CN109870471B - Single-grating-detection cone-beam CT angle sequence scattering acquisition method - Google Patents
Single-grating-detection cone-beam CT angle sequence scattering acquisition method Download PDFInfo
- Publication number
- CN109870471B CN109870471B CN201910143974.8A CN201910143974A CN109870471B CN 109870471 B CN109870471 B CN 109870471B CN 201910143974 A CN201910143974 A CN 201910143974A CN 109870471 B CN109870471 B CN 109870471B
- Authority
- CN
- China
- Prior art keywords
- scattering
- projection
- grating
- region
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Analysing Materials By The Use Of Radiation (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a single-grating-detection cone-beam CT angle sequence scattering acquisition method, which is characterized in that a scattering field obtained by detecting a small number of gratings in circumferential scanning is utilized, scattering of different areas is divided according to projection tensors, projection-scattering data are fitted, and parameter knowledge of a scattering model is acquired. The scattering model parameter information under different angles is realized through interpolation, the influence of a structure on scattering is simultaneously integrated, the accuracy of scattering acquisition is improved, and then correction is completed to improve the image quality. The single-grating-detection cone-beam CT angle sequence scattering acquisition method provided by the invention is suitable for the scattering estimation of different angle sequences of a measured object with any complexity, has good reliability and stability, can improve the accuracy of a scanning object structure on scattering distribution to a great extent, completes the scattering field estimation under different angle sequences through less scattering field detection, obviously improves the quality of cone-beam CT images and simultaneously improves the scattering acquisition efficiency.
Description
Technical Field
The invention belongs to the field of medical imaging and industrial nondestructive testing related to cone beam CT application, and relates to a cone beam CT angle sequence scattering acquisition method based on single-grating detection.
Background
Cone Beam Computed Tomography (CBCT) is an advanced medical imaging and industrial nondestructive testing technology, and can clearly, accurately and intuitively display the internal structure of a detected object in the form of a two-dimensional or three-dimensional tomographic image without damaging the object, and quantitatively provide the position and size of a defect in the object.
Aliasing of scatter information with projection information is an important factor that limits the quality of imaging. The projection information of the high-density object is seriously influenced by Compton scattering, so that image details are submerged by the scattering, the edge information and the image contrast are reduced, and the trouble is brought to the defect identification of cone beam CT (computed tomography) nondestructive testing. Although the scattering information is superposed in the signal in the form of low-frequency components, and can be suppressed by a hardware method, the collection efficiency of all angle scattering under the circumstance of circular scanning is low, and the influence of scattering at an interested angle position on projection reconstruction cannot be effectively solved.
Currently, the scattered field acquisition method generally includes a hardware method and a software method. Hardware is mainly used for adding hardware equipment, and a scattered field is obtained through related physical operation, so that correction of the scattered field is completed. The hardware method mainly comprises a collimator, an air gap, a wire filter, a scanning slit, a radiopaque lead bar and the like. The software method is based on the projection image, completes image analysis and estimation of the illuminated object through a digital image processing method, and obtains a scattering distribution rule. Including convolution, deconvolution, monte carlo simulations, and the like. In cone beam CT, Huang, Kuidong et al, in Chinese Physics C (2016,40(6): 068202), propose a CBCT scatter correction method based on staggered slits, first complementing missing data of a raster scatter image by a Gaussian filtering method, and then stitching partial scatterings obtained by staggered slit scanning. The method has certain effect, but the splicing accuracy is difficult to control. Bowen Meng et al, in the physical medical (2013,40(1):011907) article, "Single-scan partial-specific scattering in computed tomogry using the peripheral detection of scatter and compressed sensing scatter regression", propose a scatter interpolation method consisting of mixed scatter models, with estimation by the scatter obtained on the boundary regions of the scatter convolution model. A Measurement-Based compressive Sensing Scattering recovery algorithm is proposed by Yang, Fujiang et al in IEEE Transactions on Nuclear Science (2018, PP (99):1-1) by "Scattering Estimation for Cone-Beam CT Using Local Measurement Based on Compressed Sensing" which uses Local Measurement to optimize the model for the Estimation of the scattered field.
