Image correction method and device, computer device and storage medium

Abstract

The application relates to an image correction method and device, computer equipment and a storage medium. The method comprises the following steps: performing image reconstruction processing on projection images collected at different angles to obtain a reconstructed tomographic image; determining an out-of-plane artifact image corresponding to the tomographic image according to each projection image; and performing correction processing on the tomographic image according to the out-of-plane artifact image corresponding to the tomographic image to obtain a corrected tomographic image. The method can improve the image correction efficiency.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to an image correction method, apparatus, computer device, and storage medium. Background Technology

[0002] DBT (Digital Breast Tomosynthesis) can obtain cross-sectional images of the breast, allowing doctors to see lesions that appear blurred due to tissue overlap. DBT can improve the positive detection rate and the ability to differentiate between benign and malignant lesions.

[0003] However, DBT only acquires projected images within a limited scanning angle. When using the acquired projected images for image reconstruction, artifacts are easily introduced into the reconstructed layer images due to data loss.

[0004] In related technologies, high-attenuation tissues that are prone to out-of-plane artifacts are segmented from the projected image using some segmentation techniques. Then, some interpolation methods are used to interpolate this part and perform secondary reconstruction to correct the reconstructed layer image. However, this method not only requires accurate segmentation methods, but also has a complex processing flow, and has the problems of long time consumption and low efficiency. Summary of the Invention

[0005] Therefore, it is necessary to provide an image correction method, apparatus, computer equipment, and storage medium that can reduce time consumption and improve correction efficiency in response to the above-mentioned technical problems.

[0006] An image correction method, the method comprising:

[0007] Image reconstruction processing is performed on the projected images acquired from various angles to obtain the reconstructed tomographic images;

[0008] Based on each of the projected images, determine the out-of-plane artifact image corresponding to the tomographic image;

[0009] Based on the out-of-plane artifact image corresponding to the tomographic image, the tomographic image is corrected to obtain the corrected tomographic image.

[0010] In one embodiment, the image reconstruction processing based on the projected images acquired from various angles to obtain the reconstructed tomographic image includes:

[0011] Perform back-projection processing on any of the projected images to obtain the back-projected image corresponding to the projected image;

[0012] The tomographic image is obtained from the back-projection image corresponding to each of the projected images.

[0013] In one embodiment, the back-projection processing of any of the projected images to obtain the back-projected image corresponding to the projected image includes:

[0014] Perform image processing operations on any of the projected images to obtain the image-processed projected image;

[0015] Perform back-projection processing on the image-processed projected image to obtain the back-projected image corresponding to the image-processed projected image;

[0016] The image processing operation includes at least one of segmentation processing, grayscale transformation processing, and window width and window level processing.

[0017] In one embodiment, determining the out-of-plane artifact image corresponding to the tomographic image based on each of the projected images includes:

[0018] For any of the projected images, determine the out-of-plane artifact image corresponding to the projected image;

[0019] The out-of-plane artifact images corresponding to each of the projected images are superimposed to obtain the out-of-plane artifact image corresponding to the tomographic image.

[0020] In one embodiment, determining the out-of-plane artifact image corresponding to any of the projected images includes:

[0021] For the first projected image, the residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image. The first projected image is any one of the projected images acquired from each angle, and the second projected image is any projected image other than the first projected image among the projected images acquired from each angle.

[0022] Based on the back-projected image corresponding to the residual image and the first projected image, the out-of-plane artifact image corresponding to the first projected image is obtained.

[0023] In one embodiment, determining the residual image corresponding to the first projection image based on the back-projected image corresponding to the first projection image and the back-projected image corresponding to the second projection image includes:

[0024] Determine the mean image of the back-projected image corresponding to the second projected image;

[0025] The residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the mean image.

[0026] In one embodiment, obtaining the out-of-plane artifact image corresponding to the first projected image based on the back-projected image corresponding to the residual and the first projected image includes:

[0027] The residual image is corrected to obtain the corrected weight image of the first projected image;

[0028] Based on the back-projection image corresponding to the first projection image and the corrected weight image, the out-of-plane artifact image corresponding to the first projection image is obtained.

[0029] An image correction device, the device comprising:

[0030] The reconstruction module is used to perform image reconstruction processing based on the projected images acquired from various angles to obtain reconstructed tomographic images.

[0031] The determining module is used to determine the out-of-plane artifact image corresponding to the tomographic image based on each of the projected images;

[0032] The correction module is used to correct the fault image based on the out-of-plane artifact image corresponding to the fault image, so as to obtain the corrected fault image.

[0033] In one embodiment, the reconstruction module is further configured to:

[0034] Perform back-projection processing on any of the projected images to obtain the back-projected image corresponding to the projected image;

[0035] The tomographic image is obtained from the back-projection image corresponding to each of the projected images.

[0036] In one embodiment, the reconstruction module is further configured to:

[0037] Perform image processing operations on any of the projected images to obtain the image-processed projected image;

[0038] Perform back-projection processing on the image-processed projected image to obtain the back-projected image corresponding to the image-processed projected image;

[0039] The image processing operation includes at least one of segmentation processing, grayscale transformation processing, and window width and window level processing.

[0040] In one embodiment, the determining module is further configured to:

[0041] For any of the projected images, determine the out-of-plane artifact image corresponding to the projected image;

[0042] The out-of-plane artifact images corresponding to each of the projected images are superimposed to obtain the out-of-plane artifact image corresponding to the tomographic image.

[0043] In one embodiment, the determining module is further configured to:

[0044] For the first projected image, the residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image. The first projected image is any one of the projected images acquired from each angle, and the second projected image is any projected image other than the first projected image among the projected images acquired from each angle.

[0045] Based on the back-projected image corresponding to the residual image and the first projected image, the out-of-plane artifact image corresponding to the first projected image is obtained.

[0046] In one embodiment, the determining module is further configured to:

[0047] Determine the mean image of the back-projected image corresponding to the second projected image;

[0048] The residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the mean image.

[0049] In one embodiment, the determining module is further configured to:

[0050] The residual image is corrected to obtain the corrected weight image of the first projected image;

[0051] Based on the back-projection image corresponding to the first projection image and the corrected weight image, the out-of-plane artifact image corresponding to the first projection image is obtained.

[0052] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps:

[0053] Image reconstruction processing is performed on the projected images acquired from various angles to obtain the reconstructed tomographic images;

[0054] Based on each of the projected images, determine the out-of-plane artifact image corresponding to the tomographic image;

[0055] Based on the out-of-plane artifact image corresponding to the tomographic image, the tomographic image is corrected to obtain the corrected tomographic image.

[0056] A computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0057] Image reconstruction processing is performed on the projected images acquired from various angles to obtain the reconstructed tomographic images;

[0058] Based on each of the projected images, determine the out-of-plane artifact image corresponding to the tomographic image;

[0059] Based on the out-of-plane artifact image corresponding to the tomographic image, the tomographic image is corrected to obtain the corrected tomographic image.

