Among them, the lung scan image of the object to be tested includes but is not limited to CT scan images, MR scan images, PET scan images, etc., and the CT scan image may include a plain CT image or an enhanced CT image, etc.; the type of the lung scan image of the object to be tested is not specifically limited in the embodiment of the present application.

[0074] Exemplarily, the computer device can send a request for obtaining a lung scan image of the object to be tested to the server based on a trigger operation of the user, and receive the lung scan image of the object to be tested sent by the server; the computer device can also pre-receive a lung scan image of at least one object sent by the server, and store the received lung scan images of each object in a local database of the computer device, so as to obtain the lung scan image of the object to be tested from the local database, etc.

[0075] Step 202: Perform tissue segmentation processing on the lung scan image to obtain a tissue segmentation result corresponding to the lung scan image.

[0076] Among them, tissue segmentation may include segmenting tissue inside the lungs, and may also include segmenting tissue around the lungs; furthermore, tissue segmentation may segment all tissues in the lung scan image, and may also segment part of the tissue in the lung scan image; the tissue to be segmented may include preset tissues, and may also include user-defined tissues, etc. If the user does not select the tissue to be segmented, the computer device may perform tissue segmentation processing on the lung scan image of the subject according to the default preset tissue.

[0077] Exemplarily, preset tissues may include but are not limited to left lung tissue, right lung tissue, pulmonary artery tissue, pulmonary vein tissue, bronchial tissue, lesion tissue, etc.; user-defined tissues may include but are not limited to esophagus, heart, aorta, superior vena cava, bone tissue, etc.; that is, preset tissues may include at least one tissue inside the lungs, and user-defined tissues may include at least one tissue around the lungs other than the internal lung tissues, etc. Of course, since the preset tissues may be tissues segmented by default by the computer device, and the user-defined tissues are tissues that the user wants to segment additionally, therefore, for the definition of the preset tissues, it can define all or part of the tissues in the lung scan image as preset tissues, i.e., default tissues, according to actual usage requirements; and the remaining tissues belong to tissues that the user can customize and add. In the embodiments of the present application, there is no specific limitation on the tissue types contained in the preset tissues and the user-defined tissues.

[0078] In addition, it should be noted that tissues may include normal organs or tissues, and may also include abnormal tissues, such as lesions, etc. Lesions include but are not limited to lung nodule lesions, lung shadows, masses, lymph nodes, etc.

[0079] Exemplarily, a computer device may perform tissue segmentation processing on tissues to be segmented in a lung scan image based on a preset tissue segmentation model, thereby obtaining a tissue segmentation result corresponding to the lung scan image, wherein the tissue segmentation result includes segmentation results corresponding to each tissue to be segmented; optionally, the tissue segmentation result corresponding to the lung scan image may be a multi-tissue mask image including segmentation results corresponding to each tissue to be segmented, wherein different tissues to be segmented are marked based on different mask labels.

[0080] Exemplarily, for a computer device, it can not only support automatic tissue segmentation, that is, automatic tissue segmentation processing using a preset tissue segmentation model, but also support semi-automatic tissue segmentation; for example: semi-automatic tissue segmentation processing can be based on the user-marked position for segmentation, that is, the user manually marks the position of the tissue, and the computer device performs segmentation processing on the tissue in the lung scan image based on the information manually marked by the user to obtain a tissue segmentation result. It should be noted that the tissue segmentation results obtained by different segmentation methods can be fused to obtain a complete tissue segmentation result, which includes the automatically segmented tissue and the semi-automatically segmented tissue.

[0081] In addition, for the segmented tissues, that is, each tissue in the tissue segmentation results, it can also support contour editing, tissue growth and interest drawing, etc. By fully or specifically extracting the tissues in the abdomen, it is possible to more accurately identify and locate lesions and formulate treatment plans through the relative position relationship between tissues on the lung scan image.

[0082] Step 203: perform three-dimensional reconstruction on the tissue segmentation result to obtain a three-dimensional tissue image corresponding to the lung scan image.

[0083] The lung scan image of the subject to be tested may include a lung scan image of the subject to be tested in at least one orientation; the at least one orientation may include at least one of a transverse orientation, a coronal orientation and a sagittal orientation.

[0084] Exemplarily, when the lung scan image of the subject to be tested includes a lung scan image in one orientation, the computer device can reconstruct the lung scan images corresponding to the other two orientations using an interpolation algorithm based on the lung scan image in the orientation; then, three-dimensional reconstruction can be performed based on the tissue segmentation result corresponding to the lung scan image in the orientation and the lung scan images in various orientations, thereby obtaining a three-dimensional tissue image corresponding to the lung scan image. Optionally, the three-dimensional tissue image can be a three-dimensional tissue image in which each tissue is marked based on a grid model, and the computer device can construct a grid model of the tissue based on the position point information of each tissue in the tissue segmentation result corresponding to the lung scan image in the original volume data of the lung scan images in various orientations, thereby obtaining a three-dimensional tissue image including multiple tissues.

[0085] Furthermore, the three-dimensional tissue image can be output so as to intuitively and three-dimensionally display multiple tissues of the abdomen of the subject to be tested to the user. In addition, the detailed information of each tissue can be displayed separately through the lung scanning images in the transverse, coronal and sagittal directions; Figure 3 As described above, the left side is a three-dimensional tissue image, and the right side is a transverse, coronal, and sagittal lung scan image. The lung scan images and three-dimensional tissue images in various directions can provide a complete abdominal system to assist users in analysis.

[0086] Step 204 , performing region segmentation processing on the lung region of the subject to be tested according to the three-dimensional tissue image, and obtaining a region segmentation result corresponding to the lung region.

