DBT reconstruction method based on registration of multi-angle projection

Abstract

Disclosed in the present invention is a DBT reconstruction method based on registration of multi-angle projection. The method comprises the following steps: acquiring multi-angle projection data; preprocessing the projection data on the basis of Beer-Lambert law, so as to obtain preprocessed projection data; performing block coordinate translation processing on the preprocessed projection data on the basis of a displacement amount registration method for isocenter circular motion, so as to obtain translated projection data; performing image interpolation processing on a vacant pixel value in the translated projection data, so as to obtain interpolated projection data; and on the basis of the interpolated projection data and a DBT reconstruction algorithm, obtaining reconstructed DBT layers and outputting same. By means of acquiring multi-angle projection data and on the basis of a displacement amount registration method for isocenter circular motion, center projection is standardized, and different magnification requirements of projection at different angles in respect of reconstructed layers are met, so that the contrast is higher, the texture of focused reconstructed layers is clearer, the boundaries thereof are clearer, and it is less prone to generate artifacts.

Description

A DBT reconstruction method based on multi-angle projection registration Technical Field

[0001] The invention relates to a DBT reconstruction method of multi-angle projection registration, belonging to the technical field of medical image processing. Background Art

[0002] Breast cancer is a malignant tumor that develops in the mammary epithelium and poses a serious threat to women's life and health. Early detection, diagnosis, and treatment of breast cancer can extend the lives of 90% of patients, demonstrating the importance of early screening for breast cancer. Digital Breast Tomosynthesis (DBT) is a technology that uses a limited number of projections within a limited angular range to perform three-dimensional breast imaging. In DBT, as an X-ray source moves along a predetermined trajectory (typically an arc of 60° or less), multiple projections are acquired to reconstruct cross-sections parallel to the breast support plane. The entire data acquisition and scanning process is time-efficient, and the captured limited-angle projection data contains comprehensive anatomical information about the breast tissue at various angles. Compared to the widely used computed tomography (CT) imaging method, which acquires data from all angles, DBT captures data within a narrow angular range, collecting only 10 to 25 projections within a small angular range, compared to thousands of projections in CT. Therefore, the total radiation exposure associated with DBT is significantly less than that of conventional CT.

[0003] However, the current literature on DBT imaging all uses the X-ray tube's parallel operation mode relative to the stationary detection surface to derive the displacement formula, which has low contrast, poor focusing effect, blurred boundaries, and easy to produce artifacts. Summary of the Invention

[0004] The purpose of the present invention is to overcome the shortcomings of the existing technology and provide a DBT reconstruction method with multi-angle projection registration. By acquiring projection data from multiple angles and based on the displacement registration method of isocentric arc motion, the central projection standardization is achieved, and the different magnification requirements of the reconstruction layer for different angle projections are met, resulting in higher contrast, clearer texture and boundary of the focused reconstruction layer, and less prone to artifacts.

[0005] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0006] The present invention discloses a DBT reconstruction method with multi-angle projection registration, comprising the following steps:

[0007] Acquire projection data from multiple angles;

[0008] Preprocessing the projection data based on Beer's theorem to obtain preprocessed projection data;

[0009] Based on the displacement registration method of isocenter arc motion, the pre-processed projection data is subjected to block coordinate translation processing to obtain translated projection data;

[0010] performing image interpolation processing on the pixel vacancy values ​​in the translated projection data to obtain interpolated projection data;

[0011] According to the interpolated projection data, based on a DBT reconstruction algorithm, a DBT reconstruction layer is obtained and output.

[0012] Furthermore, the obtaining of multi-angle projection data includes:

[0013] Based on the X-ray tube, the center arc motion projection is performed, and the detector is used to shoot the X-ray projection at equal angle intervals to obtain multi-angle projection data.

[0014] Furthermore, the expression of the pre-processed projection data is as follows: θ =ln(I θ,0 +1);

[0015] Where θ represents the projection angle; I θ I represents the projection data of the projection angle θ after preprocessing; θ,0 represents the projection data of the projection angle θ without preprocessing; ln() represents the logarithmic operation with the constant e as the base.

[0016] Furthermore, the displacement registration method of the isocenter arc motion includes:

[0017] Obtaining preprocessed projection data;

[0018] Based on the preset isocenter arc motion model of different reconstruction layer depths, the displacement of the preprocessed projection data at different angles is calculated to obtain the displacement of the projection data along the x-axis determined by the angle and the reconstruction layer depth;

[0019] According to the displacement of the projection data along the x-axis, block coordinate translation processing is performed on the pre-processed projection data to obtain translated projection data.