[0060] The aforementioned image correction method, apparatus, computer equipment, and storage medium can perform image reconstruction processing based on projection images acquired from various angles to obtain reconstructed tomographic images. After obtaining these reconstructed tomographic images, the out-of-plane artifact images corresponding to the tomographic images are determined based on each projection image. Furthermore, the tomographic images are corrected based on these out-of-plane artifact images to obtain the corrected tomographic images. The image correction method, apparatus, computer equipment, and storage medium provided in this disclosure can correct tomographic images using out-of-plane artifact images. Correction of the tomographic image can be achieved through a single reconstruction process, thus simplifying the image correction process and improving its efficiency. Moreover, the image correction method provided in this disclosure can directly correct tomographic images using out-of-plane artifact images, thereby improving image correction accuracy. Attached Figure Description

[0061] Figure 1 This is a flowchart illustrating an image correction method in one embodiment;

[0062] Figure 2 This is a flowchart illustrating step 102 in one embodiment;

[0063] Figure 3 This is a flowchart illustrating step 202 in one embodiment;

[0064] Figure 4 This is a flowchart illustrating step 104 in one embodiment;

[0065] Figure 5 This is a flowchart illustrating step 402 in one embodiment;

[0066] Figure 6 This is an example schematic diagram of an image correction method in one embodiment;

[0067] Figures 7a-7b This is an example schematic diagram of an image correction method in one embodiment;

[0068] Figure 8 This is a flowchart illustrating step 502 in one embodiment;

[0069] Figure 9 This is a flowchart illustrating step 504 in one embodiment;

[0070] Figure 10 This is a flowchart illustrating an image correction method in one embodiment;

[0071] Figure 11 This is a flowchart illustrating an image synthesis method in one embodiment;

[0072] Figure 12 This is a flowchart illustrating step 1102 in one embodiment;

[0073] Figure 13 This is a flowchart illustrating step 1104 in one embodiment;

[0074] Figure 14 This is a flowchart illustrating step 1106 in one embodiment;

[0075] Figure 15 This is an example schematic diagram of a custom interface in one embodiment;

[0076] Figure 16 This is a flowchart illustrating an image synthesis method in one embodiment;

[0077] Figure 17 This is an example schematic diagram of an image synthesis method in one embodiment;

[0078] Figures 18a-18b This is an example schematic diagram of an image synthesis method in one embodiment;

[0079] Figure 19 This is a structural block diagram of an image correction device in one embodiment;

[0080] Figure 20 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0081] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0082] In one embodiment, such as Figure 1 As shown, an image correction method is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and further to a system including both a terminal and a server, and implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0083] Step 102: Perform image reconstruction processing on the projected images acquired from various angles to obtain the reconstructed tomographic image.

[0084] For example, the projected images of the target tissue acquired by the DBT device from various angles can be reconstructed to obtain the corresponding tomographic images. The target tissue can be any human tissue structure, such as the heart, liver, or breast. This embodiment does not specifically limit the target tissue.

[0085] For example, the projected images acquired from each angle can be filtered separately, and the filtered projected images can be back-projected. The resulting back-projected images can then be superimposed to obtain the reconstructed tomographic image.

[0086] Step 104: Determine the out-of-plane artifact image corresponding to the tomographic image based on each of the projected images.

[0087] For example, the out-of-plane artifact image corresponding to the tomographic image can be determined by using the various projection images used to reconstruct the tomographic image. For instance, the out-of-plane artifact image corresponding to each projection image can be determined, and then the out-of-plane artifact image corresponding to the tomographic image can be determined based on the out-of-plane artifact images corresponding to each projection image. Alternatively, the out-of-plane artifact image corresponding to the tomographic image can also be determined by using a pre-trained neural network for generating out-of-plane artifact images. For instance, each projection image and / or tomographic image can be input into a pre-trained neural network, and the output of the neural network will include the out-of-plane artifact image corresponding to the tomographic image. The training process of the neural network is not described in detail in the embodiments of this disclosure. Any neural network training method that can train the neural network for generating out-of-plane artifact images is acceptable.

[0088] Step 106: Based on the out-of-plane artifact image corresponding to the tomographic image, perform correction processing on the tomographic image to obtain the corrected tomographic image.

[0089] For example, after obtaining the out-of-plane artifact image corresponding to the tomographic image, the tomographic image can be corrected based on the out-of-plane artifact image to remove the artifact and obtain the corrected tomographic image. For instance, the tomographic image can be subtracted from the out-of-plane artifact image, and the result is the corrected tomographic image. Alternatively, the out-of-plane artifact image can be further corrected (e.g., by determining a corresponding correction coefficient based on prior knowledge of the artifacts and correcting the out-of-plane artifact image based on this coefficient). The tomographic image can then be subtracted from the corrected out-of-plane artifact image, and the result is the corrected tomographic image.

[0090] The aforementioned image correction method performs image reconstruction processing on projection images acquired from various angles to obtain reconstructed tomographic images. It then determines the out-of-plane artifact images corresponding to the tomographic images based on each projection image. The tomographic images are then corrected using these out-of-plane artifact images to obtain the corrected tomographic image. The image correction method provided in this disclosure can correct tomographic images using out-of-plane artifact images, achieving correction in a single reconstruction process. This simplifies the image correction process and improves efficiency. Furthermore, the method can directly correct tomographic images using out-of-plane artifact images, thus improving image correction accuracy.

[0091] In one embodiment, such as Figure 2 As shown, step 102 may include:

[0092] Step 202: Perform back-projection processing on any of the projected images to obtain the back-projected image corresponding to the projected image;

[0093] Step 204: Obtain the tomographic image based on the back-projection image corresponding to each of the projected images.

[0094] For example, for any projected image, backprojection processing can be performed to obtain the corresponding backprojection result. For instance, based on the parameters of the image acquisition device used to acquire the projected image (taking an X-ray tube as an example, the parameters of the X-ray tube may include the tube's rotation radius, the distance from the tube to the detector, etc.), the reconstructed geometry, and the acquisition angle of the projected image, backprojection processing can be performed to obtain the corresponding backprojected image. This process can be repeated to obtain the backprojected images corresponding to each projected image. After obtaining the backprojected images corresponding to each projected image, multiple backprojected images can be superimposed or weighted to obtain the corresponding tomographic image.

[0095] In this embodiment, image reconstruction processing can be performed on the back-projected images obtained by back-projecting the projected images acquired from multiple angles to obtain the corresponding tomographic images. Then, the out-of-plane artifact images corresponding to the tomographic images can be obtained based on the multiple projected images. Subsequently, the tomographic images can be corrected based on the out-of-plane artifact images of the tomographic images, thereby improving the accuracy of image correction.

[0096] In one embodiment, refer to Figure 3 Step 202 above may include:

[0097] Step 302: Perform image processing operation on any of the projected images to obtain the image-processed projected image;

[0098] Step 304: Perform back-projection processing on the image-processed projected image to obtain the back-projected image corresponding to the image-processed projected image;

[0099] The image processing operation includes at least one of segmentation processing, grayscale transformation processing, and window width and window level processing.