[0087] The region segmentation is used to indicate segmenting the whole lung tissue of the object to be tested, where the whole lung tissue includes the left lung tissue and the right lung tissue, that is, the left lung tissue and the right lung tissue are segmented respectively.

[0088] Exemplarily, when performing regional segmentation processing on the lung region of the object to be tested according to the three-dimensional tissue image, the left lung tissue and the right lung tissue in the three-dimensional tissue image can be segmented separately based on the preset anatomical structure of the lung region, so as to obtain the regional segmentation result corresponding to the lung region. The preset anatomical structure may include an anatomical structure composed of lobes-segments-subsegments; for example: the lungs can be divided into 5 lobes and 18 segments, the lobes are composed of segments, and the segments are composed of subsegments; the left lung tissue can be divided into upper and lower lobes by the oblique fissure, and the right lung tissue can be divided into upper, middle and lower lobes by the oblique fissure and the horizontal fissure; the left lung tissue includes 8 segments, and the right lung tissue includes 10 segments; further, each segment can be further divided into multiple subsegments.

[0089] Exemplarily, the computer device may also use a watershed segmentation method to perform watershed segmentation processing on the left lung tissue and the right lung tissue in the three-dimensional tissue image respectively, so as to obtain a regional segmentation result corresponding to the lung region.

[0090] Of course, other region segmentation methods may also be used when performing lung region segmentation processing, such as free segmentation method, etc., and the embodiments of the present application do not specifically limit this.

[0091] Step 205: locate the lesion region in the lung region based on the region segmentation result.

[0092] The regional segmentation result includes multiple divided regions corresponding to the left lung and multiple divided regions corresponding to the right lung, and the divided regions can be refined into lung subsegment regions.

[0093] Exemplarily, after obtaining the regional segmentation result of the lung region of the object to be tested, the computer device can locate the lesion tissue in the tissue segmentation result based on the regional segmentation result to obtain the lesion area corresponding to each lesion tissue. Optionally, the computer device can determine the target divided area corresponding to each lesion tissue according to the position information of each lesion tissue and the position information of each divided area in the regional segmentation result, and use the target divided area corresponding to each lesion tissue as the lesion area corresponding to each lesion tissue, thereby obtaining at least one lesion area in the lung region of the object to be tested; wherein one divided area may include one or more lesion tissues.

[0094] In the above-mentioned lung image data processing method, the computer device obtains a lung scan image of the object to be tested, and performs tissue segmentation processing on the lung scan image to obtain a tissue segmentation result corresponding to the lung scan image; then, the tissue segmentation result is three-dimensionally reconstructed to obtain a three-dimensional tissue image corresponding to the lung scan image; regional segmentation processing is performed on the lung area of ​​the object to be tested based on the three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung area; and the lesion area in the lung area is located based on the regional segmentation result. That is to say, in the present application, a three-dimensional tissue image including at least one tissue of the object to be tested is obtained by first performing tissue segmentation processing on the lung scan image, and a three-dimensional reconstruction is performed based on the tissue segmentation result, and the lung area is divided into small-grained areas based on the three-dimensional tissue image to obtain a regional segmentation result including multiple lung areas; then, based on the divided multiple areas, the lesion tissue in the lung is located, and the lesion area where the lesion is located is determined; that is, in the present application, a three-dimensional multi-tissue image is obtained by first segmenting and reconstructing the tissue in the lung scan image, and then the lung area is divided into multiple areas based on the three-dimensional multi-tissue image to obtain a three-dimensional lung area segmentation result. Since the three-dimensional lung area segmentation result can present multiple areas of the lung in three dimensions, it can achieve a fine division of lung lesions, and is conducive to subsequent lesion removal analysis based on the located lesion area, thereby improving the efficiency of lesion removal. In addition, when the tissue segmentation results include multiple tissues inside the lungs and multiple tissues around the lungs, the mutual correlation between the tissues can be comprehensively considered to achieve accurate and reasonable division of the lung area, thereby further improving the accuracy and rationality of the lesion area and improving the accuracy of lesion positioning.

[0095] Figure 4 FIG. 1 is a flow chart of a lung image data processing method in another embodiment. This embodiment relates to an optional implementation process in which a computer device performs tissue segmentation processing on a lung scan image to obtain a tissue segmentation result corresponding to the lung scan image when the tissue segmentation result includes multiple tissue mask images. Based on the above embodiment, Figure 4 As shown, the above step 202 includes:

[0096] Step 401: Input the lung scan image into a preset tissue segmentation model for tissue segmentation processing to obtain tissue label information of each pixel in the lung scan image.

[0097] The preset tissue segmentation model includes a model for segmenting multiple tissues, that is, the preset tissue segmentation model can be a multi-tissue segmentation model. Exemplarily, the multiple tissues can include preset tissues and user-defined tissues, and both the preset tissues and the user-defined tissues can include one or more lung tissues and lung surrounding tissues.

[0098] Exemplarily, the computer device can input the lung scan image of the subject to be tested into the preset tissue segmentation model to perform tissue segmentation processing on multiple tissues in the lung scan image, thereby obtaining tissue label information for each pixel in the lung scan image; that is, when the preset tissue segmentation model performs segmentation processing on each tissue in the lung scan image, it actually performs tissue classification on each pixel of the lung scan image, determines the tissue category to which each pixel belongs, and accordingly obtains tissue label information corresponding to each pixel; since the lungs include a variety of different tissues such as blood vessels and bronchi, the pixels inside the lungs may not only belong to lung tissue, but may also belong to vascular tissue, or bronchial tissue, lesion tissue, etc. Therefore, the pixels in the lung scan image may correspond to one or more tissue label information.

[0099] Step 402: Generate a multi-tissue mask image corresponding to the lung scan image according to the tissue label information of each pixel in the lung scan image.

