Method, processing system and program product for processing cbct projection images

Abstract

The present disclosure provides a CBCT projection image processing method, processing system and program product, the method of the present disclosure first performs local smoothing filtering on the target projection data according to the existing N mask regions respectively, obtains 2N filtered projection data, the modulation component is used for spatially selectively attenuating the rays emitted by the scanning device, the modulation component makes the first projection data dynamically change with the scanning angle, then performs scatter estimation based on the 2N filtered projection data, the first projection data and the second projection data of each scanning angle in multiple scanning angles, obtains the first scatter estimation value of each pixel position under the scanning angle, then performs local smoothing filtering on the first scatter estimation value under the same scanning angle according to the trusted region, obtains the second scatter estimation value of each pixel position under the scanning angle, and finally obtains the projection data after the artifact correction through the difference between the second projection data and the second scatter estimation value.

Description

Technical Field

[0001] This disclosure specifically relates to methods for processing CBCT projection images, systems for processing CBCT projection images, readable storage media, and computer program products. Background Technology

[0002] Because X-rays are scattered when they pass through the object being scanned, the projected images obtained during CBCT (Cone Beam Computed Tomography) scans will contain scattering artifacts, which will also appear in the reconstructed CBCT images. Summary of the Invention

[0003] This disclosure provides a method for processing CBCT projection images, a system for processing CBCT projection images, a readable storage medium, and a computer program product.

[0004] The first aspect of this disclosure proposes a method for processing CBCT projection images, comprising: taking first projection data not containing a subject and second projection data containing a subject as target projection data, and performing local smoothing filtering on the target projection data according to N existing mask regions to obtain 2N filtered projection data, where N>1; the target projection data is obtained by CBCT scanning using a scanning device equipped with a modulation component, wherein the modulation component is used to spatially selectively attenuate the rays emitted by the scanning device, thereby forming N spatial regions corresponding to N types of transmission characteristics in the first projection data, and different mask regions respectively corresponding to the N spatial regions. The modulation component causes the first projection data to change dynamically with the scanning angle. Scattering estimation is performed based on the 2N filtered projection data, the first projection data, and the second projection data for each of the multiple scanning angles to obtain a first scattering estimate for each pixel position at the scanning angle. Local smoothing filtering is applied to the first scattering estimate at the same scanning angle according to a reliable region to obtain a second scattering estimate for each pixel position at the scanning angle, where the reliable region is contained in the union of the N mask regions. Finally, artifact-corrected projection data is obtained through the difference between the second projection data and the second scattering estimate.

[0005] According to some embodiments of this disclosure, the modulation component is cylindrical and surrounds the horizontal side of the detector in the scanning device.

[0006] According to some embodiments of this disclosure, the modulation assembly includes N types of structural components, each of which has different transmission characteristics for rays.

[0007] According to some embodiments of this disclosure, N=2, the N types of structural components include a plurality of first type structural components and a second type structural component, the first type structural component has a smaller transmission characteristic to rays than the second type structural component, and the plurality of first type structural components are distributed in a grid pattern on the second type structural component.

[0008] According to some embodiments of this disclosure, before performing local smoothing filtering on the target projection data, the method further includes: performing CBCT scanning with a scanning device equipped with a modulation component to obtain first projection data when the subject is not included; and performing threshold segmentation on the first projection data to obtain N mask regions.

[0009] According to some embodiments of this disclosure, threshold segmentation is performed on the first projection data to obtain N mask regions, including: smoothing the first projection data to obtain smoothed projection data, and determining the difference between the first projection data and the smoothed projection data; and forming a high-transmission mask region by the pixel position corresponding to the difference that is higher than a first pixel threshold, and forming a low-transmission mask region by the pixel position corresponding to the difference that is lower than a second pixel threshold, to obtain N mask regions at the same scanning angle, where N=2, and the first pixel threshold is greater than the second pixel threshold.

[0010] According to some embodiments of this disclosure, local smoothing filtering is performed on the target projection data according to the existing N mask regions, including: performing local smoothing filtering on the pixel positions in the high-transmission mask regions of the existing N mask regions in the first projection data that does not contain the subject, and interpolating the pixel positions in the first projection data located in the non-high-transmission mask regions using the first projection data located in the high-transmission mask regions to obtain high-transmission flat field data; and performing local smoothing filtering on the pixel positions in the low-transmission mask regions of the existing N mask regions in the first projection data, and interpolating the pixel positions in the first projection data located in the non-low-transmission mask regions using the first projection data located in the low-transmission mask regions to obtain low-transmission flat field data.

[0011] According to some embodiments of this disclosure, local smoothing filtering is performed on the target projection data according to the existing N mask regions, including: performing local smoothing filtering on the pixel positions in the second projection data containing the subject located in the high-transmission mask region of the existing N mask regions, and interpolating the pixel positions in the second projection data located in the non-high-transmission mask region using the second projection data located in the high-transmission mask region to obtain high-transmission projection data; and performing local smoothing filtering on the pixel positions in the second projection data located in the low-transmission mask region of the existing N mask regions, and interpolating the pixel positions in the second projection data located in the non-low-transmission mask region using the second projection data located in the low-transmission mask region to obtain low-transmission projection data.

[0012] According to some embodiments of this disclosure, scattering estimation is performed based on the 2N filtered projection data, the first projection data, and the second projection data for each of the multiple scanning angles to obtain a first scattering estimate value for each pixel position under the scanning angle. This includes: for the same pixel position under each of the multiple scanning angles, determining a first quotient value between the second projection data and the first projection data, and determining a second quotient value between low-transmission flat field data and high-transmission flat field data in the 2N filtered projection data; and determining the first scattering estimate value for each pixel position under the scanning angle based on the first quotient value and the second quotient value.

[0013] According to some embodiments of this disclosure, determining a first scattering estimate of each pixel position at the scanning angle based on the first quotient and the second quotient includes: determining a first product of the first quotient raised to the power of f and low-transmission projection data in the 2N filtered projection data, and determining a second product of the second quotient and high-transmission projection data in the 2N filtered projection data, where f is a specified value; and determining a first scattering estimate of each pixel position at the scanning angle based on the first product and the second product.

[0014] According to some embodiments of this disclosure, determining a first scattering estimate of each pixel position at the scanning angle based on the first product and the second product includes: determining a first difference between the first product and the second product, and determining a second difference between the value 1 and the second quotient; and calculating the quotient of the first difference and the second difference to obtain a first scattering estimate of each pixel position at the scanning angle.

[0015] According to some embodiments of this disclosure, the first scattering estimate of each pixel location in the trusted region is positive and less than the pixel value of the corresponding pixel location in the second projection data.

[0016] According to some embodiments of this disclosure, the gradient value of each pixel position in the trusted region in the high-transmission projection data and low-transmission projection data of the 2N filtered projection data is less than a gradient threshold.