[0100] For example, before performing backprojection processing on the projected image, image processing operations can be performed on the projected image. These image processing operations can be operations used to improve the clarity of the projected image. For example, the image processing operations can include, but are not limited to, at least one of the following processing operations: segmentation processing, grayscale transformation processing, window width and window level processing.

[0101] Taking image processing operations including segmentation, grayscale transformation, and window width / level processing as examples, the embodiments of this disclosure will be described. Segmentation can identify target tissue regions and non-target tissue regions in a projected image, and extract the target tissue region from the projected image to obtain an image of the target tissue region. Grayscale transformation can map and transform all pixel values ​​in the target tissue region image according to a preset transformation relationship to obtain a grayscale-transformed image of the target tissue region, thereby improving the contrast and clarity of the target tissue region image. Window width / level processing can determine the display area image in the grayscale-transformed target tissue region image based on information such as window width and window position.

[0102] After obtaining the image-processed projected image, the image-processed projected image (such as the display area image in the example above) can be back-projected to obtain the back-projected image corresponding to the image-processed projected image. The specific back-projection processing procedure can be referred to the foregoing embodiments, and will not be repeated here.

[0103] In this embodiment, before back-projection processing, image processing operations can be performed on the projected image, and then back-projection processing can be performed on the image-processed projected image to obtain the corresponding back-projected image. Since the image-processed projected image is clearer, a back-projected image with higher accuracy can be obtained, thereby improving the accuracy of the synthesized tomographic image.

[0104] In one embodiment, refer to Figure 4 Step 104 above may include:

[0105] Step 402: For any of the projected images, determine the out-of-plane artifact image corresponding to the projected image;

[0106] Step 404: Overlay the out-of-plane artifact images corresponding to each of the projected images to obtain the out-of-plane artifact image corresponding to the tomographic image.

[0107] For example, for any projected image, the out-of-plane artifact image corresponding to that projected image can be obtained. For instance, the out-of-plane artifact image of a projected image can be determined by the difference between its back-projected image and the back-projected images of other projected images; alternatively, a pre-trained neural network can be used to obtain the out-of-plane artifact image of the projected image. For example, the projected image and / or its corresponding back-projected image can be used as input data for image processing to obtain the out-of-plane artifact image corresponding to the projected image. This process can be repeated for other projected images to obtain the out-of-plane artifact images corresponding to each projected image.

[0108] After obtaining the out-of-plane artifact images corresponding to each projection image, the out-of-plane artifact images corresponding to each projection image can be superimposed or weighted and summed to obtain the out-of-plane artifact image corresponding to the tomographic image.

[0109] In this embodiment of the present disclosure, by determining the out-of-plane artifact images corresponding to each projection image used to reconstruct the tomographic image, and obtaining the out-of-plane artifact image corresponding to the tomographic image based on the out-of-plane artifact images corresponding to each projection image, the tomographic image can be directly corrected using the out-of-plane artifact image corresponding to the tomographic image without relying on other detection algorithms. This not only improves the correction accuracy of the tomographic image but also enhances the applicability of the tomographic image correction method.

[0110] In one embodiment, refer to Figure 5 Step 402 above may include:

[0111] Step 502: For the first projected image, determine the residual image corresponding to the first projected image based on the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image, wherein the first projected image is any one of the projected images acquired from each angle, and the second projected image is the projected image other than the first projected image among the projected images acquired from each angle.

[0112] Step 504: Obtain the out-of-plane artifact image corresponding to the first projection image based on the back-projection image corresponding to the residual image and the first projection image.

[0113] For example, any one of the projected images acquired from various angles can be used as the first projected image, and all projected images other than the first projected image can be used as the second projected image. For the first projected image, the residual image between the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image can be determined.

[0114] For example, a residual image can be determined between the back-projected image corresponding to the first projected image (hereinafter referred to as the first back-projected image) and the back-projected image corresponding to at least one second projected image (hereinafter referred to as the second back-projected image), and then the residual image corresponding to the first projected image can be determined based on the residual image between the first back-projected image and at least one second back-projected image.

[0115] For example, the residual images between the first backprojected image and each of the second backprojected images can be determined separately, and the mean image of each residual image can be used as the residual image corresponding to the first projected image. Alternatively, the second backprojected images can be divided into multiple image groups. For any image group, the residual images between the first backprojected image and multiple second backprojected images in that image group can be determined, and these residual images can be used as the residual image of the first projected image relative to that image group. After obtaining the residual images of the first projected image relative to each image group, the mean image of the residual images of the first projected image relative to each image group can be determined as the residual image corresponding to the first projected image.

[0116] After obtaining the residual image corresponding to the first projected image, the out-of-plane artifact image corresponding to the first projected image can be obtained based on the residual image and the first back-projected image. This process can be repeated for other projected images to obtain the out-of-plane artifact images corresponding to each projected image.

[0117] For example, after image reconstruction of the projected images acquired at angles 1, 2, and 3, a tomographic image can be obtained. (Refer to...) Figure 6 As shown, the projection at angle 3 passes through calcification 602, and during back projection, it will leave a clear calcification image on the non-focusing calcification layer 601. This part of the image corresponds to a larger grayscale value (e.g., Figure 6 (The area marked with bold black lines) The projections at angles 1 and 2 did not penetrate the calcification. After superimposing the backprojection results, the grayscale values ​​are relatively small (no obvious calcification image). This means the residual between the backprojected image of angle 3 and the backprojected images of angles 1 and 2 is large. Therefore, the out-of-plane artifacts in the backprojected image corresponding to angle 3 can be extracted. By subtracting the out-of-plane artifact image corresponding to angle 3 from the tomographic image, the corrected tomographic image can be obtained. The tomographic image before correction can be referenced. Figure 7a As shown (before correction) Figure 7aThe image includes obvious out-of-plane artifacts (701), and the corrected tomographic image can be referenced. Figure 7b As shown.

[0118] In this embodiment of the present disclosure, the out-of-plane artifact image corresponding to the first projection image can be obtained by the residual image between the back-projection image corresponding to the first projection image and the back-projection image corresponding to the second projection image. That is, in this embodiment of the present disclosure, the out-of-plane artifact image of the projection image can be obtained by relying only on the attenuation of the material relative to the surrounding tissue without relying on other detection algorithms or statistical information. Then, the tomographic image can be directly corrected based on the out-of-plane artifact image of the projection image, which can improve the applicability of the tomographic image correction method.

[0119] In one embodiment, refer to Figure 8 Step 502 above may include:

[0120] Step 802: Determine the mean image of the back-projected image corresponding to the second projected image;

[0121] Step 804: Determine the residual image corresponding to the first projection image based on the backprojected image corresponding to the first projection image and the mean image.

[0122] For example, the backprojected images corresponding to the second projected image can be superimposed, and the resulting image can be divided by the number of second projected images to obtain the mean image of the backprojected images corresponding to the second projected image. Alternatively, the tomographic image can be subtracted from the backprojected image corresponding to the first projected image, and the difference can be divided by the number of second projected images to obtain the mean image of the backprojected images corresponding to the second projected image.