[0100] Exemplarily, after obtaining the tissue label information of each pixel in the lung scan image, multi-tissue mask data corresponding to the lung scan image can be generated based on the tissue label information of each pixel; further, based on the multi-tissue mask data, a multi-tissue mask image corresponding to the lung scan image can be generated, and the multi-tissue mask image can be used for output and displayed on a display so as to intuitively reflect the relative positional relationship between the various tissues.

[0101] Since each pixel corresponds to only one pixel value when the image is displayed, when a pixel corresponds to multiple tissue label information, it is necessary to determine one tissue label information from the multiple tissue label information corresponding to the pixel as the target tissue label information of the pixel; then, a multi-tissue mask image corresponding to the lung scan image can be generated according to the target tissue label information corresponding to each pixel and the pixel value corresponding to each tissue label information; it should be noted that in order to ensure that each tissue can be fully displayed, when determining the target tissue label information of each pixel, it should be ensured that each pixel in the same connected domain belongs to the same tissue label information to avoid the cluttered display of multiple tissues.

[0102] In this embodiment, the lung scan image is input into a preset tissue segmentation model for tissue segmentation processing to obtain tissue label information of each pixel in the lung scan image; then, a multi-tissue mask image corresponding to the lung scan image is generated based on the tissue label information of each pixel in the lung scan image; wherein the preset tissue segmentation model includes a model for segmenting multiple tissues; that is, in this embodiment, a multi-tissue segmentation model is adopted to realize automatic segmentation processing of multiple different tissues in the lung scan image, thereby obtaining a multi-tissue mask image, which can improve the accuracy and efficiency of tissue segmentation.

[0103] Figure 5 FIG. 1 is a flow chart of a lung image data processing method in another embodiment. This embodiment relates to an optional implementation process of generating a multi-tissue mask image of a lung scan image according to the tissue label information of each pixel in the lung scan image if the pixel corresponds to multiple tissue label information. Based on the above embodiment, Figure 5 As shown, the above step 402 includes:

[0104] Step 501 : for each pixel corresponding to a plurality of tissue label information in the lung scan image, determine the label level corresponding to each tissue label information of the pixel based on a preset corresponding relationship.

[0105] The preset corresponding relationship includes the corresponding relationship between different organization label information and corresponding label levels.

[0106] For multiple tissues in the lung scan image, since one tissue may also include multiple other tissue structures, when displaying a multi-tissue mask image, only the tissue label information of one tissue can be displayed for one pixel. Therefore, it is necessary to determine a target tissue label information from the multiple tissue label information corresponding to the pixel to generate a multi-tissue mask image; that is, for multiple tissues in the lung scan image, there may be overlap between the tissues. Therefore, when displaying each tissue, it is necessary to process the overlapping tissues. It should be noted that the tissue overlap mentioned here refers to the same layer of lung scan image, that is, the chest and lung tissues in the same layer of lung scan image may overlap. For example, the same voxel (i.e., pixel) may be marked as lung tissue and also marked as pulmonary artery vascular tissue, or even marked as lung nodule tissue at the same time.

[0107] Based on this, in this embodiment, in order to display the overlap of multiple tissues on the same layer of lung scan image, a hierarchical management method can be used to display each tissue; illustratively, based on the principle of hierarchical display, different label levels can be preset for the tissue label information corresponding to each tissue to form a preset corresponding relationship; for example: the tissue can be divided into 7 layers according to category and extraction method, the 0th layer is the basic layer, including unidentified tissues and bed boards, and the tissues in this layer will hardly change; the 1st layer includes the left lung and the right lung; the second layer includes the pulmonary arteries and veins and bronchi; the third layer includes manually added tissues, such as the esophagus, heart, aorta, superior vena cava, bones, etc.; the fourth layer includes lesions; the fifth layer includes lymph node tissue, which can be added multiple times and appended to the same tissue label information after addition; the sixth layer includes custom tissues, and users can define their own names and arbitrarily identify them as unincluded tissues. There can be multiple layers, which are managed separately.

[0108] Furthermore, based on the above preset corresponding relationship, when a pixel point includes multiple tissue label information, the label level corresponding to each tissue label information of the pixel point can be determined.

[0109] Step 502 : determining target tissue label information of the pixel from a plurality of tissue label information of the pixel according to label levels corresponding to the tissue label information of the pixel.

[0110] Based on the above example, when the organization's label level is higher (such as level 7), it means that the organization is an upper-level organization and should be displayed. When the organization's label level is lower (such as level 0), it means that the organization is a lower-level organization and may be covered by an upper-level organization. Based on this, the organization label information with the highest label level can be selected as the target organization label information of the pixel point according to the label levels corresponding to each organization label information of the pixel point.

[0111] Step 503: Generate a multi-tissue mask image of the lung scan image according to the target tissue label information of each pixel in the lung scan image.

[0112] For example, when the target tissue label information of each pixel in the lung scan image is determined, a multi-tissue mask image of the lung scan image can be generated according to the target tissue label information of each pixel and the pixel value corresponding to the tissue label information, such as Figure 6 Here, for a pixel point with only one tissue label information, the tissue label information can be directly used as the target tissue label information of the pixel point.

[0113] Exemplarily, the pixel value corresponding to the tissue label information may include the grayscale value of the pixel point, and may also include the transparency of the pixel point, etc. Based on this, different levels of tissues in the generated multi-tissue mask image can be displayed with different transparencies, which can improve the display effect of tissue layered display.

