CBCT reconstruction method

Abstract

The invention discloses a CBCT (Cone Beam Computed Tomography) reconstruction method, which comprises the following steps of: reading a plurality of images to be reconstructed at different angles; performing pre-reconstruction operation on the plurality of to-be-reconstructed images based on a preset reconstruction center and a preset reconstruction range to generate a pre-reconstructed image; respectively setting a target reconstruction center and a target reconstruction range on the pre-reconstructed image; and based on the target reconstruction center and the target reconstruction range, performing image reconstruction operation on the plurality of to-be-reconstructed images to generate a three-dimensional image. According to the method and the device, a user can intuitively adjust the target reconstruction center and the target reconstruction range so as to reconstruct the region of interest and adjust the spatial resolution of the reconstructed image according to requirements, and observation and diagnosis of diseases at different positions and with different sizes are facilitated.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging technology, and in particular to a CBCT reconstruction method. Background Art

[0002] As is well known, CBCT (Cone Beam Computed Tomography) is a technology that uses a large matrix X-ray detector to perform X-ray imaging of an object or the human body at multiple angles, and then uses a reconstruction algorithm for reconstruction to obtain a three-dimensional image. In order to obtain X-ray images at multiple angles, usually the X-ray detector and the gantry need to rotate around the object or the human body as the center, or the detector and the gantry are fixed while the human body or the object rotates.

[0003] In the prior art, generally the reconstruction center of CBCT is fixed, corresponding to the center of rotation, resulting in that when performing reconstruction, the reconstruction center cannot be aligned with the area of interest or the lesion. And since the reconstruction range and the matrix size are also fixed values, the spatial resolution of the reconstructed image is fixed and cannot be adjusted according to requirements. When high-resolution scanning and reconstruction of some local small areas are required clinically, the reconstruction center cannot be adjusted to align with the lesion position, nor can the reconstruction range be reduced to improve the spatial resolution, which is not convenient for observing and diagnosing local small lesions. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: to provide a CBCT reconstruction method to solve the problems that the existing CBCT reconstruction method cannot adjust the reconstruction center and the target reconstruction range.

[0005] To solve the above technical problem, the technical solution adopted by the present invention is: a CBCT reconstruction method, which includes the following steps: S1: Read multiple to-be-reconstructed images at different angles; S2: Perform a pre-reconstruction operation on the multiple to-be-reconstructed images based on a preset reconstruction center and a preset reconstruction range to generate a pre-reconstructed image; S3: Set a target reconstruction center and a target reconstruction range on the pre-reconstructed image respectively; S4: Perform an image reconstruction operation on the multiple to-be-reconstructed images based on the target reconstruction center and the target reconstruction range to generate a three-dimensional image.

[0006] Further, in the CBCT reconstruction method of the present invention, in step S3, setting the target reconstruction center on the pre-reconstructed image includes: Select a reconstruction area on the pre-reconstructed image; Determine the regional center of the reconstruction area, and use the regional center as the target reconstruction center.

[0007] Further, in the CBCT reconstruction method of the present invention, in step S3, setting a target reconstruction center on the pre-reconstructed image includes: Select a target position on the pre-reconstructed image, and use the target position as the target reconstruction center.

[0008] Further, in the CBCT reconstruction method of the present invention, in step S3, setting a target reconstruction range on the pre-reconstructed image includes: Select a circular reconstruction area on the pre-reconstructed image, and use the area diameter of the circular reconstruction area as the target reconstruction range.

[0009] Further, in the CBCT reconstruction method of the present invention, in step S3, setting a target reconstruction range on the pre-reconstructed image includes: Obtain a preset range list, where the range list includes multiple preset ranges; Select a target range from the multiple preset ranges in the range list, and use the target range as the target reconstruction range.

[0010] Further, in the CBCT reconstruction method of the present invention, in step S4, based on the target reconstruction center and the target reconstruction range, perform an image reconstruction operation on the multiple images to be reconstructed to generate a three-dimensional image, including: Perform a preprocessing operation on each of the multiple images to be reconstructed to obtain corresponding multiple preprocessed images; Based on a conversion formula, convert each of the multiple preprocessed images into a corresponding multiple attenuation maps; Perform a filtering operation on each of the multiple attenuation maps to obtain corresponding multiple filtered images; Based on a reconstruction formula, perform back-projection reconstruction on the multiple filtered images to obtain the three-dimensional image.