[0123] After obtaining the mean image of the backprojected image corresponding to the second projected image, the difference between the backprojected image corresponding to the first projected image and the mean image can be determined, and this difference is determined as the residual image of the first projected image. For example, the process of determining the residual image corresponding to the first projected image can refer to the following formula (I).

[0124]

[0125] Where Δi represents the residual image corresponding to the first projected image (or the i-th projected image), N represents the total number of projected images, and slice Total Used to represent tomographic images, slice i The backprojected image used to represent the first projected image (or the i-th projected image). Where i is a natural number greater than 1.

[0126] The same principle applies to other projected images. The residual images corresponding to each projected image can be obtained, and then the out-of-plane artifact images corresponding to each projected image can be obtained based on the residual images and the back-projected images corresponding to each projected image.

[0127] In the disclosed embodiments, the out-of-plane artifact image corresponding to the first projection image can be obtained by using the residual image between the mean image of the back-projection image corresponding to the second projection image and the back-projection image corresponding to the first projection image. Since a more accurate residual image can be obtained through the mean image, the accuracy of the obtained out-of-plane artifact image can be improved.

[0128] In one embodiment, refer to Figure 9 Step 504 above may include:

[0129] Step 902: Perform correction processing on the residual image to obtain the corrected weight image of the first projected image;

[0130] Step 904: Obtain the out-of-plane artifact image corresponding to the first projection image based on the back-projection image corresponding to the corrected weight image and the first projection image.

[0131] In this embodiment of the disclosure, the residual image can be corrected to obtain a corrected weight image of the first projected image. The data range of each pixel value in the corrected weight image can be 0 to 1. The closer the pixel value is to 1, the greater the possibility that the position corresponds to out-of-plane artifact information. Conversely, the closer the pixel value is to 0, the less likely the position corresponds to tissue result information.

[0132] For example, a correction coefficient can be determined, and the residual image can be corrected according to the correction coefficient to obtain a corrected weighted image of the first projected image. For example, the correction coefficient can be preset according to the required correction accuracy, or the corresponding correction coefficient can be determined according to the residual image. For example, the residual image can be corrected according to empirical values, and after the corrected residual image is mapped according to preset mapping conditions, the correction coefficient is determined according to the mapped residual image. This embodiment of the present disclosure does not specifically limit the method of determining the correction coefficient.

[0133] As another example, the minimum gray value Value in the residual image (labeled Δi in this example) can be determined. min Through the minimum grayscale value Value min Multiplying by the correction parameter yields the first correction value, i.e., First correction value = Correction parameter * Value minAfter obtaining the first correction value, the difference between the residual image Δi and the first correction value can be determined as the first corrected image (marked as Δi′ in this example), that is, Δi′ = Δi - correction parameter * Value. min The correction parameter can be a preset value, which ensures that the gray values ​​of the regions belonging to out-of-plane artifacts in the first corrected image corresponding to the residual image are almost all less than 0.

[0134] By setting the pixel values ​​of regions with grayscale values ​​greater than 0 in the first corrected image to 0, a second corrected image (denoted as Δi” in this example) is obtained. This second corrected image Δi” highlights the regions in the second corrected image where out-of-plane artifacts can be highlighted. By inverting the second corrected image Δi”, a third corrected image (denoted as Δi”) is obtained, i.e., Δi”’ = -Δi”. The maximum grayscale value Value in the third corrected image Δi”’ can then be determined. max By determining the third corrected image Δi”' and the maximum gray value Value max The ratio of ω to ω can be used to obtain the corrected weight image (labeled ω in this example). i ), that is After obtaining the corrected weight image ω of the first projected image i Afterwards, due to the correction of the weight image ω i The closer a pixel value is to 1, the greater the likelihood that the location corresponds to out-of-plane artifact information; the closer a pixel value is to 0, the greater the likelihood that the location corresponds to tissue result information. Therefore, the weighted image ω can be adjusted accordingly. i Slice, the back-projected image corresponding to the first projected image i A multiplication process is performed to extract the out-of-plane artifact information from the backprojection result, resulting in the out-of-plane artifact image corresponding to the first projected image. i .

[0135] The same principle applies to other projected images. We can obtain out-of-plane artifact images corresponding to each projected image. By superimposing these out-of-plane artifact images, we can obtain the out-of-plane artifact image corresponding to the tomographic image. For example, we can sequentially superimpose the out-of-plane artifact images corresponding to each projected image, and then superimpose the out-of-plane artifact image of the current projected image. i Previously, the out-of-plane artifact image obtained by superimposing the out-of-plane artifact images of i-1 projected images was called Artifact'. The out-of-plane artifact image Artifact' is obtained by superimposing the current projected image. i Subsequently, the resulting out-of-plane artifact images Artifact = Artifact' and Artifact i .

[0136] After obtaining the out-of-plane artifact image of the tomographic image, the out-of-plane artifact image can be used to correct the tomographic image, resulting in a corrected tomographic image. For example, the specific correction process can be referred to the following formula (II).

[0137] Slice correct =Slice Total -Artifact Formula (II)

[0138] Among them, Slice Correct Slice is used to characterize the corrected tomographic image. Total Used to characterize the tomographic image before correction. To enable those skilled in the art to better understand the embodiments of this disclosure, the following is combined with... Figure 10 The specific examples shown illustrate embodiments of this disclosure.

[0139] Step 1002: The X-ray tube device (the X-ray source of the DBT device) scans the breast at different angles and acquires N projection images through the X-ray detector.

[0140] Step 1004: Perform image processing operations on any projected image, including segmentation, grayscale transformation, window width and window level processing, to obtain the processed projected image.

[0141] Step 1006: Perform backprojection processing on the N processed projection images to obtain the backprojection images corresponding to the N processed projection images (hereinafter referred to as the backprojection images corresponding to the projection images), and then overlay the backprojection images corresponding to the N projection images to obtain the tomographic image.

[0142] Step 1008: Determine the residual image Δi between the backprojection image corresponding to the i-th projection image and the backprojection images corresponding to the remaining N-1 projection images.

[0143] Step 1010: Determine the correction weight image for the i-th projection image based on Δi, and obtain the out-of-plane artifact image corresponding to the i-th projection image by multiplying the correction weight image with the back-projected image corresponding to the i-th projection image. i .

[0144] Step 1012: Determine the out-of-plane artifact image of the tomographic image and the out-of-plane artifact image corresponding to the i-th projection image. i The sum of is the new out-of-plane artifact image, where when i is 1, the initial out-of-plane artifact image of the tomographic image is 0 (each pixel value is 0).

[0145] Step 1014: For projection images acquired from other directions, repeat steps 1004 to 1006 until all N projection images have been processed.

[0146] Step 1016: Subtract the out-of-plane artifact image from the tomographic image to obtain the corrected tomographic image.

[0147] It should be noted that the above method of sequentially determining the out-of-plane artifact images corresponding to each projection image is only one implementation method in the embodiments of this disclosure. In fact, the out-of-plane artifact images corresponding to each projection image can be processed in parallel, and then the out-of-plane artifact images corresponding to all projection images are summed to obtain the out-of-plane artifact image corresponding to the tomographic image. The embodiments of this disclosure do not specifically limit this.