[0020] Furthermore, the expression for the displacement of the projection data along the x-axis is as follows:

[0021] Where θ represents the projection angle; x θ Indicates the x-axis coordinate to be translated when the projection angle is θ; Z0 indicates the depth of the reconstruction layer; shift(x θ,z0) represents the displacement of the projection data along the x-axis when the reconstruction layer depth is Z0 and the projection angle is θ; d represents the radius of the isocenter arc.

[0022] Furthermore, the translated projection data is expressed as: θ ′=(x θ ′,y θ ,z0);

[0023] Among them, I θ ′ represents the projection data after translation; x θ ′ represents the x-axis coordinate after translation when the projection angle is θ, θ ′=x θ -shift(x θ ,z0),x θ Indicates the x-axis coordinate to be translated when the projection angle is θ; y θ It represents the y-axis coordinate when the projection angle is θ; Z0 represents the reconstruction layer depth.

[0024] Furthermore, the interpolated projection data is obtained, including:

[0025] Get the projection data after translation;

[0026] performing pixel vacancy search processing on the translated projection data to obtain pixel vacancy values;

[0027] According to the pixel vacancy value, a one-dimensional linear interpolation method is used to interpolate pixels on both sides of the pixel vacancy value to obtain interpolated projection data.

[0028] Furthermore, the expression of the one-dimensional linear interpolation method is as follows:

[0029] Among them, y θ- Represents the pixel vacancy value of the y-axis component in the translated projection data; y1 represents the pixel vacancy value y θ- The adjacent upper pixel value; y2 represents the pixel vacancy value y θ- The adjacent next pixel value.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The present invention obtains projection data from multiple angles and realizes the standardization of central projection based on the displacement registration method of isocentric circular arc motion, meets the different magnification requirements of reconstruction layers for projections at different angles, improves the registration accuracy of multi-angle projections, makes the contrast higher, the texture of the focused reconstruction layer clearer, the boundary clearer, and is less likely to produce artifacts. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] FIG1 is a flow chart of a DBT reconstruction method with multi-angle projection registration provided by this embodiment;

[0033] FIG2 is a schematic diagram of an isocentric arc motion mode of an optimized X-ray tube provided in this embodiment;

[0034] FIG3 is a schematic diagram of a conventional X-ray tube in parallel operation mode;

[0035] FIG4 is a projection data of mammary gland images collected from multiple angles;

[0036] FIG5 shows a focused reconstruction layer and a difference image between two focused reconstruction layers obtained by using the prior art and the method of this embodiment respectively;

[0037] FIG6 is a diagram showing the effect of magnifying the marked points in FIG5 by 800 times;

[0038] FIG. 7 shows the focused reconstruction layer and the corresponding Fourier spectrum obtained by using the prior art and the method of this embodiment, respectively. DETAILED DESCRIPTION

[0039] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention.

[0040] This embodiment provides a DBT reconstruction method with multi-angle projection registration, including the following steps:

[0041] Acquire projection data from multiple angles;

[0042] Preprocessing the projection data based on Beer's theorem to obtain preprocessed projection data;

[0043] Based on the displacement registration method of isocenter arc motion, the pre-processed projection data is subjected to block coordinate translation processing to obtain the translated projection data;

[0044] Performing image interpolation processing on the pixel vacancy values ​​in the translated projection data to obtain interpolated projection data;

[0045] According to the interpolated projection data, based on the DBT reconstruction algorithm, the DBT reconstruction layer is obtained and output.

[0046] The technical concept of the present invention is: by acquiring projection data from multiple angles, the displacement registration method based on the isocentric circular arc motion is used to achieve standardization of the central projection, meet the different magnification requirements of the reconstruction layer for projections at different angles, improve the registration accuracy of the multi-angle projection, make the contrast higher, the texture of the focused reconstruction layer clearer, the boundary clearer, and less likely to produce artifacts.

[0047] As shown in Figure 1, the specific steps are as follows:

[0048] Step 1: Obtain projection data from multiple angles.

[0049] Based on the isocentric circular motion projection of the X-ray tube, the X-ray projection is taken at equal angle intervals using an X-ray flat panel detector to obtain multi-angle projection data.