Although the different methods achieve certain scattering estimation and correction effects, the influence of the object structure on the scattering field distribution is not considered. For the current requirements of cone beam CT medical imaging and industrial nondestructive testing, the accuracy is often insufficient in practical application.
Disclosure of Invention
Aiming at the problem that the structural information of an object and the scattering correlation under an adjacent angle sequence are not considered in the cone-beam CT scattering estimation method, the invention provides a cone-beam CT angle sequence scattering acquisition method based on single-grating detection. And dividing the scattering of different areas according to the projection tensor by using a scattering field obtained by detecting a small-amplitude grating of circular scanning, and fitting projection-scattering data to obtain parameter knowledge of a scattering model. The scattering model parameter information under different angles is realized through interpolation, the influence of a structure on scattering is simultaneously integrated, the accuracy of scattering acquisition is improved, and then correction is completed to improve the image quality.
The technical scheme adopted by the invention for solving the technical problem comprises the following steps:
(1) scanning is carried out, parameters are kept unchanged in the process, and a small number of scanning projections are obtained;
(2) performing scattered field estimation to complete scattering parameter knowledge detection;
(3) and acquiring angle sequence scattered field estimation through the parameter information obtained by detection.
In the step (1) above, the acquired scan projection includes: the system comprises a single grating projection A, an object projection B and a single grating + object projection C, wherein the object projection B and the single grating + object projection C are acquired in a circular scanning mode.
In the step (2), the specific step of scattered field estimation includes:
1) collecting a grating projection image A, counting grating projection information, taking two higher peak values appearing in the statistics as grating grid area and slit area division threshold values, and marking as tau1And τ2；
2) Extracting the grid region and the slit region of the projection C of the grating + the measured object according to a threshold value, wherein the value is less than tau1Is a grid region with a value greater than tau2The region (b) is a slit region, and the grating scattering image (referred to as an image C1) of the grid region and the object + grating scattering image (referred to as an image C2) of the slit region are obtained by respectively carrying out interpolation and Gaussian filtering on the grid region image and the slit region image;
3) image C1 is subtracted from image C2 to obtain a pure projection image, referred to as image D, and image D is subtracted from object projection B to obtain object fringe field image E.
In the step (2), the specific step of acquiring the knowledge of the scattered field includes:
1) performing pixel-by-pixel gradient calculation on the projection image B to obtain a tensor matrix of the two-dimensional image, and then solving a determinant H and an eigenvalue T of the tensor matrix;
2) obtaining different regions according to different determinants and characteristic values, giving a judgment threshold value threshold, if T & ltthreshold & gt & | H | & gt 0, then calling the region as a flat region, if T & ltthreshold & gt & | H | & gt 0, then calling the region as an edge region, and if T & ltthreshold & | H | > 0, then calling the region as a corner region.
3) Selecting a non-linear model S ═ axbAnd as a fitting function, mapping the regions obtained by projection division with the regions corresponding to the obtained scattering field to obtain projection-scattering curves of different regions, obtaining scattering model parameter knowledge corresponding to different regions at the angle position, and obtaining the scattering parameter knowledge at other angles by the same method, wherein S represents the scattering field, x represents a projection value, and a and b represent scattering model parameters.
In the step (3), the specific step of obtaining the knowledge of the scattering parameters of the angle sequence includes:
1) constructing a grid with sampling position angles as variables, and interpolating the scattering model parameter information of different areas detected in the step (2) to obtain the scattering model parameter knowledge of different angle sequences;
2) substituting the parameter information corresponding to each region under the same angle sequence into a nonlinear model S ═ axbCalculating to obtain scattered fields of different area structures;
3) superposing the scattered fields obtained from the parameter information of each region under the same angle sequence, and performing Gaussian smoothing to obtain the estimation of the scattered fields of the angle sequence;
4) and traversing different angle sequences to complete the scattered field estimation of the different angle sequences.
In the method, the single grating projection A is used for area calibration in the scanning purpose, so that only one grating projection needs to be acquired, and the grid area and the slit area division performed by the subsequent grating + object projection can be directly applied without re-acquisition.
The invention has the beneficial effects that: the single-grating-detection cone-beam CT angle sequence scattering acquisition method provided by the invention is suitable for the scattering estimation of different angle sequences of a measured object with any complexity, has good reliability and stability, can improve the accuracy of a scanning object structure on scattering distribution to a great extent, completes the scattering field estimation under different angle sequences through less scattering field detection, obviously improves the quality of cone-beam CT images and simultaneously improves the scattering acquisition efficiency.