[0114] Exemplarily, when displaying a multi-tissue mask image, the multi-tissue mask image can also be converted into a multi-tissue grid image, i.e., a mesh image, and the multi-tissue grid image is output and displayed; wherein the mesh image is a three-dimensional grid in which the tissue of the specified label in the mask data is converted into a three-dimensional grid, so as to more intuitively display multiple tissues in the chest and lungs from a three-dimensional perspective; each tissue is an independent mesh model, and the positional relationship between the tissues can be assembled together through a mesh node; optionally, in addition to the mesh data of the tissue, the mesh node can also contain the transformation matrix of the tissue mesh model relative to the world coordinate system, so as to facilitate the management of multiple tissues and display them in a unified coordinate system at the same time, so that when the three-dimensional multi-tissue grid image is rotated, etc., the relative positional relationship of each tissue in space can be ensured to remain unchanged. In addition, in order to better distinguish each tissue on the multi-tissue grid image, the attribute management of each tissue can also be performed, including but not limited to the color, transparency, visibility, etc. of the tissue, and the attribute management of the tissue can record the display status attribute of the current tissue, and this display status attribute can be realized by setting the attributes of the mask and mesh node to display the display status of each tissue on the multi-tissue grid image.

[0115] In this embodiment, for each pixel point corresponding to multiple tissue label information in the lung scan image, the label level corresponding to each tissue label information of the pixel point can be determined based on a preset correspondence relationship; then, according to the label levels corresponding to each tissue label information of the pixel point, the target tissue label information of the pixel point is determined from the multiple tissue label information of the pixel point; and according to the target tissue label information of each pixel point in the lung scan image, a multi-tissue mask image of the lung scan image is generated; wherein the preset correspondence relationship includes the correspondence between different tissue label information and the corresponding label level; that is, in this embodiment, by setting the label levels corresponding to different tissues, a tissue hierarchical management method is adopted to solve the overlapping display problem of multiple tissues, and obtain a multi-tissue mask image, which can improve the display effect of each tissue.

[0116] Figure 7 FIG. 1 is a flow chart of a lung image data processing method in another embodiment. This embodiment relates to an optional implementation process of a computer device compressing and storing multi-tissue mask images. Based on the above embodiment, Figure 7 As shown, the above method also includes:

[0117] Step 701 : for each tissue label information, determine the pixel information corresponding to the tissue label information from the multi-tissue mask image.

[0118] The pixel information corresponding to the tissue label information may include the number of pixels corresponding to the tissue label information and position information of each pixel belonging to the tissue label information.

[0119] Exemplarily, the computer device can determine the number of pixels belonging to each tissue label information and the position information of each pixel belonging to the tissue label information according to the target tissue label information corresponding to each pixel in the multi-tissue mask image.

[0120] Step 702 : compress the multi-tissue mask image based on the pixel information corresponding to each tissue label information to generate compressed data.

[0121] Exemplarily, a computer device may store the number of pixels corresponding to each tissue label information and the position information of each pixel belonging to the tissue label information in the form of an array, thereby compressing the image data into array data, thereby generating compressed data corresponding to a multi-tissue mask image.

[0122] Step 703, storing the compressed data; the compressed data is used to generate a multi-tissue mask image after decompression.

[0123] Exemplarily, the computer device may store the compressed data corresponding to the multi-tissue mask image in local storage, so that when the user chooses to display the multi-tissue mask image, the computer device may obtain the compressed data corresponding to the multi-tissue mask image from the current storage, and restore the multi-tissue mask image based on the compressed data, and output it for display. Optionally, the computer device may also send the compressed data corresponding to the multi-tissue mask image to a server or other storage device for storage, etc., which is not specifically limited in the embodiments of the present application.

[0124] Exemplarily, based on the above-mentioned compression storage method for multi-tissue mask images, the same method can also be used to compress the multi-tissue mask data, which can further reduce the memory occupancy compared to the multi-tissue mask data stored in the form of an image matrix.

[0125] In this embodiment, for each tissue label information, the pixel information corresponding to the tissue label information is determined from the multi-tissue mask image, including the number of pixels corresponding to the tissue label information and the position information of each pixel belonging to the tissue label information; based on the pixel information corresponding to each tissue label information, the multi-tissue mask image is compressed to generate compressed data; the compressed data is stored; the compressed data can be used to generate the multi-tissue mask image after decompression. That is, in this embodiment, for the problem of excessive memory usage due to too many tissues in the multi-label mask image, the image data can be compressed and managed by recording the labels and the corresponding number of pixels and the position of each pixel in the mask image according to the rule that there are a large number of repetitions of each label in the mask image, thereby reducing the memory size occupied by the multi-tissue mask image and saving memory usage.

[0126] Figure 8 FIG. 1 is a flow chart of a lung image data processing method in another embodiment. This embodiment relates to an optional implementation process in which a computer device performs regional segmentation processing on the lung region of a subject to be tested according to a three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung region. Based on the above embodiment, Figure 8 As shown, the above step 204 includes:

[0127] Step 801, determine a target region segmentation method; the target region segmentation method includes at least one of a lung unit segmentation method and a watershed segmentation method.

[0128] Among them, the lung unit segmentation method is a segmentation method based on the tissue structure of the lung lobe-lung segment-lung subsegment; the watershed segmentation method is a segmentation method based on the pulmonary artery or pulmonary vein watershed, including the venous watershed method, the arterial watershed method, etc.

[0129] Exemplarily, the computer device may use a plurality of different region segmentation methods to respectively perform segmentation processing on the lung region in the three-dimensional tissue image, thereby obtaining region segmentation results corresponding to each region segmentation method; the computer device may also use any region segmentation method to perform segmentation processing on the lung region in the three-dimensional tissue image, thereby obtaining a region segmentation result corresponding to the region segmentation method.

[0130] Exemplarily, when a computer device uses one of the region segmentation methods to perform segmentation processing on a three-dimensional tissue image, it can arbitrarily select one of the region segmentation methods as the target region segmentation method; it can also determine the target region segmentation method based on the user's selection; it can also use a preset or default target region segmentation method; or, it can also determine a target region segmentation method that matches the user's habits based on the user's habits; or, it can also determine the target region segmentation method based on the number of times each region segmentation method is used in historical segmentation, etc.; the embodiments of the present application do not make specific limitations on this.