[0017] According to some embodiments of this disclosure, after obtaining artifact-corrected projection data through the difference between the second projection data and the second scattering estimate, the processing method includes: taking the difference between the second projection data and the second scattering estimate as a third difference value; for the third difference value that is less than a difference threshold, using a monotonic function to map the third difference value to a new value greater than zero, so as to smooth the artifact-corrected projection data; wherein the difference threshold value is a positive value.

[0018] A second aspect of this disclosure provides a CBCT projection image processing system, comprising: a scanning device including a radiation source and a detector; a modulation component; a memory storing execution instructions; and a processor that executes the execution instructions stored in the memory, causing the processor to perform the CBCT projection image processing method described in any of the above embodiments.

[0019] A third aspect of this disclosure provides a readable storage medium storing a computer program that, when executed by a processor, is used to implement the CBCT projection image processing method described in any of the above embodiments.

[0020] This disclosure provides a fourth aspect of a computer program product, which includes a computer program that, when executed by a processor, is used to implement the CBCT projection image processing method described in any of the above embodiments. Attached Figure Description

[0021] The accompanying drawings illustrate exemplary embodiments of the present disclosure and, together with the description thereof, serve to explain the principles of the present disclosure. These drawings are included to provide a further understanding of the present disclosure and are incorporated in and constitute a part of this specification.

[0022] Figure 1 A schematic flowchart of a CBCT projection image processing method M100 according to some embodiments of the present disclosure is shown.

[0023] Figure 2 A top view schematic diagram of a scanning device equipped with a modulation component according to some embodiments of the present disclosure is shown.

[0024] Figure 3 A schematic diagram of the modulation component as unfolded according to some embodiments of the present disclosure is shown.

[0025] Figures 4-11A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of the present disclosure is shown.

[0026] Figure 12 This is a schematic block diagram of a CBCT projection image processing apparatus according to one embodiment of the present disclosure.

[0027] Figure 13 This is a schematic block diagram of an electronic device 1000 according to one embodiment of the present disclosure. Detailed Implementation

[0028] The present disclosure will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the disclosure. Furthermore, it should be noted that, for ease of description, only the parts relevant to the present disclosure are shown in the accompanying drawings.

[0029] It should be noted that, where there is no conflict, the embodiments and features described in this disclosure can be combined with each other. The technical solutions of this disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0030] Unless otherwise stated, the exemplary implementations / embodiments shown are to be understood as providing exemplary features of various details that provide ways in which the technical concepts of this disclosure can be implemented in practice. Therefore, unless otherwise stated, the features of various implementations / embodiments may be additionally combined, separated, interchanged and / or rearranged without departing from the technical concepts of this disclosure.

[0031] The terminology used herein is for the purpose of describing particular embodiments and is not restrictive. As used herein, unless the context clearly indicates otherwise, the singular forms “a” and “the” are intended to include the plural forms as well. Furthermore, when the terms “comprising” and / or “including” and variations thereof are used in this specification, it indicates the presence of the stated features, integrals, steps, operations, parts, components, and / or groups thereof, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, parts, components, and / or groups thereof. It should also be noted that, as used herein, the terms “substantially,” “about,” and other similar terms are used as approximate terms rather than as terms of degree, thus explaining the inherent biases in measurements, calculated values, and / or provided values ​​that would be recognized by one of ordinary skill in the art.

[0032] Compared to traditional CT, CBCT offers advantages such as high spatial resolution and low dose. However, during CBCT scanning, X-rays undergo Compton scattering as they pass through the scanned object. Some photons collide with electrons in the tissue inelastically, altering their propagation direction. While the scattered X-rays do not travel along their original straight path, they may still be received by the detector. In other words, the signal received by the detector includes both the main signal and the scattered signal. The scattered signal does not carry accurate spatial location information but increases the detector reading, leading to distortion of the projection data and quality defects in the reconstructed image, such as low contrast, cupping artifacts, streaking artifacts, or overall blurring.

[0033] Therefore, this disclosure proposes a method for processing CBCT projection images.

[0034] First, without a subject, a scanning device equipped with a modulation component is used to obtain flat-field images P0 at multiple scanning angles. Gaussian smoothing filtering is then applied to the flat-field images P0 to obtain multiple baseline images Q0. For each scanning angle, the difference between the flat-field image P0 and the baseline image Q0 is calculated. The high-transmittance mask region U is then determined by the relationship between this difference and a pixel threshold. h and low-transmission mask region U l Then, mask-guided local Gaussian interpolation (local smoothing filtering) is performed on the flat-field image P0 to obtain the high-transmittance flat-field image P. h and low transmission flat field image P l An initial projection image P is obtained by scanning the subject using a scanning device equipped with a modulation component. c Then, the initial projection image P c The high-transmission projection image V is obtained by performing local Gaussian interpolation (local smoothing filtering) guided by the mask region. H and low-transmission projection image V L Then, based on the flat-field image P0 and the high-transmission flat-field image P... h Low transmission flat field image P l High-transmission projection image V H and low-transmission projection image V L Preliminary scattering estimation and beam hardening compensation are performed to obtain a preliminary scattering estimation map G1. Then, a confidence region R is defined to determine the reliability of the scattering estimation. Local Gaussian interpolation (local smoothing filtering) guided by the confidence region R is performed on the preliminary scattering estimation map G1 to obtain the final scattering estimation map G2. From the initial projection image P... c Subtracting the final scattering estimate G2 directly yields the corrected principal ray (primary ray) signal estimate, which in turn produces artifact-corrected projection images at multiple scanning angles. Reconstruction of these artifact-corrected projection images yields the CBCT 3D image.

[0035] Figure 1 A schematic flowchart of a CBCT projection image processing method M100 according to some embodiments of this disclosure is shown. Figure 1 The method shown includes steps S110, S130, S150, and S170. This method can be executed by electronic devices such as computers and servers to remove scattering artifacts and avoid introducing other artifacts.

[0036] S110, the first projection data that does not contain the subject and the second projection data that contains the subject are used as target projection data. Local smoothing filtering is then applied to the target projection data according to the existing N mask regions to obtain 2N filtered projection data. The target projection data is obtained by CBCT scanning using a scanning device equipped with a modulation component.

[0037] The first projection data is empty-field projection data acquired by a CBCT scanner without a subject (such as the human torso or head) placed on it. The image formed by the first projection data is a flat-field image P0. The second projection data is projection data acquired by a CBCT scanner with a subject placed on it. The image formed by the second projection data is the initial projection image P0. c Flat field image P0 and initial projection image P c The images are the same size.

[0038] The target projection data is obtained through CBCT scanning using a scanning device equipped with a modulation component. In other words, the scanning device is equipped with a modulation component during both the acquisition of the first and second projection data. The modulation component is a solid structure installed between the radiation source and the detector, and the path of the radiation emitted from the source passes through the modulation component.