[0011] Further, in the CBCT reconstruction method of the present invention, in step S4, the preprocessing operation includes the following steps: Perform bad pixel retrieval and correction on the image to be reconstructed; Determine whether there is an abnormal edge in the corrected image to be reconstructed. If so, compensate for the abnormal edge.

[0012] Further, in the CBCT reconstruction method of the present invention, in step S4, performing bad pixel retrieval and correction on the image to be reconstructed includes: Perform pixel inspection on the image to be reconstructed, and determine the difference between the pixel value of the target pixel point in the image to be reconstructed and the neighborhood pixel mean value; Determine whether the difference exceeds a preset threshold. If it exceeds the preset threshold, determine that the target pixel point is a bad pixel, and correct the bad pixel based on the neighborhood pixel mean value.

[0013] Further, in the CBCT reconstruction method of the present invention, in step S4, the conversion formula is:

[0014] Wherein, Ai represents the i-th attenuation map; Log(I0 / Ii) represents the natural logarithm of the ratio of I0 to Ii, Log represents the natural logarithm, I0 is the image without attenuation, Ii represents the i-th preprocessed image, and i represents the image serial number, i = 1, 2, 3... N.

[0015] Further, in the CBCT reconstruction method of the present invention, in step S4, the reconstruction formula is:

[0016] Wherein, represents the pixel value at the spatial coordinate (x, y, z) of the three-dimensional image, wherein, (x, y, z) is the coordinate of the three-dimensional image in the three-dimensional space; represents the sum of the attenuation values of all rays passing through the coordinate (x, y, z), wherein, represents the pixel value of the filtered image Fi at the coordinate Fi represents the i-th filtered image, and i represents the image serial number, i = 1, 2, 3... N, represents the coordinate where the ray passing through (x, y, z) hits the image to be reconstructed.

[0017] The beneficial effects of the present invention are as follows: The present invention is a CBCT reconstruction method that can achieve a custom center and range. It can perform reconstruction on a specific area, facilitating the observation and diagnosis of diseases at different positions and of different sizes. Specifically: First, read multiple to-be-reconstructed images at different angles. Then, based on a preset reconstruction center and a preset reconstruction range, perform a pre-reconstruction operation on the multiple to-be-reconstructed images to generate a pre-reconstructed image. By generating the pre-reconstructed image, the user can directly reset the target reconstruction center and the target reconstruction range on the pre-reconstructed image, facilitating the user's intuitive adjustment. On this basis, perform an image reconstruction operation based on the user-defined target reconstruction center and target reconstruction range to generate a three-dimensional image with a custom position and a custom range. In summary, the present invention enables the user to intuitively adjust the target reconstruction center and the target reconstruction range, and then perform reconstruction on the region of interest and adjust the spatial resolution of the reconstructed image according to needs, facilitating the observation and diagnosis of diseases at different positions and of different sizes. When high-resolution scanning and reconstruction of certain local small regions are required clinically, the reconstruction center can be adjusted to align with the lesion position, and at the same time, the target reconstruction range can be reduced to improve the spatial resolution, facilitating the observation and diagnosis of local small lesions. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a flowchart of the steps of the CBCT reconstruction method according to the present invention in one embodiment; Figure 2 It is a schematic diagram of the frequency domain of the ideal ramp filter according to the present invention; Figure 3 It is a schematic diagram of a windowing form during filtering in the CBCT reconstruction method according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] To describe the technical content, achieved objectives, and effects of the present invention in detail, the following is described in conjunction with embodiments and with reference to the accompanying drawings.

[0020] Please refer to Figures 1 to 3 , the present invention discloses a CBCT reconstruction method, which includes the following steps: S1: Read multiple to-be-reconstructed images at different angles; S2: Based on a preset reconstruction center and a preset reconstruction range, perform a pre-reconstruction operation on the multiple to-be-reconstructed images to generate a pre-reconstructed image; S3: Set a target reconstruction center and a target reconstruction range on the pre-reconstructed image respectively; S4: Based on the target reconstruction center and the target reconstruction range, perform an image reconstruction operation on the multiple to-be-reconstructed images to generate a three-dimensional image.