[0050] This embodiment collects projection data from 15 angles, with a total projection angle range of 15 degrees. Compared with traditional CT acquisition projection, its projection angle range is smaller, the number of acquisitions is smaller, and it is less harmful to people.

[0051] As shown in Figure 2, the center line of the X-ray tube isocenter arc is defined as the z-axis, and the intersection of the arc plane and the detector plane is defined as the x-axis. The intersection of the z-axis and the detector plane is the center O. Assuming that 0° is projected with the Z-axis as the center, the collected projection angle range is (-7.5°, 7.5°), and projection data at 15 angles are collected, namely -7×(7.5° / 7), -6×(7.5° / 7), -5×(7.5° / 7), -4×(7.5° / 7), -3×(7.5° / 7), -2×(7.5° / 7), -1×(7.5° / 7), 0°, 1×(7.5° / 7), 2×(7.5° / 7), 3×(7.5° / 7), 4×(7.5° / 7), 5×(7.5° / 7), 6×(7.5° / 7), and 7×(7.5° / 7).

[0052] Step 2: Preprocess the projection data based on Beer's theorem, that is, perform a logarithmic transformation operation to obtain preprocessed projection data.

[0053] The expression of the preprocessed projection data is as follows: θ =ln(I θ，0 +1);

[0054] Where θ represents the projection angle; I θ I represents the projection data of the projection angle θ after preprocessing; θ,0 represents the projection data of the projection angle θ without preprocessing; ln() represents the logarithmic operation with the constant e as the base.

[0055] It should be noted that the projection data I without preprocessing of the projection angle θ θ,0 Including the x-axis coordinate and y-axis coordinate projected on the unprocessed detector plane, that is, I θ，0 =(x θ，0 ,y θ，0 ).

[0056] The projection data I of the projection angle θ after preprocessing θ, including the x-axis coordinate and y-axis coordinate projected on the detector plane after preprocessing, I θ =(x θ ,y θ ).

[0057] Indicates step 3, based on the displacement registration method of isocenter arc motion, block coordinate translation processing is performed on the pre-processed multi-angle projection data to obtain translated projection data.

[0058] Step 3.1: Obtain preprocessed projection data.

[0059] Step 3.2: Based on the preset isocenter arc motion model for different reconstruction slice depths, calculate the displacement of the preprocessed projection data at different angles to obtain the displacement of the projection data along the x-axis, which is determined by the angle and reconstruction slice depth. Specifically, a pre-constructed isocenter arc motion model is shown in Figure 2. The centerline of the X-ray tube isocenter arc is defined as the z-axis, the intersection of the arc plane and the detector plane is defined as the x-axis, the intersection of the z-axis and the detector plane is defined as the center O, and the line perpendicular to the x-axis in the detector plane is defined as the y-axis.

[0060] The projection data of the pre-processed projection angle θ is expressed as I θ , including the x-axis coordinate and y-axis coordinate of the projection of the detector plane. Next, step 3.3 of this embodiment will calculate the x-axis coordinate x θ Perform registration.

[0061] In this step, based on the distance from the detector plane, that is, the plane formed by the x-axis and the y-axis is a non-negative integer multiple of 1 mm, 62 reconstruction layers of different depths are set above it and are numbered as: 0, 1, 2, ..., 61, that is, the maximum reconstruction layer thickness is set to 61 mm.

[0062] Taking the reconstruction layer with depth z0 of 61 as an example, let point P(x θ ,z0) is any point on the reconstruction layer. When the X-ray tube is at the center projection angle (i.e. 0°), the position of the X-ray tube is recorded as point R(0,d). Then the ray RP intersects the x-axis at point N(x N ,0), the projection of point P on the x-axis is denoted as point C(x θ ,0), the projection of point P on the z-axis is recorded as point F(θ,z0);

[0063] When the projection angle of the X-ray tube is θ, the position of the X-ray tube is recorded as point S (-dsinθ, dcosθ), then the ray SP intersects the x-axis at point M (x M ,0), the projection of point S on the x-axis is denoted as point H(-dsinθ,0), and the projection of point S on the z-axis is denoted as point D(0,dcosθ).

[0064] Where d is the rotation radius of the isocentric arc motion, z0 is the depth of the reconstructed focus layer, that is, the distance between the reconstructed focus layer and the X-ray detector plane, θ is the angle of the X-ray tube at the current projection relative to the center projection, and the center projection is 0°. θ The x-axis coordinate of a point P on the reconstruction layer to be translated and registered.