[0039] The modulation component is used to spatially selectively attenuate the rays emitted by the scanning device, thereby forming N spatial regions corresponding to N transmission characteristics in the first projection data. The degree of ray attenuation is related to the shape, structure, material, and other properties of the modulation component. If N>1, it indicates that the spatial structure of the modulation component is non-uniform, and therefore the attenuation of multiple rays passing through it is non-uniform; for example, the attenuation of some rays in space is greater than the attenuation of others.

[0040] Transmission characteristics refer to the degree or ability of an object to allow rays to pass through it. Transmission characteristics can be represented by transmittance. Higher transmittance indicates a lower degree of attenuation of the rays. N transmission characteristics can correspond to N different materials in the modulation assembly, utilizing the properties of different materials to achieve different levels of attenuation. N spatial regions are equivalent to N pixel regions. Without a subject being placed, different rays pass through different material portions of the modulation assembly and reach the detector, forming the first projection data.

[0041] Understandably, the different materials in the modulation assembly need to be configured to have a significant difference in spatial selective attenuation effect on X-rays, so that there are obvious differences in the transmission characteristics between different structures, thereby allowing N spatial regions to be distinguished and forming a spatially encoded mask (spatial modulation pattern).

[0042] It should be noted that when a subject is present, due to the presence of the modulation assembly, the X-rays pass through both the modulation assembly and the subject before reaching the detector. Therefore, different X-rays are attenuated to varying degrees by the modulation assembly, which is reflected in the second projection data. The second projection data is essentially a superposition of the subject's projection data and the first projection data. Thus, the relative differences caused by the modulation assembly exist simultaneously in both the first and second projection data. The first projection data reflects the spatial structure of the modulation assembly and the original intensity distribution of the X-ray source. The second projection data also includes the main X-ray signal and the scattered signal.

[0043] Different mask regions correspond to pixel positions in N spatial regions. Each mask region corresponds to a transmission characteristic in the modulation component. The set of pixels formed when a ray passes through a structure with that transmission characteristic in the modulation component and reaches the detector area is the mask region. The mask region can be represented by a mask image. The mask image can be a binary image and its size is the same as the flat-field image P0 and the initial projection image P. c The same applies. For example, the area formed by the pixel values ​​of 1 in the mask image is the mask region, and the values ​​of all other pixel positions in the image can be 0.

[0044] Local smoothing filtering is a mask-guided local Gaussian filtering method. It uses the masked region as a known, marked region, and applies this known region to the flat image P0 and the initial projected image P1. c Smoothing is performed. Specifically, local smoothing filtering is applied to the flat-field image P0 to obtain N filtered flat-field images P (filtered projection data). The initial projection image P... cLocal smoothing filtering also yields N filtered projection images V (filtered projection data), thus providing 2N filtered projection data for each scanning angle. Assuming a CBCT scan uses 720 scanning angles, step S110 yields 720 sets of filtered projection data, each set containing N filtered flat-field images P and N filtered projection images V.

[0045] The modulation component causes the first projection data to change dynamically with the scanning angle. As the scanning device (source and detector) rotates around the rotation center during CBCT scanning, the position or orientation of the modulation component can be adjusted once for each scan. This allows the same ray to pass through different positions of the modulation component at different scanning angles, resulting in different degrees of attenuation of the ray at different scanning angles. Consequently, the flat-field image P0 formed by each ray in the detector detection area is dynamically changed with the scanning angle.

[0046] For example, the flat-field images P0 obtained at each scanning angle can be different, i.e., the number of flat-field images P0 is equal to the number of scanning angles. Alternatively, the flat-field images P0 obtained at some scanning angles can be the same. For example, after a specified number of scans, the position and orientation of the modulation component relative to the source or detector will be the same as the position and orientation before the specified number of scans, so that the flat-field images at specified scan intervals are the same.

[0047] S130, based on 2N filtered projection data, first projection data and second projection data of each scanning angle in multiple scanning angles, scattering estimation is performed to obtain the first scattering estimate value of each pixel position under the scanning angle.

[0048] Scattering estimation is used to predict and quantify the scattered radiation generated by the scanned object (subject) during imaging, thereby separating the scattered signal from the mixed signal (main ray + scattered signal). This is achieved using the filtered pre-scan flat-field image P0 and the initial projection image P at each scanning angle. c The filtered flat-field image P and the filtered projected image V are used to estimate the scattering value of each pixel position at the scanning angle, resulting in a preliminary scattering estimation image G1 that can represent the intensity distribution of the scattered signal.

[0049] S150: Local smoothing filtering is applied to the first scattering estimate at the same scanning angle according to the reliable region to obtain the second scattering estimate at each pixel position at the scanning angle. The reliable region is contained in the union of N mask regions.

[0050] The initial scattering estimation image G1 may contain errors. For example, when there are areas with strong edges, high-contrast structures, or rich details (such as the boundary between bone and soft tissue) inside the subject, the image gradient may be large, causing errors in the filtered projected image V. Therefore, a mask-guided local Gaussian filter can be performed again, this time using a reliable region R as the mask. Using the reliable region R as a labeled known region, the initial scattering estimation image G1 is smoothed according to the known region to obtain the final scattering signal distribution, which is the final scattering estimation image G2.

[0051] A reliable region R is a pixel region that satisfies specified conditions, within which scattering estimates are considered reliable estimates. These specified conditions include a positional constraint, meaning the reliable region R is a proper subset of the union of N mask regions.

[0052] In the final scattering estimation map G2, the value of a pixel in the confidence region is equal to the value at the corresponding position in the initial scattering estimation map G1, and the value of a pixel in the non-confidence region can be filled by interpolation, for example, by calculating from a neighboring confidence region.

[0053] S170, the artifact-corrected projection data is obtained by using the difference between the second projection data and the second scattering estimate.

[0054] From the initial projection image P c Subtracting the final scattering estimate G2 directly yields the artifact-corrected principal ray signal estimate (primary ray signal estimate), which is the final projected image after correcting for scattering artifacts. This final projected image can be used for CT reconstruction. For example, after obtaining the final projected image for each scanning angle, CT reconstruction is performed to obtain CBCT image data.

[0055] According to the embodiments of this disclosure, the CBCT projection image processing method introduces a known spatial modulation mode by installing a modulation component with a known structure in the scanning device. The ray path passes through the modulation component, so that the signal received by the detector has unique spatial characteristics. The main ray signal and the scattered signal are separated by combining local filtering and scattering estimation algorithms, thereby removing scattering artifacts.

[0056] Furthermore, by dynamically adjusting the position or orientation of the modulation components during the scanning process, the first projection data can vary with the scanning angle. This results in multiple spatial modulation patterns being superimposed on the various projection data obtained throughout the scanning process. Compared to using only one spatial modulation pattern for projection data at all angles, the energy difference patterns provided by multiple spatial modulation patterns are varied, and the spatial distribution pattern of the hardening effect caused by the modulator differs at different angles. This can disrupt the consistency of artifacts. When the CT reconstruction algorithm reconstructs projection data from all angles, these angle-variable errors cannot be coherently superimposed, making it difficult to form fixed, high-contrast annular artifacts. This avoids introducing new artifacts during artifact removal.