[0021] As can be seen from the above description, the beneficial effects of the present invention are as follows: The present invention is a CBCT reconstruction method that can achieve a custom center and range, which can perform reconstruction on a specific area, facilitating the observation and diagnosis of diseases at different positions and of different sizes. Specifically: First, read multiple to-be-reconstructed images at different angles. Then, perform a pre-reconstruction operation on the multiple to-be-reconstructed images based on a preset reconstruction center and a preset reconstruction range to generate a pre-reconstructed image. By generating the pre-reconstructed image, the user can directly reset the target reconstruction center and the target reconstruction range on the pre-reconstructed image, facilitating the user's intuitive adjustment. On this basis, perform an image reconstruction operation based on the target reconstruction center and the target reconstruction range to generate a three-dimensional image for a custom position and a custom range. To sum up, the present invention enables the user to intuitively adjust the target reconstruction center and the target reconstruction range, and then perform reconstruction on the region of interest and adjust the spatial resolution of the reconstructed image according to requirements, facilitating the observation and diagnosis of diseases at different positions and of different sizes. When high-resolution scanning and reconstruction of certain local small areas are required clinically, the reconstruction center can be adjusted to align with the lesion position, and at the same time, the target reconstruction range can be reduced to improve the spatial resolution, facilitating the observation and diagnosis of local small lesions.

[0022] In practical applications, the present invention can align the reconstruction center with the lesion position, or increase or decrease the target reconstruction range to obtain tomographic images with appropriate spatial resolution, facilitating the observation and diagnosis of diseases at different positions and of different sizes. For example, when imaging a small lesion, specify the reconstruction center to the location where the lesion is located, reduce the target reconstruction range, and keep the same matrix to increase the sampling rate and thus improve the spatial resolution, enabling finer lesions to be distinguished; for another example, when imaging parts such as both lower limbs, where high resolution is not required but a larger target reconstruction range is needed to ensure that both feet can be observed simultaneously, the target reconstruction range can be increased, and the target reconstruction center can be specified between the two feet, facilitating the overall observation of lesions such as deformities of both lower limbs.

[0023] The following combines Figure 1 , and elaborates in detail on each step in steps S1 to S4 and other optional steps.

[0024] Step S1, read multiple to-be-reconstructed images at different angles; CBCT (Cone Beam Computed Tomography) is a medical imaging technology mainly used to obtain three-dimensional images. Its working principle: Using the cone beam X-ray projection technology, after circular digital projection, a three-dimensional image is obtained using computer recombination technology. During the scanning process, the X-ray generator performs circular digital projection around the object to be projected with a relatively low radiation dose to obtain multiple to-be-reconstructed images at different angles. After multiple projections, the obtained data is recombined by a computer, and finally a three-dimensional image is generated.

[0025] The to-be-reconstructed image Pi, also called the projection image. In practical applications, it may include multiple to-be-reconstructed images at different (e.g., N) angles. Before performing the pre-reconstruction operation, the to-be-reconstructed image Pi is read in first, where i = 1, 2, 3....N, and i represents the image serial number, that is, the projection image serial number.

[0026] Step S2, perform a pre-reconstruction operation on the multiple to-be-reconstructed images based on a preset reconstruction center and a preset reconstruction range to generate a pre-reconstructed image; The preset reconstruction center and the preset reconstruction range are the pre-set reconstruction center and reconstruction range, where: The reconstruction center refers to the center point of the area of interest during the scanning and reconstruction of the CBCT image. The reconstruction center is usually determined according to clinical needs and scanning purposes, and it determines the focal area of the reconstructed image. For example, in dental CBCT, the reconstruction center may be set inside the patient's oral cavity for specific teeth or jaw areas that need to be examined in detail.

[0027] The reconstruction range refers to the size of the specific area that is scanned and reconstructed into a three-dimensional image during CBCT scanning. The reconstruction range is usually determined by the field of view (FOV), and it determines the volume that can be covered by CBCT scanning. The selection of the reconstruction range has an important impact on the resolution and quality of the image. A larger reconstruction range can provide more anatomical information but may result in a lower image resolution; while a smaller reconstruction range can provide a higher-resolution image but covers a smaller anatomical area.