[0148] In this embodiment of the present disclosure, out-of-plane artifact images in the back-projection images corresponding to each projection image can be extracted using a corrected weighted image, thereby obtaining the out-of-plane artifact image of the tomographic image. This does not require reliance on other detection algorithms or statistical information, and can improve the applicability of the method.

[0149] It should be noted that after removing out-of-plane artifacts from tomographic images, the tomographic images with out-of-plane artifacts removed can be merged to obtain a merged image. For example, this can be achieved through... Figure 11 The image compositing method shown achieves image merging. Figure 11 In the example shown, the image synthesis method may include the following steps:

[0150] Step 1102: Project multiple tomographic images to be synthesized to obtain a first image and a second image corresponding to the multiple tomographic images to be synthesized.

[0151] Step 1104: Filter the first image and the second image respectively to obtain the low-frequency component image and the high-frequency component image corresponding to the multiple tomographic images to be synthesized.

[0152] Step 1106: Based on the low-frequency component image and the high-frequency component image, obtain the composite image corresponding to the multiple tomographic images to be synthesized.

[0153] For example, after setting layer intervals, multiple projected images of target human tissue can be reconstructed into multiple tomographic images through image reconstruction (the specific process of reconstructing tomographic images and removing out-of-plane artifacts is described in the foregoing embodiments, and will not be repeated here). The multiple tomographic images to be synthesized can be consecutive images from a tomographic image set. By projecting these multiple tomographic images, a first image (tissue image) and a second image (lesion image) corresponding to the multiple tomographic images to be synthesized can be obtained.

[0154] The target human tissue may include low-attenuation tissue (e.g., soft tissue) and high-attenuation tissue (e.g., lesion tissue). During the acquisition of projection images of the target human tissue, the signal attenuation in low-attenuation tissue is lower than that in high-attenuation tissue. The first image may include an image corresponding to the low-attenuation tissue within the target human tissue. The pixel value at any coordinate position in the first image is the minimum pixel value at that coordinate position among multiple tomographic images to be synthesized. The second image may include an image corresponding to the high-attenuation tissue within the target human tissue. High-attenuation tissue may include lesion tissue or similar tissues. The pixel value at any coordinate position in the second image is the maximum pixel value at that coordinate position among multiple tomographic images to be synthesized. The second image may also include contour information of the target human tissue.

[0155] This disclosure does not specifically limit the projection processing method for multiple tomographic images to be synthesized. For example, the multiple tomographic images to be synthesized can be projected using the minimum density projection algorithm to obtain a first image corresponding to the multiple tomographic images to be synthesized, and the multiple tomographic images to be synthesized can be projected using the maximum density projection algorithm to obtain a second image corresponding to the multiple tomographic images to be synthesized; or, the multiple tomographic images to be synthesized can be projected using a pre-trained neural network for generating the first image and / or the second image to obtain the first image and / or the second image corresponding to the multiple tomographic images to be synthesized. This disclosure does not specifically limit or elaborate on the training process of the neural network.

[0156] Since the image information corresponding to low-attenuation tissues such as soft tissues changes smoothly in the image, that is, it is represented as low-frequency information in the first image, after obtaining the first image corresponding to multiple tomographic images to be synthesized, the first image can be filtered to filter out the low-frequency information and remove the noise, so as to obtain the low-frequency component image corresponding to multiple tomographic images to be synthesized. This low-frequency component image is the clearer first image after noise reduction.

[0157] Similarly, since the image information corresponding to high-attenuation tissues such as the second tissue changes drastically in the image, that is, it is reflected as high-frequency information in the second image, after obtaining the second image corresponding to multiple tomographic images to be synthesized, the second image can be filtered to filter out the high-frequency information and remove the noise, so as to obtain the high-frequency component image corresponding to multiple tomographic images to be synthesized. This high-frequency component image is the second image obtained after noise reduction.

[0158] For example, after obtaining low-frequency and high-frequency component images corresponding to multiple tomographic images to be synthesized, the obtained low-frequency and high-frequency component images can be fused, for example, by superimposing the obtained low-frequency and high-frequency component images to obtain a composite image corresponding to the multiple tomographic images to be synthesized. The composite image intersects with each tomographic image to be synthesized as a thick slice image.

[0159] In the above image synthesis method, projection processing is performed on multiple tomographic images to be synthesized to obtain a first image and a second image corresponding to the multiple tomographic images to be synthesized. Filtering processing is then performed on the first image and the second image respectively to obtain low-frequency component images and high-frequency component images corresponding to the multiple tomographic images to be synthesized. Based on the low-frequency component images and the high-frequency component images, a synthesized image corresponding to the multiple tomographic images to be synthesized can be obtained. According to the image synthesis method provided in this disclosure, multiple tomographic images to be synthesized are first projected into a first image and a second image. A low-frequency component image is obtained by performing a low-pass filter on the first image, and a high-frequency component image is obtained by performing a high-pass filter on the second image. The synthesized image can then be obtained from the low-frequency component images and the high-frequency component images, reducing the number of filtering steps. Therefore, the image synthesis process can be simplified, time consumption can be reduced, and image synthesis efficiency can be improved.

[0160] In one embodiment, such as Figure 12 As shown, step 1102 above may include:

[0161] Step 1202: Perform minimum density projection processing on the multiple tomographic images to be synthesized to obtain a first image corresponding to the multiple tomographic images to be synthesized; and

[0162] Step 1204: Perform maximum density projection processing on the multiple tomographic images to be synthesized to obtain a second image corresponding to the multiple tomographic images to be synthesized.

[0163] For example, minimum density projection processing can be performed on multiple tomographic images to be synthesized to extract information corresponding to low-attenuation tissues from the multiple tomographic images and synthesize a corresponding first image. For instance, during the minimum density projection processing, the minimum value among the pixel values ​​at the same coordinate position in the multiple tomographic images to be synthesized can be extracted to project the multiple tomographic images into a single corresponding first image.

[0164] Similarly, maximum density projection processing can be performed on multiple tomographic images to be synthesized to extract information corresponding to high-attenuation tissues from the multiple tomographic images and synthesize the corresponding second image. For example, during the maximum density projection processing, the maximum value of the pixel values ​​at the same coordinate position in the multiple tomographic images to be synthesized can be extracted to project the multiple tomographic images into a corresponding second image.

[0165] It should be noted that the specific execution order of steps 1202 and 204 in this embodiment is not limited. In fact, step 1204 can be executed first and then step 1202 can be executed, or steps 1202 and 1204 can be executed simultaneously.

[0166] In this embodiment of the present disclosure, by using minimum density projection processing and maximum density projection processing, a first image corresponding to multiple tomographic images to be synthesized and a second image corresponding to multiple tomographic images to be synthesized can be obtained. Then, by performing filtering processing on the first image and the second image respectively, the low-frequency component image and the high-frequency component image obtained can be used to obtain the corresponding synthesized image. That is, the image synthesis method provided by this embodiment of the present disclosure can reduce the number of filtering times, thereby reducing the image processing time and improving the image synthesis efficiency.