[0057] Figure 2 A top view schematic diagram of a scanning device equipped with a modulation component according to some embodiments of this disclosure is shown. (See also...) Figure 2 The modulation element Md can be cylindrical and surround the detector De on the horizontal side of the scanning device. This ensures that only one layer of modulator is in the optical path of the X-rays during each scan. The modulation element Md can be a hollow cylindrical structure, and the cylinder can be designed to be relatively thin. The continuous curved surface of the cylinder ensures that the modulation pattern changes continuously and non-periodically with angle. This allows the hardening effect caused by the modulation element to be spatially averaged. Ultimately, in the reconstructed image, these artifacts do not appear as ring artifacts, but are transformed into a low-level, noise-like background, reducing interference with the image.

[0058] The inner wall of the cylinder faces the detector De, and the outer wall faces the radiation source Rs. As the radiation source Rs and the detector De rotate around the rotation center axis of the CBCT, the modulation component Md will also rotate around the rotation center axis along with the detector De, and the distance between the detector De and the modulation component Md remains constant.

[0059] As detector De rotates around its central axis of rotation, modulation component Md also rotates on its own axis. Figure 2 The crosshairs in the image represent the rotation axis of the modulation component Md. The rotation axis can be the geometric axis of a cylinder. A CBCT scan is performed to obtain the flat-field image P0 and the initial projection image P. c During the process, the modulation component Md revolves around the rotation center axis of CBCT while also rotating on its own axis, causing the position where the ray passes through the modulation component Md to change, thereby changing the spatial modulation pattern.

[0060] The revolution angle and rotation angle of the modulation component Md can be 1:1, meaning that for every 1 degree it rotates around the rotation center axis of the CBCT, it also rotates 1 degree. Understandably, this ratio can also be other ratios.

[0061] In some implementations, the modulation assembly can comprise N structural components, each with different transmission characteristics for X-rays. Spatial modulation patterns are formed using these components with varying transmission characteristics. These different transmission characteristics can be achieved using different materials; for example, the modulation assembly could include both plastic and metal structural components, assembled together. A higher value for N results in a more refined modulation effect, but also increases the computational complexity. Introducing structural components with diverse transmission characteristics enhances the spatial resolution of the modulation assembly, leading to more accurate scattering estimation. Furthermore, the diverse transmission characteristics improve the algorithm's adaptability to different types of subjects.

[0062] In some implementations, N=2. The aforementioned N types of structural components may include multiple first-type structural components and one second-type structural component. The first-type structural components have lower radiation transmission characteristics than the second-type structural components. The multiple first-type structural components may be distributed in a grid pattern on the second-type structural component.

[0063] When N=2, the N mask regions applied in step S110 are 2, namely the high-transmission mask region U. h and low-transmission mask region U l In step S110, four filtered projection data points are obtained for each scanning angle, two of which correspond to the flat-field image P0 (the first projection data), namely the high-transmission flat-field image P0. h and low transmission flat field image P l The other two correspond to the initial projected image P. c V, respectively, are high-transmission projection images H and low-transmission projection image V L .

[0064] Figure 3 A schematic diagram of the modulation component according to some embodiments of this disclosure is shown. (See also...) Figure 3 The base material of the cylindrical wall can be plastic, and multiple particles made of metal are set on the cylindrical wall. Figure 3 The multiple circles within represent metal particles. These particles can be cylinders or cuboids, etc., to ensure that X-rays travel approximately the same distance through the metal. The transmittance of the plastic region is higher than that of the metal region, thus creating a spatial modulation effect on the X-rays. The metal particles can be randomly or regularly arranged on the plastic cylinder wall, such as in a matrix, a uniform grid, or other arrangements.

[0065] By distributing high-attenuation metal particles in a grid pattern on a low-attenuation plastic substrate, a clear and stable modulation signal can be generated while ensuring sufficient X-ray flux.

[0066] Figure 4 A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of this disclosure is shown. See also... Figure 4 Before performing step S110 (local smoothing filtering of the target projection data), steps S101 and S102 are performed first. Step S101 yields the flat-field image P0, and step S102 yields the high-transmittance mask region U. h and low-transmission mask region U l .

[0067] S101, without including the subject, CBCT scanning is performed using a scanning device equipped with a modulation component to obtain first projection data.

[0068] The X-ray source Rs is activated without any object being scanned. The X-rays pass through the annular modulation assembly and reach the detector De, resulting in the first projection data, a flat-field image P0, which represents the X-ray intensity distribution and is only affected by the modulator. The flat-field image P0 roughly presents a pattern of alternating high transmission (plastic areas) and low transmission (metal particle areas). Figure 3 Under the modulation assembly shown, the flat-field image P0 generally appears as a bright background (plastic area) with multiple regularly distributed horizontal dark stripes or elliptical spots (metal particles). The flat-field image P0 contains complete spatial modulation information of the modulation assembly, which can be used to identify the modulation region, estimate the ideal flat field, and serve as a reference for final scattering correction.

[0069] S102, threshold segmentation is performed on the first projection data to obtain N mask regions.

[0070] Thresholding segmentation is used to classify pixels in an image into different categories by setting one or more thresholds. For example, it can be used to distinguish and segment high-transmittance mask regions U. h and low-transmission mask region U l .

[0071] This implementation method reduces interference caused by uneven X-ray source intensity and detector response differences by employing a strategy of smoothing before segmentation, thereby improving the accuracy of the mask region.

[0072] Figure 5 A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of this disclosure is shown. See also... Figure 5 In step S102, the method of thresholding the first projection data to obtain N mask regions may specifically include the following steps S102a and S102b.

[0073] S102a, the first projection data is smoothed by filtering to obtain smoothed projection data, and the difference between the first projection data and the smoothed projection data is determined.

[0074] The flat-field image P0 (first projection data) contains pattern information of the modulation components and may also be superimposed with low-frequency background noise caused by uneven distribution of X-ray source intensity (usually brighter in the center and darker at the edges) and differences in detector pixel response. This may cause the pixel values ​​in the central region of the high-transmission area to be higher than those in the edge region.

[0075] Therefore, the flat-field image P0 can be smoothed using a Gaussian filter. The Gaussian function is used as the convolution kernel to blur the image, reducing high-frequency details (i.e., alternating bright and dark patterns) caused by the modulation components, while preserving the inherent low-frequency inhomogeneities of the source and detector, resulting in a smooth baseline image Q0 (smoothed projection data). The baseline image Q0 has the same size as the flat-field image P0. The baseline image Q0 generally exhibits a slight brightness gradient from the center to the edges, representing the ideal flat-field image without the modulation components.

[0076] Calculate the difference image between the flat field image P0 and the baseline image Q0. By subtracting the baseline image Q0 to remove the global, slowly changing background, the resulting difference image is used to reflect the local intensity changes caused by the modulation components.