[0028] In practical applications, the reconstruction range is defined with the reconstruction center as the center point. During the CBCT scanning and reconstruction process, the reconstruction center is the focus of the entire imaging process, and the reconstruction range refers to a certain volume area around this center point. In the present invention, the reconstruction range refers to the diameter size of the cylinder, and the cylinder height is not discussed here.

[0029] The pre-reconstruction operation refers to performing pre-reconstruction on multiple to-be-reconstructed images. It can use the projection data (i.e., multiple to-be-reconstructed images) to reconstruct the three-dimensional image of the object cross-section, that is, based on the X-ray projection data at multiple angles, through a specific algorithm and calculation process, to restore the three-dimensional image of the internal structure of the object. The pre-reconstructed image refers to the three-dimensional image obtained by processing the to-be-reconstructed image through the pre-reconstruction operation.

[0030] Step S3, set a target reconstruction center and a target reconstruction range on the pre-reconstructed image respectively; The target reconstruction center is the reconstruction center selected by the user on the pre-reconstructed image, and the target reconstruction range is the reconstruction range selected by the user on the pre-reconstructed image. In practical applications, after the pre-reconstructed image is generated, the pre-reconstructed image will be presented to the user in a visual manner, such as being displayed on a display screen, facilitating the user to intuitively set the target reconstruction center and the target reconstruction range.

[0031] In practical applications, there can be multiple ways to select the target reconstruction center. For example: (1) Select a reconstruction area on the pre-reconstructed image; determine the center of the reconstruction area and use the center of the area as the target reconstruction center. In practical applications, the user can use a mouse or a stylus to draw a reconstruction area on the pre-reconstructed image displayed on the display screen, use the center of the area as the target reconstruction center, and use the coordinates of the center of the area as the coordinates of the target reconstruction center. (2) Select a target position on the pre-reconstructed image and use the target position as the target reconstruction center. In practical applications, the user can also directly use the mouse to select a target position (i.e., a target point) on the pre-reconstructed image displayed on the display screen as the reconstruction center, and use the coordinates of the target position as the coordinates of the target reconstruction center. In the present invention, the coordinates of the target reconstruction center can be denoted as (centerX, centerY). In addition, based on determining the target reconstruction center, the origin of the coordinates is changed to the rotation center in the present invention.

[0032] In practical applications, there can be multiple ways to select the target reconstruction center. For example: (1) Select a circular reconstruction area on the pre-reconstructed image and use the diameter of the circular reconstruction area as the target reconstruction range, that is, use the diameter of the area as the cylinder diameter corresponding to the target reconstruction range. In practical applications, the user can use mouse operations to draw a circular reconstruction area on the pre-reconstructed image displayed on the display screen and use the diameter of the area as the target reconstruction range. (2) Pop up an edit box for the user to edit the target reconstruction range. In practical applications, a corresponding edit box can be popped up on the display screen, and the user can determine the target reconstruction range by entering the corresponding range value in the edit box. (3) Obtain a preset range list, where the range list includes multiple preset ranges; select a target range from the multiple preset ranges in the range list and use the target range as the target reconstruction range. In practical applications, for different lesion positions, corresponding range lists can be preset in advance, and then the required target reconstruction range can be selected based on the range list. The number of preset ranges included in the range list, the actual values, etc. can all be selected according to the actual situation and are not limited here. In the present invention, the reconstruction range is denoted as (fov).

[0033] As described above, a target reconstruction center and a target reconstruction range are respectively set on the pre-reconstructed image, so as to realize user-defined target reconstruction range and target reconstruction center. For example, the target reconstruction center can be specified at the location of the lesion, and the reconstruction range can be reduced to improve the spatial resolution, so as to facilitate subsequent observation and diagnosis of diseases at different positions and of different sizes.

[0034] Step S4: Based on the target reconstruction center and the target reconstruction range, perform an image reconstruction operation on the multiple images to be reconstructed to generate a three-dimensional image.