[0065] In the constructed triangle, ΔRPF∽ΔRNO, ΔMPC∽ΔMSH, and according to the similarity relationship, we can obtain:

[0066] Substituting the corresponding coordinates into the above two equations, we can obtain:

[0067] In Figure 2, NM is the displacement along the x-axis on the detection surface relative to the center projection (projection angle is 0°) when the projection angle is θ. Let shift(x θ ,z0)=NM=x M -x N , then the expression of the displacement used for registration is as follows:

[0068] Where θ represents the projection angle; x θ Indicates the x-axis coordinate to be translated when the projection angle is θ; z0 indicates the depth of the reconstruction layer; shift(x θ ,z0) represents the displacement of the projection data along the x-axis when the reconstruction layer depth is Z0 and the projection angle is θ; d represents the radius of the isocenter arc.

[0069] The first term in the displacement formula is, It is a linear function related to the horizontal position and projection angle of the projected object. Its physical meaning is the linear transformation of the projection displacement on the x-coordinate to achieve the standardization of the central projection.

[0070] The second term of the displacement formula is, It is only related to the projection angle θ and represents the relative displacement on the x-axis required for the normalized projections at different angles θ to be calibrated to the center projection.

[0071] It is worth noting that when the detection surface also swings within a small angle range of ±φ simultaneously with the X-ray tube, each displacement obtained by the above formula needs to be multiplied by cos(±(φ*n) / 2m), where n=0,1,2,...m; 2m+1 is the number of projection angles. Usually, the φ value is very small and can be ignored, so the cosine value cos(±(φ*n) / 2m) is directly taken as 1.

[0072] Finally, the displacement of any point P along the x-axis on each reconstruction layer is obtained.

[0073] Step 3.3: According to the displacement of the projection data along the x-axis, the pre-processed projection data is subjected to block coordinate translation processing to obtain translated projection data.

[0074] It should be noted that the preprocessed projection data in step 2 includes projection data from the detector plane at 15 projection angles, with the z-axis coordinate at 0. Steps 3.1 and 3.2 obtain the x-axis displacement for different reconstruction layers and projection angles. Block-wise coordinate translation is performed on the preprocessed projection data to obtain translated projection data, which is equivalent to stacking multiple reconstruction layers at different depths on the original projection data from the detector plane.

[0075] Therefore, the projection data after translation is expressed as I θ ′,I θ ′=(x θ ′,y θ ,z0), where x θ ′ represents the x-axis coordinate after translation when the projection angle is θ, θ ′=x θ -shift(x θ ,z0),x θ Indicates the x-axis coordinate to be translated when the projection angle is θ; y θ It represents the y-axis coordinate when the projection angle is θ; Z0 represents the depth of the reconstruction layer.

[0076] Step 4: Perform image interpolation processing on the missing pixel values ​​in the translated projection data to obtain interpolated projection data.

[0077] Get the projection data after translation;

[0078] Perform pixel vacancy search processing on the translated projection data to obtain pixel vacancy values;

[0079] According to the pixel vacancy value, a one-dimensional linear interpolation method is used to interpolate the adjacent pixels on both sides of the pixel vacancy value to obtain the interpolated projection data.

[0080] Specifically, after performing block coordinate translation processing on the pre-processed projection data, that is, performing linear transformation on the x-axis component of the horizontal position, pixel vacancies on the y-axis component may easily occur.

[0081] In this step, first, the projected data I θ ′=(x θ ′,y θ , z0), find the pixel vacancy value y of the y-axis component θ- , take the pixel vacancy value y θ- The upper and lower adjacent pixels y1 and y2 are averaged and interpolated. The interpolation expression is:

[0082] Among them, y θ- Represents the pixel vacancy value of the y-axis component in the translated projection data; y1 represents the pixel vacancy value y θ- The adjacent upper pixel value; y2 represents the pixel vacancy value y θ- The adjacent next pixel value.

[0083] The image interpolation processing is shown in Figure 4. After the horizontal position linear transformation of the projection data, pixel vacancies will appear. The change of the projection angle of the X-ray tube will cause the position of the projection image on the X-ray detector to shift accordingly. Using the displacement registration formula proposed in this method to standardize, shift and register the projection image, more accurate focus reconstruction layer information can be extracted.