[0131] Step 802: based on the target region segmentation method, a target tissue corresponding to the target region segmentation method is acquired from the three-dimensional tissue image.

[0132] Among them, different regional segmentation methods are based on different lung tissues when performing regional segmentation on the lung area; for example, the lung unit segmentation method can perform lung regional segmentation based on bronchi, and the watershed segmentation method can perform lung regional segmentation based on at least one of bronchi, pulmonary arteries, and pulmonary veins.

[0133] For example, for different regional segmentation methods, tissues corresponding to different regional segmentation methods can be preset to form a preset corresponding relationship. Based on this, after determining the target regional segmentation method, the computer device can determine the target tissue corresponding to the target regional segmentation method based on the preset corresponding relationship; then, the computer device can extract the target tissue from the three-dimensional tissue image to obtain the target tissue corresponding to the target regional segmentation method.

[0134] Step 803 , based on the target tissue, performing regional segmentation processing on the lung region of the subject to be measured in the three-dimensional tissue image to obtain a regional segmentation result of the lung region.

[0135] For example, Fig. 9 As shown, in the case where the target region segmentation method is the lung unit segmentation method, the target tissue may include the bronchus. Then, in the above step 803, based on the target tissue, the lung region of the object to be measured in the three-dimensional tissue image is subjected to region segmentation processing to obtain the region segmentation result of the lung region, which may include:

[0136] Step 901, segmenting the bronchi in the three-dimensional tissue image to obtain a plurality of bronchial segments.

[0137] Step 902 , based on the multiple bronchial segments, perform lung lobe segmentation on the lung region of the subject to be measured in the three-dimensional tissue image to obtain multiple lung lobes corresponding to the lung region.

[0138] Step 903, for each lung lobe, perform lung segmentation on the lung lobe to obtain multiple lung segments corresponding to the lung lobe.

[0139] Step 904, for each lung segment, segment the lung segment into lung subsegments to obtain a plurality of lung subsegments corresponding to the lung segment, and use the plurality of lung subsegments corresponding to each lung segment as the regional segmentation result corresponding to the lung region.

[0140] That is to say, when performing lung unit segmentation of the lung region based on the bronchi, the bronchi can be segmented first by the shape of the bronchi and the oblique fissure features of the lungs (such as oblique fissures), and then the lungs can be segmented into lung units based on the segmented bronchial segments, and the left lung tissue and the right lung tissue can be further subdivided into lobes, where the left lung tissue can be divided into upper lobes and lower lobes, and the right lung tissue can be divided into upper lobes, middle lobes and lower lobes; then, each lobe is further subdivided into lung segments, and the lung segments are further subdivided into lung subsegments, and the subdivision is performed step by step according to the functional hierarchical structure of the lung tissue, so that the user can locate the lung region to which the lung lesion tissue belongs with smaller granularity regional divisions, and perform functional analysis. Figure 10(a) and 10(b) As shown, the display effects after the left lung tissue is segmented into lobes are shown respectively.

[0141] For example, Fig.11 As shown in FIG. 8 , when the target region segmentation method is the watershed segmentation method, the target tissue may include at least one of the bronchus, the pulmonary artery and the pulmonary vein. Then, in the above step 803, based on the target tissue, the lung region of the object to be measured in the three-dimensional tissue image is segmented to obtain the lung region segmentation result, which may include:

[0142] Step 1101 : Based on the target tissue, determine a target flow area corresponding to the target tissue in the lung region of the subject to be tested from the three-dimensional tissue image.

[0143] The target flow area corresponding to the target tissue in the lung region represents the lung tissue region affected by the target tissue.

[0144] Step 1102: segment the target tissue in the three-dimensional tissue image to obtain multiple tissue segments.

[0145] Step 1103, based on each tissue segmentation, the target watershed is segmented to obtain sub-watersheds corresponding to each tissue segmentation, and each sub-watershed is used as a regional segmentation result corresponding to the lung region.

[0146] That is to say, when the lung area is segmented based on at least one of the pulmonary duct tissues among the bronchi, pulmonary arteries and pulmonary veins, the lung tissue area affected by the vascular tissue can be analyzed in combination with the distribution of the vascular tissue. The vascular segmentation can be performed by analyzing the running information of the vascular tissue in the three-dimensional tissue image, and then the lung tissue affected by the segmented vascular segment can be segmented based on the segmented vascular, so that the target flow area corresponding to the pulmonary duct is divided into sub-flow areas corresponding to each pulmonary duct segment, thereby realizing the segmentation of the lung area; so that doctors can dynamically analyze the condition of lung lesions.

[0147] For example, Fig.12 As shown, the above target area segmentation method may also include a free segmentation method; based on this, the above method may also include:

[0148] Step 1201 : in response to a user's marking operation on a three-dimensional tissue image, obtaining a marking anchor point corresponding to the marking operation, and generating at least one segmentation surface according to the marking anchor point.

[0149] Among them, the user's marking operation corresponds to the free segmentation method.

[0150] Step 1202 , based on each segmentation curved surface, performing regional segmentation processing on the lung region of the subject to be tested in the three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung region.

[0151] That is to say, when the lung area is segmented by free segmentation, the user is allowed to draw a three-dimensional segmentation surface by marking anchor points on each layer of the lung scan image, and the left lung tissue and the right lung tissue are segmented by the drawn three-dimensional segmentation surface; in other words, the computer device can allow the user to identify lung features according to the two-dimensional and three-dimensional scan images in a visual way, and automatically generate a segmentation surface by marking anchor points. After the segmentation surface is generated, the user can also be supported to edit the segmentation surface, and after the user adjusts the segmentation area to a suitable position, the unilateral lung tissue is segmented based on the adjusted segmentation area, thereby obtaining the regional segmentation result corresponding to the lung area.