[0035] As described above, the user has defined the target reconstruction range and the target reconstruction center. By reconstructing (i.e., performing an image reconstruction operation) the multiple images to be reconstructed based on the user-defined target reconstruction range and target reconstruction center, three-dimensional images corresponding to diseases at different positions and of different sizes can be obtained, which is convenient for observation and diagnosis. In the present invention, the above step S4 specifically includes the following steps: Perform a preprocessing operation on each of the multiple images to be reconstructed to obtain corresponding multiple preprocessed images; Based on a conversion formula, convert the multiple preprocessed images into corresponding multiple attenuation maps respectively; Perform a filtering operation on each of the multiple attenuation maps to obtain corresponding multiple filtered images; Based on a reconstruction formula, perform back-projection reconstruction on the multiple filtered images to obtain the three-dimensional image.

[0036] In practical applications, performing an image reconstruction operation on the multiple images to be reconstructed based on the target reconstruction center and the target reconstruction range includes the following steps: (1) First, a preprocessing operation needs to be performed on each image to be reconstructed. The preprocessing operation for each image to be reconstructed is as follows: perform bad pixel retrieval and correction on the image to be reconstructed; determine whether there is an abnormal edge in the corrected image to be reconstructed. If so, compensate for the abnormal edge.

[0037] Specifically, the steps for performing bad pixel retrieval and correction on the image to be reconstructed are as follows: perform pixel inspection on the image to be reconstructed to determine the difference between the pixel value of a target pixel point in the image to be reconstructed and the average value of neighboring pixels; determine whether the difference exceeds a preset threshold. If it exceeds the preset threshold, determine that the target pixel point is a bad pixel, and correct the bad pixel based on the average value of neighboring pixels.

[0038] The bad pixel retrieval involves performing pixel inspection on each pixel point of the image to be reconstructed one by one to identify pixel points that are significantly different from surrounding pixels, which can be achieved by comparing the pixel value with the values of its neighboring pixels.

[0039] The target pixel point is a certain pixel point in the image to be reconstructed. The neighborhood pixel mean refers to: for a target pixel point in the image to be reconstructed, take the neighborhood pixels within a certain range around it (these neighborhood pixels usually form a rectangular or circular area), and calculate the average value of the gray values of these neighborhood pixels. This average value is the neighborhood pixel mean of the target pixel point.

[0040] In practical applications, the present invention can first obtain the difference between the pixel value of the target pixel point in the image to be reconstructed and the neighborhood pixel mean. If the difference exceeds a preset threshold, it can be determined that the target pixel point is a bad pixel. At this time, the bad pixel can be corrected based on the neighborhood pixel mean. For example, the pixel value of the target pixel point can be replaced with the neighborhood pixel mean to achieve the correction of the bad pixel. If the difference does not exceed the preset threshold, there is no need to correct the target pixel point. It should be noted that the above preset threshold can be selected according to the actual situation.

[0041] After performing bad pixel retrieval and correction on the image to be reconstructed, image abnormal edge compensation can be performed on the corresponding image to be reconstructed. Specifically: determine whether there is an abnormal edge in the corrected image to be reconstructed. If there is, compensate for the abnormal edge. If not, there is no need to compensate for the corrected image to be reconstructed.

[0042] The abnormal edge usually refers to abnormal or irregular phenomena that appear in the edge part of the image to be reconstructed. These abnormalities may be manifested as edge blurring, breakage, dislocation, distortion, or discontinuity with other parts. The appearance of abnormal edges is usually caused by non-photosensitive pixel columns / rows of the detector itself on the four sides of the image to be reconstructed. These abnormal edges may cause problems such as edge blurring, distortion, or pseudo-color in the image to be reconstructed. Therefore, it is necessary to compensate for the abnormal edge image, and noise reduction can be performed on the corresponding image to be reconstructed if necessary. In practical applications, for the case of abnormal edges, the nearest normal pixel value can be used for filling. For example, when it is detected that a certain pixel in the corrected image to be reconstructed is an abnormal pixel, its nearest normal pixel can be searched for, and the pixel value of one of the normal pixels can be selected for filling. In addition, the above noise reduction can be performed using common noise reduction methods such as mean filtering, Gaussian filtering, and frequency domain low-pass filters.