The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
Drawings
FIG. 1 is a flow chart of the algorithm of the present invention.
Fig. 2 is a schematic diagram of a grating structure.
FIG. 3 is a comparison of angular series scatter estimation and scatter obtained from grating detection.
Detailed Description
The method comprises the following steps of performing projection sampling on a titanium alloy part by using the conventional industrial cone-beam CT equipment (AN X-ray source is MXR-451HP/11 of Comet, a flat panel detector is XRD 1621 AN15 ES of PerkinElmer, and the equipment is provided with a scanning mechanism and a computer for system control and calculation), and applying the method to obtain the cone-beam CT angle sequence scattering detected by a single grating, wherein the method comprises the following steps:
(1) selecting the voltage of 420kV and the current of 0.18mA of a ray source through an industrial cone-beam CT device of the multi-energy spectrum ray source, wherein the scanning geometric parameters are as follows: the distance between the source and the detector is 1241.832886mm, and the distance between the source and the rotation center is 965.323242 mm; the reconstruction resolution is 512 x 512, and the circular scanning obtains a single grating projection a at an angular position of 0 ° and an object projection B and a single grating + object projection C at angular positions of 0 °, 60 °, 120 °, 180 °, 240 °, 300 °.
(2) Performing scattered field estimation on the collected sample to finish scattering parameter knowledge detection, and the method specifically comprises the following steps:
1) selecting a grating projection image A at an angle position of 0 degree, counting grating projection information, and taking two higher peaks appearing in the statistics as grating grid area and slit area division thresholds, wherein tau1＝3000，τ2＝41000；
2) Extracting a grid region and a slit region of a projection C of the grating and the measured object at an angle position of 0 degree according to a threshold value, wherein the region with the angle position of less than or equal to 3000 is the grid region, the region with the angle position of more than or equal to 41000 is the slit region, and respectively carrying out interpolation and Gaussian filtering on the grid region image and the slit region image to obtain a grating scattering image (called as an image C1) of the grid region and an object and grating scattering image (called as an image C2) of the slit region;
3) subtracting image C1 from image C2 to obtain a projection image, referred to as image D, and subtracting image D from object projection B at an angular position of 0 ° to obtain an object fringe field image E;
4) performing pixel-by-pixel gradient calculation on the projection image B at the angle position of 0 degrees to obtain a gradient matrix of the two-dimensional image, and then solving a determinant H and an eigenvalue T of a tensor matrix;
5) obtaining different areas according to different determinants and characteristic values, and selecting a judgment threshold value of 1.0 multiplied by 10 through experiments-11If T ≈ threshold&If H | ≈ 0, the region is said to be flat, if T > threshold&If H | ≈ 0, the region is called as edge region, if T > threshold&If H > 0, the region is called an angular point region;
6) selecting a non-linear model S ═ axbAnd as a fitting function, mapping the regions obtained by projection division with the regions corresponding to the obtained scattering field to obtain projection-scattering curves of different regions, obtaining scattering model parameter knowledge corresponding to different regions at the angle position, and obtaining the scattering model parameter knowledge at other angles by the same method, wherein S represents the scattering field, x represents a projection value, and a and b represent scattering model parameters.
(3) Through the detection information of the angle positions of 0 degrees, 60 degrees, 120 degrees, 180 degrees, 240 degrees and 300 degrees, the estimation of the scattered field of different angle sequences is obtained, and the method specifically comprises the following steps:
1) constructing a grid with sampling position angles as variables, and interpolating the scattering model parameter information of different areas detected in the step (2) to obtain the scattering model parameter knowledge of different angle sequences;
2) substituting the parameter information corresponding to each region under the same angle sequence into a nonlinear model S ═ axbCalculating to obtain scattered fields of different area structures;
3) superposing the scattered fields obtained from the parameter information of each region under the same angle sequence, and performing Gaussian smoothing to obtain the estimation of the scattered fields of the angle sequence;
4) and traversing different angle sequences to complete the scattered field estimation of the different angle sequences.