[0077] S102b: A high-transmittance mask region is formed by the pixel positions corresponding to the difference above the first pixel threshold, and a low-transmittance mask region is formed by the pixel positions corresponding to the difference below the second pixel threshold, resulting in N mask regions at the same scanning angle. N=2, and the first pixel threshold is greater than the second pixel threshold.

[0078] If the difference between the pixel value of the same pixel in the flat field image P0 and the pixel value in the baseline image Q0 is higher than the first pixel threshold, then the pixel is considered to belong to the high transmittance mask region U. h If the difference between the pixel value of the same pixel in the flat field image P0 and the pixel value in the baseline image Q0 is less than the second pixel threshold, then the pixel is considered to belong to the low-transmission mask region U. l The first pixel threshold can be a positive number, and the second pixel threshold can be a negative number. The absolute values ​​of the first and second pixel thresholds can be equal to ensure symmetry.

[0079] In the high-transmission mask region U h In the resulting mask image, the high-transmittance mask region U h The corresponding pixel has a value of 1, while the values ​​of the remaining pixels are 0. In the low-transmission mask region U... l In the resulting mask image, the low-transmission mask region U l The value of the corresponding pixel is 1, and the value of the other pixels is 0.

[0080] It is understandable that in the difference image obtained in step S101a, some pixels have a difference value lower than the first pixel threshold and higher than the second pixel threshold. These pixels do not belong to the high-transmittance mask region U. h It also does not belong to the low-transmission mask region U. l The areas formed by these pixels can be called transition regions or unclassified regions. These pixels are typically located at the transition edges of a modulation component pattern, such as a gradient area from metal particles to a plastic substrate.

[0081] This embodiment extracts the spatial modulation information of the modulation component from the background noise and uses two thresholds for segmentation, thereby improving the accuracy and symmetry of high and low transmission region identification and obtaining a high-quality region mask.

[0082] Figure 6 A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of this disclosure is shown. See also... Figure 6 In step S110, the method of performing local smoothing filtering on the target projection data according to the existing N mask regions can include steps S111 and S112. Step S111 is used to generate a high-transmission flat-field image P. h Step S112 is used to generate a low-transmission flat-field image P. l The execution order between steps S111 and S112 is arbitrary.

[0083] S111, perform local smoothing filtering on the pixel positions in the high-transmission mask regions of the existing N mask regions in the first projection data that does not contain the subject, and interpolate the pixel positions in the non-high-transmission mask regions of the first projection data using the first projection data located in the high-transmission mask regions to obtain high-transmission flat field data.

[0084] S112, perform local smoothing filtering on the pixel positions in the low-transmission mask region among the existing N mask regions in the first projection data, and interpolate the pixel positions in the non-low-transmission mask region in the first projection data using the first projection data located in the low-transmission mask region to obtain low-transmission flat field data.

[0085] Local smoothing filtering is a filtering method that combines local smoothing and spatial interpolation. Its input is an image that is partially known and partially unknown, as well as a mask marked with known regions. By performing smoothing filtering in the known regions and interpolation in the unknown regions, a complete estimated image is output.

[0086] For each pixel in the flat-field image P0 (first projection data), if the pixel belongs to the high-transmittance mask region U h(For the plastic portion), a Gaussian kernel is used to perform a weighted average of the neighboring pixels to obtain the value of that pixel. If the pixel does not belong to the high-transmittance mask region U... h Then, a Gaussian kernel is used to target the high-transmittance mask region U in the neighborhood. h Interpolation is performed on the pixels to obtain an estimated pixel value. This interpolation value is used to indicate whether the pixel belongs to the high-transmittance mask region U. h The expected intensity value. This yields the high-transmission flat-field image P. h (High transmission flat field data).

[0087] If the pixels in the flat field image P0 belong to the low-transmission mask region U l (For the metallic portion), a Gaussian kernel is used to perform a weighted average of neighboring pixels to obtain the value of that pixel. If the pixel does not belong to the low-transmission mask region U... l Then, a Gaussian kernel is used to target the low-transmission masking region U in the neighborhood. l Interpolation is performed on the pixels to obtain an estimated pixel value. This interpolation value is used to indicate whether the pixel belongs to the low-transmission mask region U. l The expected intensity value. This yields the low-transmission flat-field image P. l (Low transmission flat field data).

[0088] High transmission flat field image P h and low transmission flat field image P l The image size is the same as the flat-field image P0. High-transmission flat-field image P h Flat-field images used to represent the assumption that the entire modulation assembly is made of a high-transmittance material (such as plastic) are generally presented as smooth and bright images without dark stripes. Low-transmittance flat-field image P l Flat-field images are used to represent images where the entire modulation assembly is assumed to be made of a low-transmittance material (such as metal), and are generally presented as smooth and dark images.

[0089] This embodiment constructs high-transmission flat-field images and low-transmission flat-field images to obtain reference benchmarks for both full high-transmission and full low-transmission states.

[0090] Figure 7 A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of this disclosure is shown. See also... Figure 7 In step S110, the method of performing local smoothing filtering on the target projection data according to the existing N mask regions can include steps S113 and S114. Step S113 is used to generate a high-transmission projection image V. H Step S114 is used to generate a low-transmission projection image V LThe execution order of steps S111, S112, S113 and S114 is arbitrary. Figure 7 Steps S111 and S112 are not shown in the diagram.

[0091] S113, perform local smoothing filtering on the pixel positions in the high-transmission mask regions of the existing N mask regions in the second projection data containing the subject, and interpolate the pixel positions in the non-high-transmission mask regions in the second projection data using the second projection data located in the high-transmission mask regions to obtain high-transmission projection data.

[0092] S114, perform local smoothing filtering on the pixel positions in the low-transmission mask region among the existing N mask regions in the second projection data, and interpolate the pixel positions in the non-low-transmission mask region in the second projection data using the second projection data located in the low-transmission mask region to obtain low-transmission projection data.

[0093] For the initial projection image P c For each pixel in (second projection data), if the pixel belongs to the high-transmittance mask region U h (For the plastic portion), a Gaussian kernel is used to perform a weighted average of the neighboring pixels to obtain the value of that pixel. If the pixel does not belong to the high-transmittance mask region U... h Then, a Gaussian kernel is used to target the high-transmittance mask region U in the neighborhood. h Interpolation is performed on the pixels to obtain estimated pixel values. This yields the high-transmission projection image V. H (High-transmission projection data).

[0094] If the initial projection image P c The pixels in the image belong to the low-transmission mask region U. l (For the metallic portion), a Gaussian kernel is used to perform a weighted average of neighboring pixels to obtain the value of that pixel. If the pixel does not belong to the low-transmission mask region U... l Then, a Gaussian kernel is used to target the low-transmission masking region U in the neighborhood. l Interpolation is performed on the pixels to obtain an estimated pixel value. This interpolation value is used to indicate whether the pixel belongs to the low-transmission mask region U. l The required intensity value at that time. This yields the low-transmission projection image V. L (Low-transmission projection data).