[0167] In one embodiment, such as Figure 13 As shown, step 1104 above may include:

[0168] Step 1302: Perform low-pass filtering on the first image to obtain the low-frequency component image corresponding to the multiple tomographic images to be synthesized; and

[0169] Step 1304: Perform high-pass filtering on the second image to obtain the high-frequency component image corresponding to the multiple tomographic images to be synthesized.

[0170] For example, after obtaining the first image and the second image, a low-pass filter can be used to perform low-pass filtering on the first image to remove noise signals, resulting in multiple low-frequency component images corresponding to the tomographic images to be synthesized. These low-frequency component images constitute a smoother and clearer first image. Similarly, a high-pass filter can be used to perform high-pass filtering on the second image to remove noise signals, resulting in multiple high-frequency component images corresponding to the tomographic images to be synthesized. These high-frequency component images constitute a clearer second image. This disclosure does not specifically limit the use of low-pass and high-pass filters; any low-pass filter (including but not limited to mean filters, Gaussian filters, nonlinear bilateral filters, median filters) and high-pass filter (including but not limited to Canny, Sobel, and other edge filters) is applicable to this disclosure.

[0171] It should be noted that the specific execution order of steps 1302 and 1304 in this embodiment is not limited. In fact, step 1304 can be executed first and then step 1302, or steps 1302 and 1304 can be executed simultaneously.

[0172] In this embodiment of the present disclosure, by performing a low-pass filtering process on the first image to obtain low-frequency component images corresponding to multiple tomographic images to be synthesized, and by performing a high-pass filtering process on the second image to obtain high-frequency component images corresponding to multiple tomographic images to be synthesized, the synthesized image corresponding to multiple tomographic images to be synthesized can be obtained through the low-frequency component images and the high-frequency component images. This reduces the number of filtering processes, thereby reducing image processing time and improving image synthesis efficiency.

[0173] In one embodiment, refer to Figure 14 As shown, step 1106 above may include:

[0174] Step 1402: The low-frequency component image and the high-frequency component image are superimposed to obtain a superimposed image;

[0175] Step 1404: Based on the superimposed image, obtain the composite image corresponding to the multiple tomographic images to be synthesized.

[0176] For example, after obtaining the low-frequency component images and high-frequency component images corresponding to multiple tomographic images to be synthesized, the low-frequency component images and the high-frequency component images can be fused and superimposed to obtain a superimposed image. The superimposed image can then be used to obtain a composite image corresponding to the multiple tomographic images to be synthesized; for example, the obtained superimposed image can be directly used as the composite image corresponding to the multiple tomographic images to be synthesized.

[0177] In this embodiment of the disclosure, by superimposing low-frequency component images and high-frequency component images, a composite image corresponding to multiple tomographic images to be synthesized can be obtained from the superimposed image. Since the low-frequency component images are obtained through a low-frequency filtering process and the high-frequency component images are obtained through a high-frequency filtering process, the number of filtering processes is reduced, which can reduce the image processing time and improve the image synthesis efficiency.

[0178] In one embodiment, step 1404 above may further include:

[0179] The superimposed image and the second image are subjected to weighted summation to obtain the composite image corresponding to the multiple tomographic images to be synthesized.

[0180] For example, after superimposing low-frequency component images with high-frequency component images to obtain the corresponding superimposed image, the superimposed image and the second image can be weighted and summed to obtain a composite image corresponding to multiple tomographic images to be synthesized. The sum of the weight values ​​of the superimposed image and the second image is 1. The larger the weight value of the superimposed image, the more obvious the lesion tissue in the composite image. Conversely, the larger the weight value of the second image, the more obvious the structure of the target human tissue in the composite image.

[0181] For example, in the weighted summation process, the weights corresponding to the superimposed image and the second image can be preset weight values ​​or weight values ​​determined according to factors such as the imaging requirements of the synthesized image. This embodiment does not specifically limit the method of determining the weights.

[0182] In this embodiment of the disclosure, a composite image corresponding to multiple tomographic images to be synthesized can be obtained by weighted summation of the superimposed image and the second image, which can result in a composite image that better meets the user's personalized needs and thus improve the user experience.

[0183] In one embodiment, the above method may further include:

[0184] In response to the weight setting operation, a first weight corresponding to the overlay image and a second weight corresponding to the second image are determined.

[0185] For example, users can customize the first weight for the overlaid image and the second weight for the second image. For instance, if it is necessary to increase the salience of lesions in the synthesized image, a relatively large first weight and a relatively small second weight can be set; or, if it is necessary to improve the integrity of the synthesized image, that is, to display the structural information of the target human tissue, a relatively small first weight and a relatively large second weight can be set.

[0186] For example, refer to Figure 15The diagram illustrates a custom interface for the first weight and the second weight. The custom interface can include a first area and a second area. The first area displays setting boxes for the overlaid image and the first weight, and setting boxes for the second image and the second weight. The second area displays the composite image. Users can set the weight value corresponding to the first weight in the first weight setting box through a first setting operation, which can include entering a weight value in the first weight setting box or selecting a corresponding weight value from a drop-down list. Alternatively, users can set the weight value corresponding to the second weight through a second setting operation, which can also include entering a weight value in the second weight setting box or selecting a corresponding weight value from a drop-down list.

[0187] In this embodiment of the disclosure, after setting the weight value of either the first weight or the second weight, the weight value of the other weight is set to the difference between 1 and the already set weight value (i.e., first weight = 1 - second weight, second weight = 1 - first weight). In other words, the setting of the other weight can be completed according to the already set weight.

[0188] After setting the first weight and the second weight, the superimposed image and the second image can be weighted and summed according to the set first weight and the second weight to obtain the corresponding composite image, which is then displayed in the second area. Users can continue to adjust the first weight and the second weight according to the obtained composite image until a composite image that meets the user's needs is obtained.

[0189] In this embodiment of the disclosure, the first weight corresponding to the superimposed image and the second weight corresponding to the second image can be determined in accordance with the user's setting operation on the weight. This allows the user to customize the weight settings and obtain a composite image that meets the user's needs. This not only enriches the synthesis methods of composite images but also greatly improves the user experience.

[0190] In one embodiment, refer to Figure 16 The above methods may also include:

[0191] Step 1602: Reconstruct the images based on the projection images acquired from multiple angles and the corresponding layer intervals to obtain multiple continuous tomographic images.

[0192] Step 1604: According to the division rules, the plurality of tomographic images are divided into at least one thick slice group, and any thick slice group includes a preset number of consecutive tomographic images.

[0193] Step 1606: Use the tomographic images in the target thick section group as the tomographic images to be synthesized. The target thick section group is any thick section group in the at least one thick section group, and the thick section group is also the tomographic image group to be synthesized.