[0152] In this embodiment, the computer device first determines a target area segmentation method, and based on the target area segmentation method, obtains a target tissue corresponding to the target area segmentation method from the three-dimensional tissue image; then, based on the target tissue, performs regional segmentation processing on the lung area of ​​the object to be measured in the three-dimensional tissue image to obtain a regional segmentation result of the lung area; wherein, the target area segmentation method includes at least one of a lung unit segmentation method and a watershed segmentation method; that is, in this embodiment, when the computer device performs segmentation processing on the lung area in the three-dimensional tissue image, it can implement the segmentation of the lung area based on a plurality of different regional segmentation methods, that is, comprehensively considers the segmentation processing of the lung area from multiple dimensions, and can provide users with a variety of lung area segmentation methods to meet the different scenario requirements of users.

[0153] In one embodiment, reference Fig.13As shown, it shows a schematic diagram of the organizational structure of a computer device executing a lung image data processing method. The lung scan image may be a CT scan image, including but not limited to chest and abdomen CT, plain scan CT or enhanced CT; the lung scan image is usually a transverse three-dimensional scan image, based on the transverse three-dimensional scan image, a coronal three-dimensional scan image and a sagittal three-dimensional scan image of the object to be tested can be obtained by tomographic reconstruction, thereby obtaining a multi-planar reconstruction (MPR) image of the object to be tested, and the MPR image can fully display the anatomical structure of the lung scan image.

[0154] When performing multi-tissue segmentation on lung scan images, the internal tissues of the lungs can be segmented, including the left and right lungs, pulmonary arteries and veins, and bronchi. The internal tissues of the lungs can also include lesion tissues, such as lung nodules, lymph nodes, and customized lesion tissues. When performing multi-tissue segmentation, the tissues surrounding the lungs can also be segmented, such as the esophagus, aorta, heart, bones, etc., to display all tissue structures of the chest and abdomen, providing users with a more comprehensive tissue morphology. Of course, when performing multi-tissue segmentation, the tissues used for lung region segmentation can also be extracted in advance, including lung unit segmentation tissues, watershed segmentation tissues, and free segmentation tissues.

[0155] Next, after segmenting multiple tissues in the chest and abdomen and obtaining the tissue segmentation results, the multiple tissues can be managed and analyzed, including but not limited to hierarchical management of each tissue in the multi-tissue mask image, attribute management of each tissue in the multi-tissue grid image, etc. The attribute information can be customized and adjusted by the user, including but not limited to color attributes, transparency attributes, and visibility attributes, etc. For each tissue, users are also supported to perform adaptive adjustments, such as contour editing, VOI drawing, multi-point connection, tissue generation and other tissue adjustment operations.

[0156] The adjusted multi-tissue mask image and multi-tissue grid image can be displayed separately, and the MPR image of the object to be tested can also be displayed. Next, for the multi-tissue grid image, three different regional segmentation methods can be used to perform regional segmentation processing on the lung area, so as to accurately locate the lesion tissue inside the lung area, thereby providing users with a more comprehensive lesion analysis.

[0157] refer to Fig.14As shown, it shows a schematic diagram of the structure of lung tissue segmentation. First, a multi-tissue segmentation model can be obtained based on deep learning model training, and the initial deep learning network is trained using multiple sample lung scan images and calibration data corresponding to each sample lung scan image to obtain a multi-tissue segmentation model; wherein the calibration data corresponding to the sample lung scan image can be a gold standard multi-tissue mask image corresponding to the sample lung scan image, and the calibration data can be obtained by manual calibration by the user, or by other calibration methods.

[0158] Next, for the chest and abdomen CT scan data of the subject to be tested, that is, the lung scan image, a pre-trained multi-tissue segmentation model is used to perform multi-tissue segmentation processing to obtain a multi-tissue mask image; including but not limited to segmentation processing of the whole lung, blood vessels, chest tissue, and lung lesions. Next, for the segmented whole lung tissue, that is, including the left lung tissue and the right lung tissue, the above three regional segmentation methods can be used to perform regional segmentation processing on the whole lung tissue respectively to obtain different regional segmentation results of the whole lung tissue; the regional segmentation results can be used to perform regional positioning of the lesion tissue inside the lung, so as to obtain the lesion area corresponding to the lesion tissue.

[0159] The above-mentioned lung image data processing method comprehensively manages and analyzes various chest and abdominal tissues from abstract data structures such as original lung scan images (Image data), multi-tissue mask images (mask data), multi-tissue mesh images (mesh data), assembly relationships of multi-tissue mesh models (mesh assembly relationships), and tissue attributes of multi-tissue mesh models. It is not limited to the direct segmentation of a single lung tissue, and can systematically analyze lung lesions in combination with other tissues in the chest cavity to achieve accurate regional positioning of lung lesions; in addition, in order to solve the problem of a large number of tissues and tissue overlap, an efficient tissue management method is provided through mask data compression and mask hierarchical management, which is convenient for the management and display of multiple tissues; furthermore, by providing three different regional segmentation methods such as lung unit segmentation, free segmentation, and watershed segmentation to assist in the analysis of lung lesions, a multi-dimensional lung analysis method is provided.

[0160] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0161] Based on the same inventive concept, the embodiment of the present application also provides a lung image data processing device for implementing the lung image data processing method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations of one or more lung image data processing device embodiments provided below can refer to the limitations of the lung image data processing method above, and will not be repeated here.