[0095] High transmission projection image V H and low-transmission projection image V L Image size and initial projected image P c Same. High-transmission projection image V HThis is used to represent a projected image assuming the entire modulation assembly is made of a high-transmittance material (such as plastic). Low-transmittance projected image V L Used to represent a projected image assuming the entire modulation assembly is made of a low-transmittance material (such as metal).

[0096] Figure 8 A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of this disclosure is shown. See also... Figure 8 In step S130, the method of obtaining the first scattering estimate of each pixel position under the scanning angle by performing scattering estimation based on 2N filtered projection data, first projection data and second projection data of each scanning angle can include the following steps S131 and S132.

[0097] S131, for the same pixel position under each of the multiple scanning angles, determine the first quotient of the second projection data and the first projection data, and determine the second quotient of the low transmission flat field data and the high transmission flat field data among the 2N filtered projection data.

[0098] S132, determine the first scattering estimate of each pixel position under the scanning angle based on the first quotient and the second quotient.

[0099] For each pixel, the first quotient D1 can be represented as P. c / P0, the first quotient D1 reflects the total attenuation along the ray path corresponding to the pixel. The second quotient D2 can be represented as P l / P h D2 describes the ratio of the signal intensity in the low-transmission region to the signal intensity in the high-transmission region under ideal no-scattering conditions. D2 is determined by the physical characteristics of the modulation components (such as the thickness and material of metals and plastics) and is independent of the object being scanned. The first scattering estimate at the pixel is obtained by calculating D1 and D2, thus obtaining the preliminary scattering estimate map G1.

[0100] This implementation combines the inherent modulation characteristics with the object information of the current scanning scene through two quotients. This separate processing method allows the algorithm to be applied to different subjects while maintaining high sensitivity to the scattered signal.

[0101] Figure 9 A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of this disclosure is shown. See also... Figure 9 In step S132, the method of determining the first scattering estimate of each pixel position under the scanning angle based on the first quotient and the second quotient may include steps S133 and S135.

[0102] S133, determine the first product of the first quotient raised to the power of f and the low-transmission projection data in the 2N filtered projection data, and determine the second quotient and the second product of the high-transmission projection data in the 2N filtered projection data. Where f is a specified value.

[0103] S135, determine the first scattering estimate of each pixel position under the scanning angle based on the first product and the second product.

[0104] The first product U1 can be represented as D1 f ×V L f is used to adjust the intensity of hardening compensation and can be calibrated according to the source and the scanned object. Since the first quotient D1 reflects the degree of attenuation, the greater the attenuation, the more severe the beam hardening (the phenomenon where low-energy photons are preferentially absorbed when the beam passes through an object, leading to an increase in the average energy of the remaining beam) becomes, and the weaker the effective blocking capability of the modulation component, meaning the actual transmittance will increase. Correcting this with f compensates for the overestimation of scattering caused by hardening, improving the accuracy of correction on complex objects (such as the human body with skeletons). The second product U2 can be expressed as D2 × V H The first scattering estimate at the pixel is obtained by performing operations on U1 and U2, thus obtaining the preliminary scattering estimate map G1.

[0105] This implementation improves robustness and practicality by introducing a power operation of the empirical parameter f to correct the non-uniform hardening effect introduced by the fixed modulator.

[0106] Figure 10 A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of this disclosure is shown. See also... Figure 10 In step S135, the method of determining the first scattering estimate of each pixel position under the scanning angle based on the first product and the second product may include steps S135a and S135b.

[0107] S135a, determine the first difference between the first product and the second product, and determine the second difference between the value 1 and the second quotient.

[0108] S135b calculates the quotient of the first difference and the second difference to obtain the first scattering estimate of each pixel position under the scanning angle.

[0109] The first difference J1 can be expressed as U1-U2, and the second difference J2 can be expressed as 1-D2. The first scattering estimate can be expressed as (U1-U2) / (1-D2), and the complete formula for calculating the first scattering estimate is: [(P c / P0) f ×V L -(P l / P h ×V H )] / [1-(P l / P h This formula separates the high-frequency modulation structure information from the low-frequency scattering information, resulting in a preliminary scattering estimation image G1 that only reflects the scattering distribution and does not include the periodic stripes of the modulation components.

[0110] Understandably, due to the high transmission projection image V H and low-transmission projection image V L The absolute intensities are different, therefore it can be determined by (P) l / P h ) to V H and V L Normalization is performed so that V can be normalized at the same scale. H and V L The calculations are performed to separate the scattering components from the mixed signal.

[0111] This implementation can separate high-frequency modulation structure information from low-frequency scattering information, directly generating a preliminary estimate of the scattering signal.

[0112] In some implementations, regarding the trusted region in step S150 (locally smoothing and filtering the first scattering estimate at the same scanning angle according to the trusted region), the first scattering estimate of each pixel position in the trusted region is positive and less than the pixel value of the corresponding pixel position in the second projection data.

[0113] When determining the reliable region R, in addition to requiring it to belong to the high-transmission mask region U, h and low-transmission mask region U l In addition to being a proper subset of the union of the two sets, the confidence region R can be constrained by the following conditions: Condition 1: The first scattering estimate of the pixels in the confidence region is greater than zero. Condition 2: The first scattering estimate of the pixels in the confidence region is less than the initial projected image P. c The values ​​at the corresponding positions are used. Condition 1 is used to filter out pixels with negative scattering intensity. Condition 2 is used to filter out pixels with intensity equal to or exceeding the total received signal. This ensures that pixels with reasonable values ​​and conforming to the physical laws of imaging are used for final scattering correction, improving the fidelity and stability of the reconstructed image.

[0114] In other embodiments, the aforementioned reliable region can also be constrained by condition 3. Condition 3: The gradient value of each pixel position in the reliable region among the high-transmission projection data and low-transmission projection data in the 2N filtered projection data is less than a gradient threshold. The gradient value is used to represent the magnitude of the numerical change of the pixel.

[0115] Due to the high transmission projection image V H and low-transmission projection image V L The algorithm relies heavily on interpolation, and pixel values ​​can change abruptly (with large gradients) at object boundaries or in severely hardened areas, resulting in significant errors. Condition 3 filters out high-frequency regions such as object boundaries, preventing these large error values ​​from degrading the quality of the final scattering map. This ensures that only pixels with sufficiently small gradients (i.e., located in smooth regions) are used to generate the final scattering map. If any of the three conditions above are not met, the pixel is not considered a reliable region.

[0116] The gradient value can be calculated using numerical differentiation methods, such as the Sobel or Scharr operators. The calculation of the gradient value may include the following steps: First, apply a horizontal gradient convolution kernel to the high-transmittance projection image V... H and low-transmission projection image V L Perform convolution to obtain the horizontal gradient components of each pixel, and then apply the vertical gradient convolution kernel to V. H and V L Perform convolution to obtain the vertical gradient components of each pixel. Then, calculate the gradient magnitude of the pixel based on the horizontal and vertical gradient components.