[0194] For example, the slice interval can be a preset value or a user-defined value. After obtaining projection images acquired from multiple angles, image reconstruction can be performed based on the projection images acquired from multiple angles and the slice interval to obtain multiple consecutive tomographic images. For instance, if the target human tissue is a breast, assuming the breast thickness is 50mm, if a slice interval of 1mm is used to reconstruct the tomographic images, then a 5cm compressed breast will be reconstructed into 50 tomographic images. These 50 tomographic images are consecutive, corresponding to thicknesses of 1mm, 2mm, 3mm, etc.

[0195] A pre-defined partitioning rule can be set to indicate the number of thick slice groups and the preset number of tomographic images in each thick slice group. The number of tomographic images in each thick slice group can be the same or different and / or the tomographic images in each thick slice group can be completely continuous and non-overlapping, or they can have overlapping parts. For example, after the tomographic images are coded and marked according to the slice thickness, thick slice group 1 includes tomographic images 1 to 8, thick slice group 2 includes tomographic images 9 to 20, thick slice group 3 can include tomographic images 18 to 28, and so on.

[0196] After obtaining the various thick slice groups, any thick slice group of the tomographic image to be synthesized can be used as the target thick slice group, and the tomographic image in the target thick slice group is the tomographic image to be synthesized. By performing the aforementioned image synthesis method on the target thick slice group, the corresponding synthesized image can be obtained.

[0197] In this embodiment of the disclosure, the tomographic image can be divided into corresponding thick slice groups according to the division rules. Users can set different division rules according to their needs, which can be applied to different scenarios, improve the applicability of the image synthesis method, and meet the user's customization needs.

[0198] To enable those skilled in the art to better understand the embodiments of this disclosure, the following describes... Figure 17 The examples shown illustrate embodiments of this disclosure.

[0199] Step 1702: Obtain projection images from different angles, and reconstruct the corresponding tomographic images using a reconstruction algorithm based on the given layer intervals.

[0200] Step 1704: Several consecutive fault images are determined from the reconstructed fault images and synthesized into a thick section group, and the fault images in the thick section group are used as the fault images to be synthesized.

[0201] Step 1706: Perform minimum density projection on multiple tomographic images to be synthesized to obtain the first image MinIPSyn corresponding to the multiple tomographic images.

[0202] Step 1708: Perform maximum density projection on multiple tomographic images to be synthesized to obtain a second image MaxIPSyn corresponding to the multiple tomographic images.

[0203] Step 1710: Perform low-pass filtering on the first image MinIPSyn to obtain the low-frequency component image MinIPSynLow corresponding to multiple tomographic images.

[0204] Step 1712: Perform high-pass filtering on the second image MaxIPSyn to obtain the high-frequency component image MaxIPSynHigh corresponding to multiple tomographic images.

[0205] It should be noted that this example does not specifically limit the execution order of the aforementioned steps 1706 and 1708, or steps 1710 and 1712. In fact, steps 1708 can be executed first, followed by steps 1704; steps 1712 can be executed first, followed by steps 1710; or steps 1706 and 1708 can be executed in parallel and synchronously, or steps 1710 and 1712 can be executed in parallel and synchronously.

[0206] Step 1714: Combine the low-frequency component image MinIPSynLow and the high-frequency component image MaxIPSynHigh to obtain the superimposed image img1 = MinIPSynLow + MaxIPSynHigh;

[0207] Step 1716: Combine the superimposed image img1 with the second image MaxIPSyn according to a certain weight to obtain the synthesized image img.

[0208] For example, suppose any tomographic image to be synthesized is as follows: Figure 18a As shown, the synthesized image obtained after synthesis can be referred to Figure 18b As shown, calcifications located in different tomographic images to be synthesized are displayed in a single synthesized image.

[0209] It should be understood that, although Figure 1-18b The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1-18b At least some of the steps in the process may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but may be executed at different times. The execution order of these steps or stages is not necessarily sequential, but may be executed in turn or alternately with other steps or at least some of the steps or stages in other steps.

[0210] In one embodiment, such as Figure 19 As shown, an image correction device is provided, including: a reconstruction module 1901, a determination module 1902, and a correction module 1903, wherein:

[0211] The reconstruction module 1901 is used to perform image reconstruction processing based on the projected images acquired from various angles to obtain the reconstructed tomographic images.

[0212] The determining module 1902 is used to determine the out-of-plane artifact image corresponding to the tomographic image based on each of the projected images;

[0213] The correction module 1903 is used to correct the fault image based on the out-of-plane artifact image corresponding to the fault image to obtain the corrected fault image.

[0214] The aforementioned image correction device can perform image reconstruction processing based on projection images acquired from various angles to obtain reconstructed tomographic images. After obtaining these reconstructed tomographic images, it determines the out-of-plane artifact images corresponding to the tomographic images based on each projection image, and then corrects the tomographic images based on these out-of-plane artifact images to obtain the corrected tomographic images. The image correction device provided in this embodiment can correct tomographic images using out-of-plane artifact images, achieving correction in a single reconstruction process. This simplifies the image correction process and improves its efficiency. Furthermore, the image correction device provided in this embodiment can directly correct tomographic images using out-of-plane artifact images, thus also improving image correction accuracy.

[0215] In one embodiment, the reconstruction module 1901 described above can also be used for:

[0216] Perform back-projection processing on any of the projected images to obtain the back-projected image corresponding to the projected image;

[0217] The tomographic image is obtained from the back-projection image corresponding to each of the projected images.

[0218] In one embodiment, the reconstruction module 1901 described above can also be used for:

[0219] Perform image processing operations on any of the projected images to obtain the image-processed projected image;

[0220] Perform back-projection processing on the image-processed projected image to obtain the back-projected image corresponding to the image-processed projected image;

[0221] The image processing operation includes at least one of segmentation processing, grayscale transformation processing, and window width and window level processing.

[0222] In one embodiment, the determining module 1902 described above can also be used for:

[0223] For any of the projected images, determine the out-of-plane artifact image corresponding to the projected image;

[0224] The out-of-plane artifact images corresponding to each of the projected images are superimposed to obtain the out-of-plane artifact image corresponding to the tomographic image.

[0225] In one embodiment, the determining module 1902 described above can also be used for:

[0226] For the first projected image, the residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image. The first projected image is any one of the projected images acquired from each angle, and the second projected image is any projected image other than the first projected image among the projected images acquired from each angle.

[0227] Based on the back-projected image corresponding to the residual image and the first projected image, the out-of-plane artifact image corresponding to the first projected image is obtained.

[0228] In one embodiment, the determining module 1902 described above can also be used for:

[0229] Determine the mean image of the back-projected image corresponding to the second projected image;

[0230] The residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the mean image.

[0231] In one embodiment, the determining module 1902 described above can also be used for:

[0232] The residual image is corrected to obtain the corrected weight image of the first projected image;

[0233] Based on the back-projection image corresponding to the first projection image and the corrected weight image, the out-of-plane artifact image corresponding to the first projection image is obtained.

[0234] Specific limitations regarding the image correction device can be found in the limitations of the image correction method described above, and will not be repeated here. Each module in the aforementioned image correction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in hardware or independently of the processor in the computer device, or stored in software in the memory of the computer device, so that the processor can call and execute the corresponding operations of each module.