[0084] Step 5: Output the DBT reconstruction layer based on the interpolated projection data.

[0085] In this step, the ShiftAndAdd (SAA) method is used for DBT reconstruction. It mainly superimposes and averages the registered multiple projections, which has the advantages of being fast and simple. The advantages of the present invention will be demonstrated and analyzed based on this reconstruction method.

[0086] In addition, DBT reconstruction can also be performed using the filtered back projection method (FBP), the maximum likelihood expectation maximization method (MLEM), and the blind source separation based separating the focusing plane (BSFP) method.

[0087] In the existing technical literature, most of the displacement formulas are derived by assuming that the X-ray tube runs parallel to the detection surface. The derivation process is as follows:

[0088] Establish the coordinate system shown in Figure 3. When the X-ray tube is at the first projection angle, its projection on the x-axis is at a1. At this time, the X-ray passes through z and z on the z-axis. f , respectively fall to x1 and b1 on the x-axis, corresponding to the X-ray tube being located at the center projection angle and the rays emitted falling to x0 and 0 on the x-axis. Define the projection magnification M f =D / (Dz f ), D represents the height of the X-ray tube. The depth at this time is z f When the projection of the first projection angle is relative to the center projection, the displacement on the x-axis is x1(zf )for:

[0089] Where a1 represents the x-axis projection component of the X-ray tube; M f Indicates the reconstruction layer depth is Z f The above formula shows that the displacement on the x-axis is related to the reconstruction depth and the position of the X-ray tube.

[0090] For the reconstruction depth z, the x displacement is: x1(z)=a1(1-M z );

[0091] The reconstruction depth is z, z f The relative displacement between the projections on the x-axis is: x1′(z)=x1(z)-x1(z f )=a1(M f -M z );

[0092] Similarly, when the X-ray tube is projected on the x-axis as a k When the reconstruction depth is z, z f The relative displacement between the projections on the x-axis is: shift k =x′ k (z) = a k (M f -M z ).

[0093] In order to compare and verify that the displacement formula of the present invention can be used to align multiple projections more accurately, the focus layer reconstruction effects after alignment with the displacement formula of the prior art and the displacement formula proposed in the present invention are compared, and the focus layer reconstruction based on SAA of 15 projections collected by the DBT system of HOLOGIC Company in Figure 4 is taken as an example.

[0094] As shown in Figure 5, reconstructed focus layers at the same depth were selected for visual comparison and analysis. Figure 5(a) shows the reconstructed focus layer corrected by the original method shown in Figure 3, followed by the correction of the present invention. Figure 5(b) shows the reconstructed focus layer image after the correction of the present invention minus the reconstructed image after the original method. Figure 5(c) is the difference between Figures 5(a) and 5(b). It can be seen that the difference between the calibration points of the two is very small near the center, but there is a significant difference at the edges and sides of the gland in the reconstructed image, and the difference increases the further away from the gland's central axis.

[0095] The four reconstructed calibration points marked in Figures 5(a) and (b) are respectively intercepted with small dot images at the same position, and the reconstructed small dots are proportionally magnified 800 times for observation, as shown in Figure 6. By comparing Figures 6(a) and (b), Figure 6(a) has a more obvious artifact on the side close to the middle axis; although the result of the displacement formula processing proposed by this method in Figure 6(b) also has a small amount of artifacts, it greatly reduces the artifacts of the reconstructed points compared with Figure 6(a); It can be seen from Figures 6(g) and 6(h) that when the background is bright, the contrast of the reconstructed points obtained by using the original displacement formula in Figure 6(g) is greatly reduced, and its boundary points become blurred; Figure 6(h) is the corrected image, which is not only better focused, but also has a significantly increased contrast and a clearer boundary. From the comparison between Figure 6(a) and Figure 6(b), Figure 6(c) and Figure 6(d), Figure 6(e) and Figure 6(f), and Figure 6(g) and Figure 6(h), it can be verified that the displacement formula of the present invention is significantly higher than the original displacement formula in terms of registration accuracy, and reduces the blur and artifacts of the focus point.