[0117] A trustworthy region can be represented as a binary mask. For a trustworthy region after multiple constraints, if there are strong edges such as bones or air cavities inside the subject, then due to the large gradient values ​​near these edges, thin black lines or regions will appear to represent untrustworthy regions.

[0118] Figure 11 A schematic flowchart of a CBCT projection image processing method M100 according to other embodiments of this disclosure is shown. See also... Figure 11 After obtaining the artifact-corrected projection data through the difference between the second projection data and the second scattering estimate (step S170), step S190 can be executed. Step S190 is used to ensure that the artifact-corrected projection data is non-negative.

[0119] S190, the difference between the second projection data and the second scattering estimate is used as the third difference. For the third difference that is less than the difference threshold, a monotonic function is used to map the third difference to a new value greater than zero, so as to smooth the projection data after artifact correction. Here, the difference threshold is a positive value.

[0120] During scattering estimation, the scattering intensity of some pixels may be overestimated, causing the third difference D3 to be less than zero. Therefore, negative value regions can be smoothed to avoid introducing discontinuities. The monotonic function can be (ε / e)×exp(D3 / ε), where ε is the difference threshold, e is the base of the natural logarithm, and exp() is a function with base e. ε can be a small positive number, such as e^(-ε / ε). -3 or e -4 When D3 > ε, the original value of the second scattering estimate is directly output. When D3 = ε, the output value is ε, making the value continuous. When D3 < ε, the output value approaches 0 but will not equal 0 or be negative. This ensures that the smoothed value is always positive, and its first derivative is continuous. Therefore, the signal change is smooth when transitioning from the positive signal region to the corrected negative signal region, avoiding the introduction of new artifacts.

[0121] This implementation method ensures physical rationality and signal smoothness, avoids introducing new artifacts due to hard truncation, and ensures the quality of the final reconstructed image. Based on any of the above embodiments, this disclosure also provides a CBCT projection image processing system. See also... Figure 2 The CBCT projection image processing system includes: a scanning device, a modulation component Md, a memory, and a processor. Figure 2 (Not shown in the image). The scanning device includes a radiation source Rs and a detector De. The memory stores execution instructions, and the processor executes the execution instructions stored in the memory, causing the processor to perform the CBCT projection image processing method M100 of any of the above embodiments.

[0122] Figure 12 This is a schematic block diagram of a CBCT projection image processing apparatus according to one embodiment of the present disclosure. Figure 12 As shown, the processing device includes: a filtering module 110, a first scattering estimation module 130, a second scattering estimation module 150, and an artifact correction module 170.

[0123] The filtering module 110 is used to take the first projection data that does not contain the subject and the second projection data that contains the subject as target projection data, and to perform local smoothing filtering on the target projection data according to the existing N mask regions to obtain 2N filtered projection data, where N>1. The target projection data is obtained by CBCT scanning through a scanning device equipped with a modulation component. The modulation component is used to perform spatial selective attenuation of the rays emitted by the scanning device, thereby forming N spatial regions corresponding to N kinds of transmission characteristics in the first projection data. Different mask regions correspond to the pixel positions of the N spatial regions. The modulation component makes the first projection data dynamically change with the scanning angle.

[0124] The first scattering estimation module 130 is used to perform scattering estimation based on 2N filtered projection data, first projection data and second projection data of each scanning angle in multiple scanning angles, to obtain the first scattering estimation value of each pixel position under the scanning angle.

[0125] The second scattering estimation module 150 is used to perform local smoothing filtering on the first scattering estimation value under the same scanning angle according to the reliable region, so as to obtain the second scattering estimation value of each pixel position under the scanning angle. The reliable region is contained in the union of N mask regions.

[0126] The artifact correction module 170 is used to obtain the artifact-corrected projection data by the difference between the second projection data and the second scattering estimate.

[0127] The aforementioned CBCT projection image processing device can be in the form of computer software, and each module of the processing device can be implemented through computer software modules. The specific implementation process of the functions and roles of each module in the aforementioned device is detailed in the corresponding steps of the above method, and will not be repeated here.

[0128] The execution subject of the CBCT projection image processing method in the specific embodiments of this disclosure can be an electronic device such as a computer.

[0129] Therefore, based on any of the above embodiments, this disclosure also provides an electronic device that can perform the CBCT projection image processing method of any of the embodiments described above.

[0130] Figure 13 This is a schematic block diagram of an electronic device 1000 according to one embodiment of the present disclosure.

[0131] The hardware architecture of the electronic device 1000 can be implemented using a bus architecture. The bus architecture can include any number of interconnect buses and bridges, depending on the specific application of the hardware and overall design constraints. Bus 1100 connects various circuits, including one or more processors 1200, memory 1300, and / or hardware modules. Bus 1100 can also connect various other circuits 1400, such as peripheral devices, voltage regulators, power management circuits, external antennas, etc.

[0132] Bus 1100 can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, this diagram uses only one connection line, but this does not imply that there is only one bus or one type of bus.

[0133] The processor 1200 can be a central processing unit (CPU). The processor 1200 can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations of the above types of chips.

[0134] The memory 1300 can serve as a non-transitory computer-readable storage medium, used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as program instructions of the computer program in the embodiments of this disclosure. The processor 1200 implements the CBCT projection image processing method by running the non-transitory software programs, instructions, and modules stored in the memory 1300.

[0135] The memory 1300 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and application programs required for at least one function; the data storage area may store data created by the processor 1200. Furthermore, the memory 1300 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 1300 may optionally include memory remotely located relative to the processor 1200, and these remote memories may be connected to the processor 1200 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0136] This disclosure also provides a readable storage medium storing a computer program that, when executed by a processor, is used to implement the methods described above. A "readable storage medium" can be any means capable of containing, storing, communicating, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples of a readable storage medium include: an electrical connection with one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable read-only memory (CD-ROM), etc.

[0137] This disclosure also provides a computer program product, the methods of which can be implemented wholly or partially through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented wholly or partially as a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed, the processes or functions of this disclosure are performed wholly or partially. The computer can be a general-purpose computer, a special-purpose computer, a computer network, network equipment, user equipment, core network equipment, OAM, or other programmable device.

[0138] Computer programs or instructions can be stored in a readable storage medium or transferred from one readable storage medium to another. For example, the computer program or instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The readable storage medium can be any available medium capable of access, or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; an optical medium, such as a digital video optical disc; or a semiconductor medium, such as a solid-state drive. The computer-readable storage medium can be a volatile or non-volatile storage medium, or it can include both volatile and non-volatile types of storage media.

[0139] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, apparatus, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus, and computer program products according to this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, produce a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0141] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0143] In the description of this specification, the references to terms such as "one embodiment / mode," "some embodiments / modes," "example," "specific example," or "some examples," etc., refer to specific features, structures, or characteristics described in connection with that embodiment / mode or example, which are included in at least one embodiment / mode or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment / mode or example. Moreover, the specific features, structures, or characteristics described may be combined in any suitable manner in one or more embodiments / modes or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments / modes or examples described in this specification, as well as the features of different embodiments / modes or examples.