[0235] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 20 As shown, the computer device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. When executed by the processor, the computer program implements an image correction method. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0236] Those skilled in the art will understand that Figure 20 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0237] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0238] Image reconstruction processing is performed on the projected images acquired from various angles to obtain the reconstructed tomographic images;

[0239] Based on each of the projected images, determine the out-of-plane artifact image corresponding to the tomographic image;

[0240] Based on the out-of-plane artifact image corresponding to the tomographic image, the tomographic image is corrected to obtain the corrected tomographic image.

[0241] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0242] Perform backprojection processing on any of the projected images to obtain the backprojected image corresponding to the projected image; obtain the tomographic image based on the backprojected images corresponding to each of the projected images.

[0243] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0244] An image processing operation is performed on any of the projected images to obtain the image-processed projected image; a back-projection operation is performed on the image-processed projected image to obtain the back-projected image corresponding to the image-processed projected image; wherein the image processing operation includes at least one of segmentation processing, grayscale transformation processing, and window width and window level processing.

[0245] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0246] For any of the projected images, determine the out-of-plane artifact image corresponding to the projected image; superimpose the out-of-plane artifact images corresponding to each of the projected images to obtain the out-of-plane artifact image corresponding to the tomographic image.

[0247] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0248] For the first projected image, a residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image. The first projected image is any one of the projected images acquired from each angle, and the second projected image is any projected image other than the first projected image acquired from each angle. Based on the residual image and the back-projected image corresponding to the first projected image, an out-of-plane artifact image corresponding to the first projected image is obtained.

[0249] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0250] Determine the mean image of the back-projected image corresponding to the second projected image; determine the residual image corresponding to the first projected image based on the back-projected image corresponding to the first projected image and the mean image.

[0251] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0252] The residual image is corrected to obtain the corrected weight image of the first projected image; based on the corrected weight image and the back-projected image corresponding to the first projected image, the out-of-plane artifact image corresponding to the first projected image is obtained.

[0253] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0254] Image reconstruction processing is performed on the projected images acquired from various angles to obtain the reconstructed tomographic images;

[0255] Based on each of the projected images, determine the out-of-plane artifact image corresponding to the tomographic image;

[0256] Based on the out-of-plane artifact image corresponding to the tomographic image, the tomographic image is corrected to obtain the corrected tomographic image.

[0257] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0258] Perform backprojection processing on any of the projected images to obtain the backprojected image corresponding to the projected image; obtain the tomographic image based on the backprojected images corresponding to each of the projected images.

[0259] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0260] An image processing operation is performed on any of the projected images to obtain the image-processed projected image; a back-projection operation is performed on the image-processed projected image to obtain the back-projected image corresponding to the image-processed projected image; wherein the image processing operation includes at least one of segmentation processing, grayscale transformation processing, and window width and window level processing.

[0261] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0262] For any of the projected images, determine the out-of-plane artifact image corresponding to the projected image; superimpose the out-of-plane artifact images corresponding to each of the projected images to obtain the out-of-plane artifact image corresponding to the tomographic image.

[0263] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0264] For the first projected image, a residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image. The first projected image is any one of the projected images acquired from each angle, and the second projected image is any projected image other than the first projected image acquired from each angle. Based on the residual image and the back-projected image corresponding to the first projected image, an out-of-plane artifact image corresponding to the first projected image is obtained.

[0265] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0266] Determine the mean image of the back-projected image corresponding to the second projected image; determine the residual image corresponding to the first projected image based on the back-projected image corresponding to the first projected image and the mean image.

[0267] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0268] The residual image is corrected to obtain the corrected weight image of the first projected image; based on the corrected weight image and the back-projected image corresponding to the first projected image, the out-of-plane artifact image corresponding to the first projected image is obtained.

[0269] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this 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, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0270] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0271] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. An image correction method, characterized in that, The method includes: Image reconstruction processing is performed on the projected images acquired from various angles to obtain the reconstructed tomographic images; Determining the out-of-plane artifact image corresponding to the tomographic image based on each of the projected images includes: determining the out-of-plane artifact image corresponding to any one of the projected images; and superimposing the out-of-plane artifact images corresponding to each of the projected images to obtain the out-of-plane artifact image corresponding to the tomographic image. Based on the out-of-plane artifact image corresponding to the fault image, the fault image is corrected to obtain the corrected fault image. Wherein, determining the out-of-plane artifact image corresponding to any of the projected images includes: For the first projected image, the residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image; wherein, the first projected image is any one of the projected images acquired from each angle, and the second projected image is any projected image other than the first projected image among the projected images acquired from each angle. The residual image is corrected to obtain the corrected weight image of the first projected image; Based on the back-projection image corresponding to the first projection image and the corrected weight image, the out-of-plane artifact image corresponding to the first projection image is obtained.

2. The method according to claim 1, characterized in that, The image reconstruction process based on the projected images acquired from various angles to obtain the reconstructed tomographic image includes: Perform back-projection processing on any of the projected images to obtain the back-projected image corresponding to the projected image; The tomographic image is obtained from the back-projection image corresponding to each of the projected images.

3. The method according to claim 2, characterized in that, The step of performing backprojection processing on any of the projected images to obtain the backprojected image corresponding to the projected image includes: Perform image processing operations on any of the projected images to obtain the image-processed projected image; Perform back-projection processing on the image-processed projected image to obtain the back-projected image corresponding to the image-processed projected image; The image processing operation includes at least one of segmentation processing, grayscale transformation processing, and window width and window level processing.

4. The method according to claim 1, characterized in that, The step of determining the residual image corresponding to the first projection image based on the back-projected image corresponding to the first projection image and the back-projected image corresponding to the second projection image includes: Determine the mean image of the back-projected image corresponding to the second projected image; The residual image corresponding to the first projected image is determined based on the back-projected image corresponding to the first projected image and the mean image.

5. An image correction device, characterized in that, The device includes: The reconstruction module is used to perform image reconstruction processing based on the projected images acquired from various angles to obtain reconstructed tomographic images. The determining module is configured to determine the out-of-plane artifact image corresponding to the tomographic image based on each of the projected images, including: determining the out-of-plane artifact image corresponding to any one of the projected images; and performing superposition processing on the out-of-plane artifact images corresponding to each of the projected images to obtain the out-of-plane artifact image corresponding to the tomographic image. The correction module is used to correct the fault image based on the out-of-plane artifact image corresponding to the fault image to obtain the corrected fault image. The determining module is further configured to, for the first projected image, determine a residual image corresponding to the first projected image based on the back-projected image corresponding to the first projected image and the back-projected image corresponding to the second projected image; wherein, the first projected image is any one of the projected images acquired from each angle, and the second projected image is any projected image other than the first projected image among the projected images acquired from each angle; perform correction processing on the residual image to obtain a correction weight image of the first projected image; and obtain an out-of-plane artifact image corresponding to the first projected image based on the correction weight image and the back-projected image corresponding to the first projected image.

6. 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, it implements the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images