[0162] In one embodiment, Fig.15 As shown, a lung image data processing device is provided, comprising: an acquisition module 1501, a tissue segmentation module 1502, a three-dimensional reconstruction module 1503, a region segmentation module 1504 and a positioning module 1505, wherein:

[0163] The acquisition module 1501 is used to acquire a lung scan image of the subject to be tested.

[0164] The tissue segmentation module 1502 is used to perform tissue segmentation processing on the lung scan image and obtain a tissue segmentation result corresponding to the lung scan image.

[0165] The three-dimensional reconstruction module 1503 is used to perform three-dimensional reconstruction on the tissue segmentation result to obtain a three-dimensional tissue image corresponding to the lung scan image.

[0166] The region segmentation module 1504 is used to perform region segmentation processing on the lung region of the object to be tested according to the three-dimensional tissue image to obtain a region segmentation result corresponding to the lung region.

[0167] The positioning module 1505 is used to locate the lesion region in the lung region based on the region segmentation result.

[0168] In one embodiment, the tissue segmentation result includes a multi-tissue mask image; the tissue segmentation module 1502 includes:

[0169] The tissue segmentation submodule is used to input the lung scan image into a preset tissue segmentation model for tissue segmentation processing to obtain tissue label information of each pixel in the lung scan image; the preset tissue segmentation model includes models for segmenting multiple tissues;

[0170] The generation submodule is used to generate a multi-tissue mask image corresponding to the lung scan image according to the tissue label information of each pixel point in the lung scan image.

[0171] In one embodiment, if a pixel corresponds to multiple tissue label information, a submodule is generated, including:

[0172] A label level determination unit is used to determine, for each pixel corresponding to a plurality of tissue label information in the lung scan image, a label level corresponding to each tissue label information of the pixel based on a preset corresponding relationship; the preset corresponding relationship includes a corresponding relationship between different tissue label information and corresponding label levels;

[0173] A target label determination unit, configured to determine target tissue label information of a pixel point from a plurality of tissue label information of the pixel point according to label levels respectively corresponding to the tissue label information of the pixel point;

[0174] The generating unit is used to generate a multi-tissue mask image of the lung scanning image according to the target tissue label information of each pixel point in the lung scanning image.

[0175] In one embodiment, the device further comprises:

[0176] A determination module, for determining pixel information corresponding to each tissue label information from the multi-tissue mask image; the pixel information includes the number of pixels corresponding to the tissue label information and the position information of each pixel belonging to the tissue label information;

[0177] A compression module, used to compress the multi-tissue mask image based on the pixel information corresponding to each tissue label information to generate compressed data;

[0178] The storage module is used to store compressed data; the compressed data is used to generate a multi-tissue mask image after decompression.

[0179] In one embodiment, the region segmentation module 1504 includes:

[0180] A determination submodule is used to determine a target region segmentation method; the target region segmentation method includes at least one of a lung unit segmentation method and a watershed segmentation method;

[0181] An acquisition submodule, used for acquiring a target tissue corresponding to the target region segmentation method from a three-dimensional tissue image based on the target region segmentation method;

[0182] The region segmentation submodule is used to perform region segmentation processing on the lung region of the object to be tested in the three-dimensional tissue image based on the target tissue, and obtain the region segmentation result of the lung region.

[0183] In one embodiment, if the target region segmentation method is a lung unit segmentation method, the target tissue includes a bronchus, and the region segmentation submodule includes:

[0184] A segmentation processing unit, used for segmenting the bronchus in the three-dimensional tissue image to obtain a plurality of bronchial segments;

[0185] A lung lobe segmentation unit, used to perform lung lobe segmentation on the lung region of the object to be tested in the three-dimensional tissue image based on multiple bronchial segments, and obtain multiple lung lobes corresponding to the lung region;

[0186] A lung segmentation unit is used to segment each lung lobe to obtain a plurality of lung segments corresponding to the lung lobe;

[0187] The lung subsegment segmentation unit is used to perform lung subsegment segmentation on each lung segment to obtain multiple lung subsegments corresponding to the lung segment, and use the multiple lung subsegments corresponding to each lung segment as the regional segmentation result corresponding to the lung region.

[0188] In one embodiment, if the target region segmentation method is a watershed segmentation method, the target tissue includes at least one of a bronchus, a pulmonary artery, and a pulmonary vein, and the region segmentation submodule includes:

[0189] A watershed determination unit, configured to determine, based on the target tissue, a target watershed corresponding to the target tissue in the lung region of the subject to be tested from the three-dimensional tissue image;

[0190] A segmentation processing unit, used for segmenting the target tissue in the three-dimensional tissue image to obtain multiple tissue segments;

[0191] The segmentation processing unit is used to perform watershed segmentation processing on the target watershed based on each tissue segment, obtain sub-watersheds corresponding to each tissue segment, and use each sub-watershed as a regional segmentation result corresponding to the lung region.

[0192] In one embodiment, the target area segmentation method further includes a free segmentation method; the device further includes:

[0193] A segmentation surface generation module, used to respond to a user's marking operation on a three-dimensional tissue image, obtain a marking anchor point corresponding to the marking operation, and generate at least one segmentation surface according to the marking anchor point; the user's marking operation corresponds to a free segmentation mode;

[0194] The region segmentation module 1504 is further used to perform region segmentation processing on the lung region of the subject to be tested in the three-dimensional tissue image based on each segmentation curved surface, and obtain a region segmentation result corresponding to the lung region.

[0195] Each module in the above-mentioned lung image data processing device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0196] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig.16 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external computer device in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for processing lung image data is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0197] Those skilled in the art will understand that Fig.16 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0198] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the lung image data processing method in any of the above embodiments are implemented.

[0199] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the lung image data processing method in any of the above embodiments are implemented.

[0200] In one embodiment, a computer program product is provided, including a computer program, which implements the steps of the lung image data processing method in any of the above embodiments when executed by a processor.