[0144] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this disclosure, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0145] Those skilled in the art should understand that the above embodiments are merely for illustrating the present disclosure and are not intended to limit the scope of the disclosure. Those skilled in the art can make other changes or modifications based on the above disclosure, and these changes or modifications still fall within the scope of the present disclosure.

Claims

1. A method for processing CBCT projection images, characterized in that, include: The first projection data without the subject and the second projection data containing the subject are used as target projection data. The target projection data are locally smoothed and filtered according to the existing N mask regions to obtain 2N filtered projection data, where N>1. The target projection data are obtained by CBCT scanning using a scanning device equipped with a modulation component. The modulation component is used to spatially selectively attenuate the rays emitted by the scanning device, thereby forming N spatial regions corresponding to N kinds of transmission characteristics in the first projection data. Different mask regions correspond to the pixel positions of the N spatial regions. The modulation component makes the first projection data dynamically change with the scanning angle. Scattering estimation is performed based on the 2N filtered projection data, the first projection data, and the second projection data for each of the multiple scanning angles to obtain the first scattering estimate value of each pixel position under the scanning angle. Local smoothing filtering is performed on the first scattering estimate at the same scanning angle according to the trusted region to obtain the second scattering estimate at each pixel position at the scanning angle. The trusted region is contained in the union of the N mask regions. as well as The artifact-corrected projection data is obtained by using the difference between the second projection data and the second scattering estimate.

2. The method for processing CBCT projection images according to claim 1, characterized in that, The modulation component is cylindrical and surrounds the horizontal side of the detector in the scanning device.

3. The method for processing CBCT projection images according to claim 1 or 2, characterized in that, The modulation assembly includes N types of structural components, each with different transmission characteristics for rays.

4. The method for processing CBCT projection images according to claim 3, characterized in that, N=2, the N types of structural components include multiple first-type structural components and one second-type structural component. The transmission characteristics of the first-type structural components to rays are less than those of the second-type structural components to rays. The multiple first-type structural components are distributed in a grid pattern on the second-type structural component.

5. The method for processing CBCT projection images according to claim 1, characterized in that, Before performing local smoothing filtering on the target projection data, the method further includes: First projection data is obtained by CBCT scanning using a scanning device equipped with a modulation component without including the subject. The first projection data is thresholded to obtain N mask regions.

6. The method for processing CBCT projection images according to claim 5, characterized in that, Threshold segmentation of the first projection data yields N mask regions, including: The first projection data is smoothed by filtering to obtain smoothed projection data, and the difference between the first projection data and the smoothed projection data is determined; and A high-transmission mask region is formed by the pixel position corresponding to the difference value above the first pixel threshold, and a low-transmission mask region is formed by the pixel position corresponding to the difference value below the second pixel threshold, resulting in N mask regions at the same scanning angle, where N=2, and the first pixel threshold is greater than the second pixel threshold.

7. The method for processing CBCT projection images according to claim 1, characterized in that, Local smoothing filtering is performed on the target projection data according to the existing N mask regions, including: Local smoothing filtering is applied to the pixel positions within the high-transmission mask regions of the existing N mask regions in the first projection data that does not contain the subject. Then, the pixel positions within the non-high-transmission mask regions of the first projection data are interpolated using the first projection data located within the high-transmission mask regions to obtain high-transmission flat-field data; and The pixel positions in the first projection data located within the low-transmission mask region of the existing N mask regions are locally smoothed and filtered. The pixel positions in the first projection data located within the low-transmission mask region are interpolated to obtain low-transmission flat field data.

8. The method for processing CBCT projection images according to claim 1 or 7, characterized in that, Local smoothing filtering is performed on the target projection data according to the existing N mask regions, including: Local smoothing filtering is performed on the pixel positions within the high-transmission mask regions of the existing N mask regions in the second projection data containing the subject. Then, the pixel positions within the non-high-transmission mask regions of the second projection data are interpolated using the second projection data located within the high-transmission mask regions to obtain high-transmission projection data. The pixel positions in the second projection data located within the low-transmission mask region of the existing N mask regions are locally smoothed and filtered. The pixel positions in the second projection data located within the low-transmission mask region are interpolated to obtain the low-transmission projection data.

9. The method for processing CBCT projection images according to claim 1, characterized in that, Scattering estimation is performed based on the 2N filtered projection data, the first projection data, and the second projection data for each of the multiple scanning angles to obtain the first scattering estimate value for each pixel position at the scanning angle, including: For the same pixel position at each of the multiple scanning angles, determine a first quotient between the second projection data and the first projection data, and determine a second quotient between the low-transmission flat field data and the high-transmission flat field data in the 2N filtered projection data; and The first scattering estimate of each pixel position at the scanning angle is determined based on the first quotient and the second quotient.

10. The method for processing CBCT projection images according to claim 9, characterized in that, Determining the first scattering estimate of each pixel position at the scanning angle based on the first quotient and the second quotient includes: Determine the first product of the f-th power of the first quotient and the first product of the low-transmission projection data in the 2N filtered projection data, and determine the second product of the second quotient and the second product of the high-transmission projection data in the 2N filtered projection data, where f is a specified value; and The first scattering estimate of each pixel position at the scanning angle is determined based on the first product and the second product.

11. The method for processing CBCT projection images according to claim 10, characterized in that, Determining the first scattering estimate of each pixel position at the scanning angle based on the first product and the second product includes: Determine a first difference between the first product and the second product, and determine a second difference between the value 1 and the second quotient; and The quotient of the first difference and the second difference is calculated to obtain the first scattering estimate of each pixel position at the scanning angle.

12. The method for processing CBCT projection images according to claim 1, characterized in that, The first scattering estimate of each pixel location in the trusted region is positive and less than the pixel value of the corresponding pixel location in the second projection data.

13. The method for processing CBCT projection images according to claim 1 or 12, characterized in that, The gradient value of each pixel position in the trusted region is less than the gradient threshold in the high-transmission projection data and low-transmission projection data of the 2N filtered projection data.

14. The method for processing CBCT projection images according to claim 1, characterized in that, After obtaining the artifact-corrected projection data through the difference between the second projection data and the second scattering estimate, the processing method includes: The difference between the second projection data and the second scattering estimate is used as the third difference. For the third difference that is less than the difference threshold, a monotonic function is used to map the third difference to a new value greater than zero in order to smooth the projection data after artifact correction. The difference threshold is a positive value.

15. A CBCT projection image processing system, characterized in that, include: Scanning equipment, including a radiation source and a detector; Modulation components; The memory stores execution instructions; as well as A processor that executes execution instructions stored in the memory, causing the processor to perform the CBCT projection image processing method according to any one of claims 1 to 14.

16. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, is used to implement the CBCT projection image processing method according to any one of claims 1 to 14.

Images