[0043] As can be seen from the above description, by performing the above preprocessing operations on each image to be reconstructed, the quality of the image to be reconstructed can be effectively improved, facilitating subsequent image reconstruction and visual presentation. The preprocessed image can be represented as Ii, where i = 1, 2, 3... N.

[0044] (2)Based on the conversion formula, convert the multiple preprocessed images into corresponding multiple attenuation maps respectively. The conversion formula is as follows:

[0045] Among them, Ai represents the i-th attenuation map; Log(I0 / Ii) represents the natural logarithm of the ratio of I0 to Ii. Log represents the natural logarithm, that is, the logarithm with the constant e as the base. I0 is the image without attenuation, which is obtained by taking a blank shot under the same exposure conditions. Ii represents the i-th preprocessed image, and i represents the image serial number, where i = 1, 2, 3... N. The above I0 / Ii represents the one-to-one division of the gray levels of each pixel in I0 and Ii, and its physical meaning is to convert the original brightness of the image into the attenuation value of the substance.

[0046] In practical applications, after the preprocessing operation, each preprocessed image can be converted into a corresponding attenuation map. Among them, converting the preprocessed image into an attenuation map means converting the gray level values of the original image to be reconstructed into the attenuation values after the rays pass through the human body or object.

[0047] (3) Perform filtering operations on multiple said attenuation maps respectively to obtain multiple filtered images; After converting the preprocessed image into the corresponding attenuation map, each attenuation map can be filtered. Specifically: windowing is performed on the basis of an ideal ramp filter, and then frequency domain filtering is performed on each row of the attenuation map Ai. The frequency domain representation of the ideal ramp filter is (as Figure 2 shown). Among them, the form of the above windowing can be defined according to requirements. In practical applications, the windowing as Figure 3 shown can better suppress the high frequency, reduce noise, and at the same time enhance the intensity of the intermediate frequency, which helps to enhance the intermediate frequency information. In the present invention, the filtered image can be represented as Fi, where i = 1, 2, 3... N (4) Based on the reconstruction formula, perform back-projection reconstruction on multiple said filtered images to obtain the three-dimensional image and the corresponding three-dimensional slice images.

[0048] The reconstruction formula is:

[0049] Among them, represents the pixel value at the spatial coordinates (x, y, z) of the three-dimensional image, where (x, y, z) are the coordinates of the three-dimensional image in the three-dimensional space; represents the sum of the attenuation values (after filtering) of all rays passing through the coordinates (x, y, z), where represents the pixel value of the filtered image Fi at the coordinates Fi represents the i-th filtered image, and i represents the image serial number, where i = 1, 2, 3... N, Denote the coordinates where the ray passing through (x, y, z) hits the image to be reconstructed. In addition, it should be noted that x ∈ [0, fov], y ∈ [0, fov], fov is the target reconstruction range, and the coordinates of the above target reconstruction center are (centerX, centerY). Therefore:

[0050]

[0051] Among them, SID refers to the distance from the X-ray focus to the detector plane. Based on the above three-dimensional image V(x, y, z), the corresponding three-dimensional slice image can be obtained. For example, V(z) in the three-dimensional image V(x, y, z) represents the three-dimensional slice image at the z position.

[0052] In practical applications, after obtaining the corresponding three-dimensional image and three-dimensional slice image, the three-dimensional image and three-dimensional slice image can be displayed for the user to observe. In addition, the user can also store the three-dimensional image and three-dimensional slice image in a preset storage module for subsequent use.

[0053] In summary, the CBCT reconstruction method provided by the present invention: can align the reconstruction center to the lesion position, or increase or decrease the target reconstruction range. When keeping the same matrix, the sampling rate can be changed to obtain a tomographic image with an appropriate spatial resolution, which is convenient for observing and diagnosing diseases at different positions and sizes. For example, when imaging a small lesion, the reconstruction center is specified to the position where the lesion is located, the target reconstruction range is reduced, and the same matrix is maintained to increase the sampling rate and thus improve the spatial resolution, so that finer lesions can be distinguished; for another example, when imaging parts such as both lower limbs, a high resolution is not required but a larger target reconstruction range is needed to ensure that both feet can be observed simultaneously. Then, the target reconstruction range can be increased, and the reconstruction center is specified between the two feet to facilitate the overall observation of lesions such as deformities of both lower limbs.