[0096] Figure 7 (a) and Figure 7 (b) are comparisons of the effects of the reconstructed layers after registration using the original displacement method and the projection after registration using the method of the present invention, respectively. The upper part is the reconstructed layer, and the lower part is its corresponding Fourier Transform (FT) spectrum. From the comparison of the reconstructed focus layers in Figure 7 (a) and (b), after the angle projection registration is corrected using the displacement registration formula of the present invention, the small metal focused in the reconstructed layer is clearer and the image contrast is higher. Analyzing from the spectrum of the reconstructed layer, the frequency of Figure 7 (a) is cleaner in the horizontal direction, but there is more high-frequency noise in the vertical direction, which indicates that high-frequency noise is generated during the longitudinal calibration process. In contrast, the spectrum corresponding to the reconstructed layer in Figure 7 (b) is cleaner in both the horizontal and vertical directions, indicating that after the displacement correction is performed using the method of the present invention, the projection registration is more accurate, and more noise is avoided during the correction process, which can make the texture of the focus layer clearer, the contrast is higher, and the blur is suppressed.

[0097] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0098] The present application is described with reference to the flow chart and / or block diagram of the method, device (system), and computer program product according to the embodiment of the present application. It should be understood that each flow process and / or box in the flow chart and / or block diagram and the combination of the flow process and / or box in the flow chart and / or block diagram can be realized by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processing machine or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for realizing the function specified in one flow chart flow or multiple flows and / or one box or multiple boxes of the block diagram.

[0099] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0101] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A DBT reconstruction method with multi-angle projection registration, characterized in that: The steps include: Acquire projection data from multiple angles; Preprocessing the projection data based on Beer's theorem to obtain preprocessed projection data; Based on the displacement registration method of isocenter arc motion, block coordinate translation processing is performed on the preprocessed projection data to obtain translated projection data; Performing image interpolation processing on the pixel vacancy values ​​in the translated projection data to obtain interpolated projection data; According to the interpolated projection data, based on a digital breast tomosynthesis (DBT) reconstruction algorithm, a digital breast tomosynthesis (DBT) reconstruction layer is obtained and output; The expression of the preprocessed projection data is as follows: I θ =ln(I θ,0 +1); Where θ represents the projection angle; I θ I represents the projection data of the projection angle θ after preprocessing; θ,0 represents the projection data of the projection angle θ without preprocessing; ln() represents the logarithmic operation with the constant e as the base; The displacement registration method of the isocenter arc motion includes: Obtaining preprocessed projection data; Based on the preset isocenter arc motion model of different reconstruction layer depths, the displacement of the preprocessed projection data at different angles is calculated to obtain the displacement of the projection data along the x-axis determined by the angle and the reconstruction layer depth; According to the displacement of the projection data along the x-axis, the preprocessed projection data is subjected to block coordinate translation processing to obtain translated projection data; The expression of the displacement of the projection data along the x-axis is as follows: Where θ represents the projection angle; x θ represents the x-axis coordinate to be translated when the projection angle is θ; z0 represents the depth of the reconstruction layer; shift(x θ , z0) represents the displacement of the projection data along the x-axis when the reconstruction layer depth is z0 and the projection angle is θ; d represents the radius of the isocenter arc; The translated projection data is expressed as: I θ ′=(x θ ,y θ ,z0); Among them, I θ ′ represents the projection data after translation; x θ ′ represents the x-axis coordinate after translation when the projection angle is θ, θ ′=x θ -shift(x θ , z0), x θ Indicates the x-axis coordinate to be translated when the projection angle is θ; y θ It represents the y-axis coordinate when the projection angle is θ; z0 represents the reconstruction layer depth.

2. The DBT reconstruction method of multi-angle projection registration according to claim 1, characterized in that: The obtaining of multi-angle projection data includes: Based on the X-ray tube doing isocentric circular motion projection, the detector is used to shoot X-ray projections at equal angle intervals to obtain multi-angle projection data.

3. The DBT reconstruction method of multi-angle projection registration according to claim 1, characterized in that: The interpolated projection data is obtained, including: Obtain the projection data after translation; Performing pixel vacancy search processing on the translated projection data to obtain pixel vacancy values; According to the pixel vacancy value, a one-dimensional linear interpolation method is adopted to interpolate the adjacent pixels on both sides of the pixel vacancy value to obtain interpolated projection data.

4. The DBT reconstruction method of multi-angle projection registration according to claim 3, characterized in that: The expression of the one-dimensional linear interpolation method is as follows: Among them, y θ- Represents the pixel vacancy value of the y-axis component in the translated projection data; y1 represents the pixel vacancy value y θ- The adjacent upper pixel value; y2 represents the pixel vacancy value y θ- The adjacent next pixel value.

Images