[0201] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0202] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0203] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0204] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A lung image data processing method, characterized in that: The method comprises: Acquire a lung scan image of the subject to be tested; Performing tissue segmentation processing on the lung scan image to obtain a tissue segmentation result corresponding to the lung scan image; Performing three-dimensional reconstruction on the tissue segmentation result to obtain a three-dimensional tissue image corresponding to the lung scan image; performing a regional segmentation process on the lung region of the subject to be tested according to the three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung region; The lesion region in the lung region is located based on the region segmentation result.

2. The method according to claim 1, characterized in that The tissue segmentation result includes a multi-tissue mask image; The performing tissue segmentation processing on the lung scan image to obtain a tissue segmentation result corresponding to the lung scan image includes: Inputting the lung scan image into a preset tissue segmentation model for tissue segmentation processing to obtain tissue label information of each pixel in the lung scan image; The preset tissue segmentation model includes models for segmenting multiple tissues; A multi-tissue mask image corresponding to the lung scan image is generated according to the tissue label information of each pixel in the lung scan image.

3. The method according to claim 2, characterized in that If the pixel point corresponds to multiple tissue label information, then generating a multi-tissue mask image of the lung scan image according to the tissue label information of each pixel point in the lung scan image includes: For each pixel point corresponding to a plurality of tissue label information in the lung scan image, based on a preset corresponding relationship, determining a label level corresponding to each of the tissue label information of the pixel point; the preset corresponding relationship includes a corresponding relationship between different tissue label information and corresponding label levels; Determining target tissue label information of the pixel point from a plurality of tissue label information of the pixel point according to label levels respectively corresponding to the tissue label information of the pixel point; A multi-tissue mask image of the lung scan image is generated according to the target tissue label information of each pixel in the lung scan image.

4. The method according to claim 3, characterized in that The method further comprises: For each tissue label information, determining pixel information corresponding to the tissue label information from the multi-tissue mask image; the pixel information includes the number of pixels corresponding to the tissue label information and position information of each pixel belonging to the tissue label information; Based on the pixel information corresponding to each of the tissue label information, compressing the multi-tissue mask image to generate compressed data; The compressed data is stored; the compressed data is used to generate the multi-tissue mask image after decompression.

5. The method according to any one of claims 1 to 4, characterized in that The performing a regional segmentation process on the lung region of the subject to be tested according to the three-dimensional tissue image to obtain a regional segmentation result corresponding to the lung region includes: Determine a target region segmentation method; the target region segmentation method includes at least one of a lung unit segmentation method and a watershed segmentation method; Based on the target region segmentation method, acquiring a target tissue corresponding to the target region segmentation method from the three-dimensional tissue image; Based on the target tissue, a region segmentation process is performed on the lung region of the subject to be measured in the three-dimensional tissue image to obtain a region segmentation result of the lung region.

6. The method according to claim 5, characterized in that If the target region segmentation method is the lung unit segmentation method, the target tissue includes a bronchus, and based on the target tissue, performing region segmentation processing on the lung region of the object to be measured in the three-dimensional tissue image to obtain a region segmentation result of the lung region includes: Segmenting the bronchus in the three-dimensional tissue image to obtain a plurality of bronchus segments; Based on the multiple bronchial segments, performing lung lobe segmentation on the lung region of the subject to be tested in the three-dimensional tissue image to obtain multiple lung lobes corresponding to the lung region; For each of the lung lobes, segmenting the lung lobes to obtain a plurality of lung segments corresponding to the lung lobes; For each of the lung segments, lung subsegment segmentation is performed on the lung segment to obtain multiple lung subsegments corresponding to the lung segment, and the multiple lung subsegments corresponding to each of the lung segments are used as the regional segmentation results corresponding to the lung region.

7. The method according to claim 5, characterized in that If the target region segmentation method is the watershed segmentation method, the target tissue includes at least one of the bronchus, the pulmonary artery and the pulmonary vein, and based on the target tissue, performing regional segmentation processing on the lung region of the object to be measured in the three-dimensional tissue image to obtain the regional segmentation result of the lung region includes: Based on the target tissue, determining a target flow domain corresponding to the target tissue in the lung region of the subject to be tested from the three-dimensional tissue image; Segmenting the target tissue in the three-dimensional tissue image to obtain a plurality of tissue segments; Based on each of the tissue segments, the target watershed is segmented to obtain sub-watersheds corresponding to each of the tissue segments, and each of the sub-watersheds is used as a regional segmentation result corresponding to the lung region.

8. The method according to claim 5, characterized in that The target area segmentation method also includes a free segmentation method; the method also includes: In response to a marking operation of the user on the three-dimensional tissue image, a marking anchor point corresponding to the marking operation is acquired, and at least one segmentation surface is generated according to the marking anchor point; the marking operation of the user corresponds to the free segmentation mode; Based on each of the segmentation curved surfaces, a region segmentation process is performed on the lung region of the subject to be tested in the three-dimensional tissue image to obtain a region segmentation result corresponding to the lung region.

9. A lung image data processing device, characterized in that: The device comprises: An acquisition module, used for acquiring a lung scan image of a subject to be tested; A tissue segmentation module, used to perform tissue segmentation processing on the lung scan image to obtain a tissue segmentation result corresponding to the lung scan image; A three-dimensional reconstruction module, used to perform three-dimensional reconstruction on the tissue segmentation result to obtain a three-dimensional tissue image corresponding to the lung scan image; A region segmentation module, used to perform region segmentation processing on the lung region of the subject to be tested according to the three-dimensional tissue image, and obtain a region segmentation result corresponding to the lung region; A positioning module is used to locate the lesion area in the lung area based on the region segmentation result.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

12. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images