In this embodiment, the method for obtaining cone-beam CT angle sequence scattering based on single-grating detection is characterized in that:
(1) according to projection information of a detection object and gradient representation of projection, obtaining a projected structural region by different determinants H and eigenvalues T of tensor information;
(2) selecting a nonlinear scattering model to complete projection-scattering fitting under different structural regions to obtain scattering model parameters of the different structural regions;
(3) the angle sequence scattered field estimation is obtained by scattered field knowledge detection under a small uniform sampling angle, the scattering correction of projection is completed, and a high-quality image is obtained.
Fig. 3 is a comparison between angle sequence scattering estimation and scattering obtained by grating detection, and it can be seen that the method of the present invention almost matches the scattering field distribution obtained by grating detection with respect to the scattering field estimation under the angle sequence, and can obtain the scattering field estimation of other angle sequences by using the small angle scattering detection with high accuracy, thereby completing the scattering correction of the projection.
Claims (2)
1. A single grating detected cone beam CT angle sequence scattering acquisition method is characterized by comprising the following steps:
(1) scanning is carried out, parameters are kept unchanged in the process, and a small number of scanning projections are obtained;
(2) performing scattered field estimation to complete scattering parameter knowledge detection;
(3) acquiring angle sequence scattered field estimation through parameter information obtained by detection;
in the step (1), the acquired scan projection includes: the system comprises a single grating projection A, an object projection B, a single grating and an object projection C, wherein the object projection B, the single grating and the object projection C are obtained in a circular scanning mode;
in the step (2), the specific step of scattered field estimation comprises:
1) collecting a grating projection image A, counting grating projection information, taking two higher peak values appearing in the statistics as grating grid area and slit area division threshold values, and marking as tau1And τ2；
2) Extracting the grid region and the slit region of the projection C of the grating + the measured object according to a threshold value, wherein the value is less than tau1Is a grid region with a value greater than tau2The region (2) is a slit region, and the grid region image and the slit region image are subjected to interpolation and Gaussian filtering respectively to obtain a grating scattering image C1 of the grid region and an object + grating scattering image C2 of the slit region;
3) subtracting image C1 from image C2 to obtain a pure projection image, referred to as image D, and subtracting image D from object projection B to obtain object fringe field image E;
in the step (2), the step of detecting the knowledge of the scattering parameters includes:
1) performing pixel-by-pixel gradient calculation on the object projection B to obtain a tensor matrix of the two-dimensional image, and then solving a determinant H and an eigenvalue T of the tensor matrix;
2) obtaining different regions according to different determinants and characteristic values, giving a judgment threshold value threshold, if T & ltthreshold & gt and | H | & gt are & lt 0 & gt, the region is called a flat region, if T & ltthreshold & gt and | H | & lt 0 & gt, the region is called an edge region, and if T & ltthreshold & gt and | H | > 0, the region is called an angular point region;
3) selecting a non-linear model S ═ axbAs a fitting function, mapping the regions obtained by projection division with the regions corresponding to the obtained scattering field to obtain projection-scattering curves of different regions, obtaining scattering model parameter knowledge corresponding to different regions at an angle position, and obtaining scattering parameter knowledge at other angles by the same method, wherein S represents the scattering field, x represents a projection value, and a and b represent scattering model parameters;
in the step (3), the specific step of obtaining the angle sequence scattered field estimation by detecting the obtained parameter information includes:
1) constructing a grid with sampling position angles as variables, and interpolating the scattering model parameter information of different areas detected in the step (2) to obtain the scattering model parameter knowledge of different angle sequences;
2) substituting the parameter information corresponding to each region under the same angle sequence into a nonlinear model S ═ axbCalculating to obtain scattered fields of different area structures;
3) superposing the scattered fields obtained from the parameter information of each region under the same angle sequence, and performing Gaussian smoothing to obtain the estimation of the scattered fields of the angle sequence;
4) and traversing different angle sequences to complete the scattered field estimation of the different angle sequences.