[0054] The above are only the embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent transformation made by using the specification and drawings of the present invention, or directly or indirectly applied in the relevant technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A CBCT reconstruction method, characterized in that: The following steps are involved: S1: Read multiple images to be reconstructed at different angles; S2: performing a pre-reconstruction operation on the plurality of images to be reconstructed based on a preset reconstruction center and a preset reconstruction range to generate a pre-reconstructed image; S3: setting a target reconstruction center and a target reconstruction range on the pre-reconstructed image respectively; S4: Based on the target reconstruction center and the target reconstruction range, perform an image reconstruction operation on the plurality of images to be reconstructed to generate a three-dimensional image.

2. The CBCT reconstruction method according to claim 1, characterized in that: In step S3, setting a target reconstruction center on the pre-reconstructed image includes: selecting a reconstruction area on the pre-reconstructed image; A regional center of the reconstruction region is determined, and the regional center is used as a target reconstruction center.

3. The CBCT reconstruction method according to claim 1, characterized in that: In step S3, setting a target reconstruction center on the pre-reconstructed image includes: A target position is selected on the pre-reconstructed image, and the target position is used as the target reconstruction center.

4. The CBCT reconstruction method according to claim 1, characterized in that: In step S3, setting a target reconstruction range on the pre-reconstructed image includes: A circular reconstruction area is selected on the pre-reconstructed image, and the area diameter of the circular reconstruction area is used as a target reconstruction range.

5. The CBCT reconstruction method according to claim 1, characterized in that: In step S3, setting a target reconstruction range on the pre-reconstructed image includes: Obtain a preset range list, wherein the range list includes multiple preset ranges; A target range is selected from a plurality of preset ranges in the range list, and the target range is used as a target reconstruction range.

6. The CBCT reconstruction method according to claim 1, characterized in that: In step S4, based on the target reconstruction center and the target reconstruction range, an image reconstruction operation is performed on the plurality of images to be reconstructed to generate a three-dimensional image, including: Performing preprocessing operations on the plurality of images to be reconstructed respectively to obtain a corresponding plurality of preprocessed images; Based on a conversion formula, converting the plurality of preprocessed images into corresponding plurality of attenuation maps respectively; Performing filtering operations on the plurality of attenuation maps respectively to obtain a plurality of corresponding filtered images; Based on the reconstruction formula, back-projection reconstruction is performed on the plurality of filtered images to obtain the three-dimensional image.

7. The CBCT reconstruction method according to claim 6, characterized in that: In step S4, the pre-processing operation includes the following steps: Performing bad pixel retrieval and correction on the image to be reconstructed; Determine whether the corrected image to be reconstructed has an abnormal edge, and if so, compensate for the abnormal edge.

8. The CBCT reconstruction method according to claim 7, characterized in that: In step S4, bad pixel retrieval and correction are performed on the image to be reconstructed, including: Performing pixel inspection on the image to be reconstructed to determine the difference between the pixel value of the target pixel in the image to be reconstructed and the mean value of the neighboring pixels; It is determined whether the difference exceeds a preset threshold. If it exceeds the preset threshold, the target pixel is determined to be a bad pixel, and the bad pixel is corrected based on the neighborhood pixel mean.

9. The CBCT reconstruction method according to claim 6, characterized in that: In step S4, the conversion formula is: Among them, Ai represents the i-th attenuation image; Log(I0 / Ii) represents the natural logarithm of the ratio of I0 to Ii, Log represents the natural logarithm, I0 is the image without attenuation, Ii represents the i-th preprocessed image, i represents the image sequence number, i=1, 2, 3...N.

10. The CBCT reconstruction method according to claim 1, characterized in that: In step S4, the reconstruction formula is: in, Represents the pixel value of the three-dimensional image at the spatial coordinate (x, y, z), where (x, y, z) is the coordinate of the three-dimensional image in the three-dimensional space; It represents the sum of the attenuation values ​​of all rays passing through the coordinates (x, y, z), where: Indicates the filtered image Fi at coordinates , Fi represents the i-th filtered image, i represents the image number, i=1, 2, 3...N, Indicates the coordinates of the ray passing through (x, y, z) hitting the image to be reconstructed.

Images