2. The method as claimed in claim 1, wherein the single-grating-detection cone-beam CT angle-sequence scattering acquisition method comprises: in the method, the single grating projection A is used for area calibration in the scanning purpose, so that only one grating projection needs to be acquired, and the grid area and slit area division performed by the subsequent grating + object projection can be directly applied without re-acquisition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910143974.8A CN109870471B (en) | 2019-02-27 | 2019-02-27 | Single-grating-detection cone-beam CT angle sequence scattering acquisition method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910143974.8A CN109870471B (en) | 2019-02-27 | 2019-02-27 | Single-grating-detection cone-beam CT angle sequence scattering acquisition method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109870471A CN109870471A (en) | 2019-06-11 |
| CN109870471B true CN109870471B (en) | 2021-09-14 |
Family
ID=66919356
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910143974.8A Active CN109870471B (en) | 2019-02-27 | 2019-02-27 | Single-grating-detection cone-beam CT angle sequence scattering acquisition method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109870471B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113034636B (en) * | 2021-03-09 | 2022-04-29 | 浙江大学 | Cone beam CT image quality improvement method and device based on cross-scale multi-energy spectrum CT label |
| CN113177991A (en) * | 2021-04-29 | 2021-07-27 | 中国科学技术大学 | Method for correcting scattering artifacts in CBCT (cone beam computed tomography) based on planned CT (computed tomography) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104161536A (en) * | 2014-07-30 | 2014-11-26 | 西北工业大学 | Cone beam CT scatter correction method and device based on complementary gratings |
-
2019
- 2019-02-27 CN CN201910143974.8A patent/CN109870471B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104161536A (en) * | 2014-07-30 | 2014-11-26 | 西北工业大学 | Cone beam CT scatter correction method and device based on complementary gratings |
Non-Patent Citations (1)
| Title |
|---|
| Scatter measurement and correction method for cone-beam CT based on single grating scan;Kuidong Huang等;《Optical Engineering》;20170629;第56卷(第6期);第064106-2至064106-6页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109870471A (en) | 2019-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5254810B2 (en) | Local motion compensation based on list mode data | |
| US7912180B2 (en) | Scattered radiation correction method and scattered radiation correction apparatus | |
| US8111889B2 (en) | Method and apparatus for efficient calculation and use of reconstructed pixel variance in tomography images | |
| US20190213715A1 (en) | System and method for controlling noise in multi-energy computed tomography images based on spatio-spectral information | |
| EP3238136B1 (en) | System and method for reconstructing ect image | |
| Zhao et al. | Patient-specific scatter correction for flat-panel detector-based cone-beam CT imaging | |
| US7176916B2 (en) | Object identifying system for segmenting unreconstructed data in image tomography | |
| JP2017209498A (en) | X-ray computer tomography apparatus and medical image processor | |
| US20110082368A1 (en) | Reconstruction of dynamical cardiac spect for measuring tracer uptake and redistribution | |
| CN109920020A (en) | A kind of Cone-Beam CT morbid state backprojection reconstruction artifact suppressing method | |
| Yang et al. | Imaging and measuring methods for coating layer thickness of TRISO-coated fuel particles with high accuracy | |
| JP2015536732A (en) | Image processing apparatus and method | |
| CN109870471B (en) | Single-grating-detection cone-beam CT angle sequence scattering acquisition method | |
| Li et al. | Multienergy cone-beam computed tomography reconstruction with a spatial spectral nonlocal means algorithm | |
| CN109919868B (en) | Method for detecting and projection weighting correction of cone beam CT beam hardening curve | |
| Martin et al. | An in silico performance characterization of respiratory motion guided 4DCT for high-quality low-dose lung cancer imaging | |
| CN105844678A (en) | Low dose X-ray CT image reconstruction method based on completely generalized variational regularization | |
| US11337671B2 (en) | Methods and systems for improved spectral fidelity for material decomposition | |
| CN114365192A (en) | Confidence map for neural network based limited angle artifact reduction in cone beam CT | |
| CN101893586B (en) | Simplified cone beam computed tomography (CT) scattering detection method | |
| CN111899312B (en) | Iterative compensation finite angle CT projection reconstruction method | |
| US20060233459A1 (en) | Imaging method and device for the computer-assisted evaluation of computer-tomographic measurements by means of direct iterative reconstruction | |
| JP6741158B2 (en) | Scattering estimation method and image processing apparatus | |
| CN115423700A (en) | Scattering correction method for cone beam CT scattering projection ratio estimation | |
| Yossontikul et al. | A comparative study of capacitance image reconstruction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |