CN112419434A - Gamma photon 3D imaging noise suppression method and application - Google Patents
Gamma photon 3D imaging noise suppression method and application Download PDFInfo
- Publication number
- CN112419434A CN112419434A CN202011217523.3A CN202011217523A CN112419434A CN 112419434 A CN112419434 A CN 112419434A CN 202011217523 A CN202011217523 A CN 202011217523A CN 112419434 A CN112419434 A CN 112419434A
- Authority
- CN
- China
- Prior art keywords
- gamma photon
- lor
- data
- tof
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000003384 imaging method Methods 0.000 title claims abstract description 27
- 230000001629 suppression Effects 0.000 title claims abstract description 21
- 239000011159 matrix material Substances 0.000 claims abstract description 43
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 39
- 238000001514 detection method Methods 0.000 claims abstract description 36
- 238000005215 recombination Methods 0.000 claims abstract description 16
- 230000006798 recombination Effects 0.000 claims abstract description 16
- 230000008569 process Effects 0.000 claims description 11
- 230000008521 reorganization Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 6
- 230000000670 limiting effect Effects 0.000 claims description 3
- 150000001875 compounds Chemical class 0.000 claims description 2
- 238000013507 mapping Methods 0.000 abstract description 4
- 230000002401 inhibitory effect Effects 0.000 abstract description 2
- 230000002829 reductive effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000149 penetrating effect Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
- G06T11/006—Inverse problem, transformation from projection-space into object-space, e.g. transform methods, back-projection, algebraic methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N23/00—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00
- G01N23/02—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
- G06T11/008—Specific post-processing after tomographic reconstruction, e.g. voxelisation, metal artifact correction
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2223/00—Investigating materials by wave or particle radiation
- G01N2223/03—Investigating materials by wave or particle radiation by transmission
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2223/00—Investigating materials by wave or particle radiation
- G01N2223/10—Different kinds of radiation or particles
- G01N2223/101—Different kinds of radiation or particles electromagnetic radiation
- G01N2223/1013—Different kinds of radiation or particles electromagnetic radiation gamma
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2223/00—Investigating materials by wave or particle radiation
- G01N2223/40—Imaging
- G01N2223/401—Imaging image processing
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Analysis (AREA)
- Chemical & Material Sciences (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Algebra (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Optimization (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Nuclear Medicine (AREA)
Abstract
The invention relates to a gamma photon 3D imaging noise suppression method and application, belonging to the technical field of industrial detection, and the invention provides an improved OSEM algorithm of ToF data recombination based on the mapping principle of gamma photon ToF information and positron annihilation position, establishes a constrained image reconstruction model of ToF index solution, performs data recombination on LoR by taking ToF information as an index, adjusts system matrix elements by taking ToF information weight coefficients, compresses the system matrix by thinning the system matrix elements, and improves the image reconstruction efficiency; and an improved OSEM algorithm for solving constraint is adopted to reconstruct the image, the contribution of the image pixel to LoR is limited to be matched with the distribution probability of a positron annihilation 'position-ToF' curve of the image pixel in ToF information, and the purpose of inhibiting the propagation of statistical noise is achieved, so that the detection resolution is improved.
Description
Technical Field
The invention relates to the technical field of industrial detection, in particular to a gamma photon 3D imaging noise suppression method and application.
Background
The mass density of complex parts of industrial equipment such as engine parts, a pneumatic runner of a tail nozzle and the like in an aviation propulsion system is high, the structure of an inner cavity is complex, nondestructive detection of the defects of the inner cavity structure and the inner wall of the complex parts becomes a difficult problem which needs to be solved urgently in the detection field, and a conventional detection method is limited by the mass density of the complex parts, the complexity of the inner cavity structure and other factors, so that undisturbed and nondestructive high-resolution detection cannot be realized. The gamma photon 3D imaging technology becomes a new method for detecting the inner cavity of the complex piece of the industrial equipment, and in addition, the gamma photon has the advantages of strong penetrating power, strong anti-interference capability and the like, thereby ensuring the technical stability of the application of the gamma photon three-dimensional imaging technology in the industrial detection.
The system matrix describes the probability distribution of gamma photon pairs recorded by the detector, which serves as a bridge connecting the image space and the projection space, and is the key to the iterative reconstruction algorithm. Generally speaking, the system matrix reflects two aspects, namely, the coupling positioning between a pixel and an LOR, namely whether a photon emitted by a certain pixel is detected by a certain LOR; the second is the degree of coupling between a pixel and an LOR, that is, the probability that a photon emitted by a pixel is detected by a LOR.
The gamma photon image reconstruction OSEM algorithm belongs to an unconstrained optimization solving algorithm, and in the image reconstruction process, it is assumed that pixels of an image through which LoRs pass contribute to the generation of the LoRs according to corresponding components of a system matrix, and different pixels belonging to the same LoR are mutually influenced, so that statistical noise occurs between different pixels belonging to the same LoR and is amplified and propagated, and the reconstructed image has the condition of excessively smooth edge and the like, so that the signal-to-noise ratio of the reconstructed image is reduced, and the detection resolution is adversely influenced.
Disclosure of Invention
The invention provides a gamma photon 3D imaging noise suppression method and application for solving the technical problems.
The invention is realized by the following technical scheme:
a noise suppression method for gamma photon 3D imaging, comprising;
s1, introducing gamma photon ToF information, and performing data recombination on the gamma photon LoR by taking the ToF information as an index;
s2, performing gamma photon image reconstruction by using a solution constrained improved OSEM algorithm: in the image reconstruction process, the improved OSEM algorithm for solving the constraint limits that only image pixels in a feasible domain participate in image reconstruction according to the time resolution of the gamma photon detector and the ToF information of the gamma photon pair.
Further, in S1, data reorganization is performed under the premise that the dimension of the system matrix is kept unchanged.
Further, in S1, the system matrix elements are adjusted by ToF information weight coefficients, and the system matrix is compressed by thinning the system matrix elements.
Further, in S1, the LoR data are grouped by gamma photon for ToF information difference Δ t, and then data reconstruction is performed on the LoR data in each group by using a fourier reconstruction method.
Further, the data reorganization specifically includes:
storing LoR data belonging to different gamma photon ToF information differences in a discretization mode;
dividing a time window of a gamma photon detection system into a plurality of intervals according to the time resolution of a gamma photon detector, and matching LoR data in a Root file into the divided time window intervals by reading gamma photon data to store the Root file;
and recombining LoR data in each time window interval into Sinogrm data by adopting a Fourier recombination mode.
Further, the data reorganization is expressed by the following formula:
in the formula (I), the compound is shown in the specification,
the included angle between the vertical line from the center O of the detection system to the projection line l and the X axis is determined;
δ is tan (θ), and θ is an included angle between the projection line l and the Z axis; l denotes the length of the LoR projection line and t is a variable in the time domain.
Further, S2 specifically includes;
determining a positron annihilation 'position-ToF' curve corresponding to each LoR according to the time resolution of the gamma photon detector and the ToF information of the gamma photon pair;
limiting the contribution of the image pixel to LoR to be matched with the distribution probability of a positron annihilation 'position-ToF' curve of the image pixel in ToF information, so that the feasible solution of a reconstruction algorithm is limited in the neighborhood of a positron annihilation calculation position;
and searching an optimal solution in the neighborhood by adopting an improved OSEM algorithm, and reconstructing the image.
Wherein the iterative formula of the solution-constrained improved OSEM algorithm is as follows:
s.t.xj∈[xj-δx，xj+δx]
in the formula, xjA pixel value representing the j (j ═ 1,2, …, n) th image pixel, aijRepresents the probability that a gamma photon pair resulting from positron annihilation at the jth image pixel, along the ith (i ═ 1,2, …, m (m-1)/2) bar LoR, is registered by the gamma photon detector pair;
pijtis a weight coefficient; bi is the number of gamma photon pairs recorded by the gamma photon detector pair corresponding to the ith (i ═ 1,2, …, m (m-1)/2) LoR; δ x represents the domain length of a neighborhood centered around the positron annihilation computation location.
The application of the gamma photon 3D imaging noise suppression method in industrial detection comprises the following steps;
step 1, starting a gamma photon detection system, and scanning a workpiece by using a detector;
and 2, performing three-dimensional image reconstruction on the gamma photon data subjected to scattering correction by adopting the gamma photon 3D imaging noise suppression method to obtain a reconstructed image.
Compared with the prior art, the invention has the following beneficial effects:
the invention provides an improved OSEM algorithm of ToF data recombination based on the mapping principle of gamma photon ToF information and positron annihilation position, establishes a constraint image reconstruction model of ToF index solution, performs data reconstruction on LoR by taking the ToF information as an index, adjusts system matrix elements by using ToF information weight coefficients, compresses the system matrix by thinning the system matrix elements, and improves the image reconstruction efficiency; and an improved OSEM algorithm for solving constraint is adopted to reconstruct the image, the contribution of the image pixel to LoR is limited to be matched with the distribution probability of a positron annihilation 'position-ToF' curve of the image pixel in ToF information, and the purpose of inhibiting the propagation of statistical noise is achieved, so that the detection resolution is improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention.
FIG. 1 is a schematic diagram of ToF data reorganization;
FIG. 2 is a schematic of gamma photon versus ToF information;
FIG. 3 is a schematic diagram of the image reconstruction principle with gamma photon ToF information;
FIG. 4 is a schematic diagram of the improved OSEM algorithm image reconstruction of the present invention;
FIG. 5 three-dimensional drawing of a hydraulic housing;
FIG. 6 prior art imaging results;
FIG. 7 imaging results of the method of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
The invention discloses a gamma photon 3D imaging noise suppression method, which comprises the following steps of;
and S1, introducing gamma photon ToF information, and performing data reorganization on the gamma photon LoR by taking the ToF information as an index.
S2, performing gamma photon image reconstruction by using a solution constrained improved OSEM algorithm: in the image reconstruction process, the improved OSEM algorithm for solving the constraint limits that only image pixels in a feasible domain participate in image reconstruction according to the time resolution of the gamma photon detector and the ToF information of the gamma photon pair.
Because the gamma photon detection system records gamma photon LoR data through an energy window and a time window and stores the gamma photon LoR data in a Root file, before image reconstruction, the Root file LoR data needs to be recombined into a Sinogram, the Sinogram file stores data according to two components of a projection angle and a projection distance of the LoR in the detection system and does not contain ToF information of a gamma photon pair on the LoR, and because the dimension of a system matrix is huge, when image reconstruction is performed by using the system matrix, the system matrix needs to be called once when a pixel corresponding to each LoR is calculated, so the gamma photon image reconstruction speed is generally slow.
When gamma photon ToF information is introduced, the dimension of the system matrix needs to be increased to store the gamma photon ToF information. In the conventional image reconstruction method, the weight coefficients of a system matrix are calculated by using ToF information, each image pixel corresponds to a plurality of weight coefficients, and when the weight coefficients corresponding to each image pixel are stored by increasing the dimension of the system matrix, the system matrix is excessively large, the calculation amount of a computer is increased, and thus the gamma photon image reconstruction efficiency is greatly reduced.
Therefore, in the embodiment, on the premise of keeping the dimension of the system matrix unchanged, data reorganization is performed on the LoR by using ToF information as an index, the elements of the system matrix are adjusted by using ToF information weight coefficients, the system matrix is compressed by thinning the elements of the system matrix, and the image reconstruction efficiency is improved.
The principle of the LoR data recombination of the ToF information in the invention is as follows:
as shown in FIG. 1, assuming that the radius of the detection system is R, in the space coordinate system X-Y-Z, f (X, Y, Z) is a nuclide spatial distribution function, the projection of LoR to the center section of the gamma photon detection system is a projection line l, the included angle between the projection line l and the Z axis is θ, the distance from the center O of the detection system to the projection line l is s, and the included angle between the perpendicular line from the center O of the detection system to the projection line l and the X axis is
tAAnd tBToF where gamma photons reach detector a and detector B, respectively; let Δ t be tA-tBAnd grouping the ToF information difference deltat by the LoR data according to gamma photons, and then performing data recombination on the LoR data in each group by adopting a Fourier recombination method to form a data recombination method taking the ToF information as an index.
In order to facilitate the recombination of ToF information index LoR data, the embodiment firstly adopts a discretization mode to store LoR data belonging to different gamma photon ToF information differences, and then divides a time window of a gamma photon detection system into a plurality of intervals [ t ] according to the time resolution of a gamma photon detector0,t1),[t1,t2),…,[tn-1,tn) The method comprises the steps of reading a gamma photon data storage Root file, matching LoR data in the Root file into divided time window intervals, and finally recombining the LoR data in each time window interval into Sinogram data in a Fourier recombination mode for gamma photon image reconstruction, wherein ToF data recombination can be represented by a formula (1):
in formula (1), δ is tan (θ), l represents the length of the LoR projection line, and t is a variable in the time domain.
The principle of the OSEM algorithm of the present invention is described in detail below:
the gamma photon pair generated after positron annihilation hits a detector and falls into an energy window and a time window which are arranged by a detection system, a LoR is formed, the position of positron annihilation is located on the LoR, the process of gamma photon image reconstruction is to determine the specific position of positron annihilation, and the position of positron annihilation cannot be judged due to the lack of other information related to the gamma photon pair in LoR data.
After gamma photon ToF information is introduced, the position of positron annihilation can be calculated by utilizing the time difference and the light speed of the gamma photon pair reaching the detector pair, and the position is called as the position of positron annihilation calculation. Due to the inherent temporal resolution of the detector itself, the true positron annihilation site is no longer a point but is distributed within a neighborhood centered on the computed position of the positron annihilation. In the image reconstruction process, the true positron annihilation position is limited in the neighborhood as a constraint, so that the number of image pixels participating in image reconstruction can be reduced, and the probability that the true positron annihilation position falls on an error position can be reduced.
As shown in fig. 2, point P is a positron annihilation position, a pair of gamma photons emitted from the position are respectively recorded by a detector a and a detector B and form a LoR, point O is a LoR center, and assuming that the time t for a gamma photon to reach the detector a is tAThe other gamma photon arrives at the detector B at time tBThen, the time difference Δ t between the two γ photons respectively arriving at the detector a and the detector B is: Δ t ═ tA-tB|。
As shown in fig. 3, when the time difference between the gamma photon pair and the two detectors is determined, the positron annihilation computed position can be accurately located, the distance between the positron annihilation computed position and the center of the LoR is Δ x, and the Δ x is computed by adopting a formula (2);
in the formula (2), c represents the speed of light.
In the gamma photon image reconstruction process, the annihilation number of positrons at a certain position is represented by image pixels, and due to the existence of the time resolution of a gamma photon detector, the true positron annihilation position is not a point any more but a neighborhood centered on the positron annihilation calculation position, the length of the neighborhood is deltax, and the deltax is calculated by adopting a formula (3);
in the formula (3), c is the speed of light, deltatIs the time resolution of the gamma photon detector.
A pair of gamma photons from the jth image pixel xjThe probability of emitting and being recorded by a gamma photon detector pair along the ith LoR is aij，aijThe composed matrix is called the system matrix, δ xjRepresenting the distance between the center of the jth image pixel and the center of the bar LoR, and assuming that the time of flight difference of the pair of gamma photons to reach the detector pair is t, the positron annihilation position recorded by the detector system falls within the theoretical positron annihilation positionPut 2 δ xjProbability in neighborhood is pijtIs represented by the formula pijtThe resulting curve is called the positron annihilation "position-ToF" curve, pijtCan be expressed by the following gaussian function, namely formula (4);
in the formula (4), σ is a gaussian coefficient.
The main process of image reconstruction using the method of the present invention is as follows:
firstly, data recombination is carried out on gamma photon LoR by taking ToF information as an index, and then a system matrix weight coefficient is determined according to ToF information of gamma photon pairs reaching a gamma photon detector pair; secondly, calculating a positron annihilation position-ToF curve in the ToF information according to the inherent time resolution of the gamma photon detector, and limiting the contribution of the image pixel to the LoR to be matched with the distribution probability of the image pixel in the positron annihilation position-ToF curve;
and finally, the feasible solution of the reconstruction algorithm is not any position on the LoR any more, but is limited in the neighborhood of the positron annihilation calculation position, and the optimal solution is searched in the neighborhood by adopting the OSEM algorithm.
The improved OSEM algorithm for ToF data reconstruction according to the present invention is described in detail below with reference to the accompanying drawings.
Assuming that the gamma photon detection system consists of m gamma photon detectors, as shown in fig. 4, each two-dimensional slice image to be reconstructed
Containing n image pixels, xjThe pixel value representing the jth (j-1, 2, …, n) image pixel (whose magnitude is proportional to the nuclide activity at that pixel), i.e., the
aijRepresenting a gamma photon pair resulting from positron annihilation at the jth image pixel,probability of being recorded by a gamma photon detector pair along the ith (i ═ 1,2, …, m (m-1)/2) th bar LoR, by aijThe composed matrix is called the system matrix, biThe number of gamma photon pairs (observation projection data) recorded for the gamma photon detector pair corresponding to the ith (i is 1,2, …, m (m-1)/2) LoR, assuming that the ToF information difference of the gamma photon pair corresponding to the LoR reaching the detection system is Deltat, a positron annihilation 'position-ToF' curve can be determined according to the time resolution of the gamma photon detector and the ToF information of the gamma photon pair, and the system matrix expresses the mapping relation between the LOR data and the reconstructed image pixel, and after the ToF information difference Deltat is introduced, a weight coefficient p determined according to Deltat can be obtainedijtAdjusting the system matrix, i.e. the system matrix elements are adjusted to aij×pijt。
At this time, the adjusted system matrix is a mapping relationship between LOR data including ToF information and reconstructed image pixels, and when an image pixel is not in the positron annihilation "position-ToF" curve confidence interval, the weight coefficient p isijtThe numerical value of (2) is zero, so that after ToF information is introduced, some non-zero elements in the system matrix become zero after being adjusted by weight coefficients, thereby achieving the purpose of thinning the system matrix and further improving the image reconstruction efficiency.
Suppose δ xjRepresenting the distance between the center of the jth image pixel and the center of the bar LoR, the time-of-flight difference of a pair of gamma photons reaching the detector pair is t, pijtThe positron annihilation position recorded for the detector system falls within the theoretical positron annihilation position 2 δ xjProbability in the neighborhood, the image reconstruction likelihood function can be expressed by equation (5):
taking logarithm of two sides of the formula (5) to obtain a formula (6):
for a set of gammaPhoton sampling data zijAre multiplied by
The set of gamma photons is sampled for data zijThe likelihood function of (7) can be expressed as:
solving the likelihood function of the sampling data by adopting an EM algorithm, and assuming that the image pixel of the current iteration times is x, the maximum likelihood function can be expressed by a formula (8):
in formula (8), z'ijFor sampled data zijThe conditional expectation value of (9) can be expressed by the following equation:
and (3) expanding and arranging the formula (8) to obtain a formula (10):
substituting equation (9) into equation (10) to obtain the maximum value of equation (10), and applying equation (10) to xjThe first derivative is calculated and the derivative value is 0, and equation (11) is obtained:
finally, from the K-T condition, only when
When the maximum value is obtained, ln L (b | x) is simultaneously obtained on both sides of the formula (11)Multiplied by xjObtaining the formula (12):
since equation (12) is a nonlinear equation set, cannot be solved directly by an analytical method, and needs to be solved by a numerical method, and equation (12) can be expressed as an expression form of x ═ f (x), equation (12) can be expressed by an iterative equation (13) of the numerical method by the law of fixed points:
s.t.xj∈[xj-δx，xj+δx]
to improve the convergence speed of the EM algorithm, LoR observation projection data recorded by a gamma photon detection system is divided into N subsets, Os(s 1, 2.. N.) the LoR observation projection data in the s subset is formed into the OSEM algorithm, however, the OSEM algorithm belongs to an unconstrained optimization algorithm, and during the image reconstruction process, it is assumed that each image pixel has equal probability of contributing to the generation of the LoR. However, in the present invention, the improved OSEM algorithm limits that only image pixels in the feasible region contribute to the generation of the LoR according to the time-of-flight of the gamma photon pair and the temporal resolution of the detector, and the iterative formula of the improved OSEM algorithm can be expressed as formula (14):
s.t.xj∈[xj-δx，xj+δx]
it can be seen from the formula (14) that, on the basis of the OSEM algorithm, by introducing the gamma photon ToF information and the time resolution of the gamma photon detector, in the image reconstruction process, the improved OSEM algorithm limits the contribution of the image pixel to the LoR to be matched with the distribution probability of the positron annihilation position-ToF curve of the image pixel in the ToF information, so that the OSEM algorithm is called as a constrained optimization method, and the image reconstruction quality is improved.
The invention discloses an application of a gamma photon 3D imaging noise suppression method in industrial detection, which comprises the following steps;
step 1, starting a gamma photon detection system, and scanning a workpiece by using a detector;
and 2, performing three-dimensional image reconstruction on the gamma photon data subjected to scattering correction by adopting the gamma photon 3D imaging noise suppression method to obtain a reconstructed image.
In order to verify the effectiveness of the invention, the industrial hydraulic part shown in fig. 5 is taken as a detection object, and gamma photon three-dimensional imaging detection is carried out on the cylinder body structure of the hydraulic part and the shape of the piston rod nut. Because of the limitation of the time resolution of the gamma photon detector, the detection experiment of the hydraulic part inner cavity structure is carried out in the gamma photon detection system built by the GATE simulation platform, the 18F nuclide with the activity of 0.85mCi is arranged in the cylinder body of the hydraulic part to immerse a piston rod nut, the hydraulic part is arranged in the center of the gamma photon detection system, the time resolution of the gamma photon detector is 30ps, the time window of the gamma photon detection system is 330ps, the energy window is 434KeV-587KeV, the acquisition time of the gamma photon is 5s,
the improved OSEM algorithm of OSEM algorithm and ToF data recombination of the invention are respectively adopted to iterate four times to reconstruct the three-dimensional image of the gamma photon data after scattering correction, the subsets are divided into 4, and as the piston rod nut is immersed in the nuclide, the nut imaging area in the gamma photon three-dimensional reconstruction image is covered by the nuclide distribution area, so the embodiment selects the section shape of the nut in the gamma photon two-dimensional slice image for analysis.
Because the piston rod nut does not generate gamma photons, the section where the piston rod nut is located is imaged as a black image, the edge of the cylinder body of the hydraulic part in fig. 6 is fuzzy, and meanwhile, the black area in the section image where the piston rod nut is located is not obvious, so that the shape of the piston rod nut cannot be judged;
in the process of image reconstruction, the improved OSEM algorithm limits that only image reconstruction in a feasible domain participates in image reconstruction, so that the image reconstruction quality is improved. In fig. 7, the edge of the cylinder body of the hydraulic part is clear, and meanwhile, the shape of the piston rod nut can be accurately judged to be a regular hexagon from a black area in a cross-sectional image where the piston rod nut is located.
Meanwhile, the peak signal-to-noise ratio (PSNR) is adopted to measure the gamma photon scattering correction effect in the reconstructed image after the algorithm is improved, the PSNR of the image is 24.9797 in fig. 6, the PSNR of the image is 26.7039 in fig. 7, and the effectiveness of the gamma photon single scattering and multiple scattering correction method provided by the invention can be seen from the PSNR of the reconstructed image.
According to the method, on the premise that the dimension of a system matrix is kept unchanged, ToF information is used as an index to carry out data reconstruction on LoR, ToF information weight coefficients are used for adjusting system matrix elements, the system matrix is compressed through thinning the system matrix elements, image reconstruction efficiency is improved, an improved OSEM algorithm for solving constraint is adopted to carry out image reconstruction, the contribution of image pixels to LoR is limited to be matched with the distribution probability of positron annihilation 'position-ToF' curves of the image pixels in the ToF information, and therefore gamma photon image reconstruction quality is improved.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (9)
1. A gamma photon 3D imaging noise suppression method is characterized in that: comprises the following steps of;
s1, introducing gamma photon ToF information, and performing data recombination on the gamma photon LoR by taking the ToF information as an index;
s2, performing gamma photon image reconstruction by using a solution constrained improved OSEM algorithm: in the image reconstruction process, the improved OSEM algorithm for solving the constraint limits that only image pixels in a feasible domain participate in image reconstruction according to the time resolution of the gamma photon detector and the ToF information of the gamma photon pair.
2. The gamma photon 3D imaging noise suppression method according to claim 1, characterized in that: in S1, data reorganization is performed while maintaining the dimension of the system matrix unchanged.
3. The gamma photon 3D imaging noise suppression method according to claim 2, characterized in that: in S1, the system matrix elements are adjusted by ToF information weight coefficients, and the system matrix is compressed by thinning the system matrix elements.
4. The gamma photon 3D imaging noise suppression method according to claim 2 or 3, characterized in that: in S1, the LoR data are grouped according to gamma photons for ToF information difference Δ t, and then data reconstruction is performed on the LoR data in each group by using a fourier reconstruction method.
5. The gamma photon 3D imaging noise suppression method according to claim 4, characterized in that: the data reorganization specifically comprises:
storing LoR data belonging to different gamma photon ToF information differences in a discretization mode;
dividing a time window of a gamma photon detection system into a plurality of intervals according to the time resolution of a gamma photon detector, and matching LoR data in a Root file into the divided time window intervals by reading gamma photon data to store the Root file;
and recombining LoR data in each time window interval into Sinogrm data by adopting a Fourier recombination mode.
6. The gamma photon 3D imaging noise suppression method according to claim 5, characterized in that: the data reorganization is expressed by the following formula:
in the formula (I), the compound is shown in the specification,
for detecting systemsThe included angle between the perpendicular line from the center O to the projection line l and the X axis;
δ is tan (θ), and θ is an included angle between the projection line l and the Z axis; l denotes the length of the LoR projection line and t is a variable in the time domain.
7. The gamma photon 3D imaging noise suppression method according to claim 1,2, 3, 5 or 6, characterized in that: specifically included in S2 is;
determining a positron annihilation 'position-ToF' curve corresponding to each LoR according to the time resolution of the gamma photon detector and the ToF information of the gamma photon pair;
limiting the contribution of the image pixel to LoR to be matched with the distribution probability of a positron annihilation 'position-ToF' curve of the image pixel in ToF information, so that the feasible solution of a reconstruction algorithm is limited in the neighborhood of a positron annihilation calculation position;
and searching an optimal solution in the neighborhood by adopting an improved OSEM algorithm, and reconstructing the image.
8. The gamma photon 3D imaging noise suppression method according to claim 7, characterized in that: the iterative formula of the solution-constrained improved OSEM algorithm is:
s.t. xj∈[xj-δx，xj+δx]
in the formula, xjA pixel value representing the j (j ═ 1,2, …, n) th image pixel, aijRepresents the probability that a gamma photon pair resulting from positron annihilation at the jth image pixel, along the ith (i ═ 1,2, …, m (m-1)/2) bar LoR, is registered by the gamma photon detector pair;
pijtis a weight coefficient; bi is the number of gamma photon pairs recorded by the gamma photon detector pair corresponding to the ith (i ═ 1,2, …, m (m-1)/2) LoR; δ x represents the domain length of a neighborhood centered around the positron annihilation computation location.
9. Use of the gamma photon 3D imaging noise suppression method according to any one of claims 1 to 8 in industrial detection, characterized in that: comprises the following steps;
step 1, starting a gamma photon detection system, and scanning a workpiece by using a detector;
and 2, performing three-dimensional image reconstruction on the gamma photon data subjected to scattering correction by adopting the gamma photon 3D imaging noise suppression method to obtain a reconstructed image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011217523.3A CN112419434B (en) | 2020-11-04 | Gamma photon 3D imaging noise suppression method and application |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011217523.3A CN112419434B (en) | 2020-11-04 | Gamma photon 3D imaging noise suppression method and application |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112419434A true CN112419434A (en) | 2021-02-26 |
| CN112419434B CN112419434B (en) | 2024-07-09 |
Family
ID=
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113205462A (en) * | 2021-04-06 | 2021-08-03 | 武汉大学 | Photon reflectivity image denoising method based on neural network learning prior |
| CN113670961A (en) * | 2021-08-19 | 2021-11-19 | 南京航空航天大学 | Gamma photon multiple scattering correction method based on space distribution fitting |
| CN113671557A (en) * | 2021-08-16 | 2021-11-19 | 中国科学院近代物理研究所 | Positive electron nuclide activity distribution calculation method, system, device and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104077763A (en) * | 2014-07-08 | 2014-10-01 | 中国科学院高能物理研究所 | TOF-PET image reconstruction method based on compressed sensing theory |
| CN110168410A (en) * | 2016-12-14 | 2019-08-23 | 皇家飞利浦有限公司 | The dead pixel correction rebuild for digital PET |
| CN111638544A (en) * | 2020-05-28 | 2020-09-08 | 清华大学 | Multi-gamma photon coincidence imaging system and method based on slit-hole hybrid collimator |
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104077763A (en) * | 2014-07-08 | 2014-10-01 | 中国科学院高能物理研究所 | TOF-PET image reconstruction method based on compressed sensing theory |
| CN110168410A (en) * | 2016-12-14 | 2019-08-23 | 皇家飞利浦有限公司 | The dead pixel correction rebuild for digital PET |
| CN111638544A (en) * | 2020-05-28 | 2020-09-08 | 清华大学 | Multi-gamma photon coincidence imaging system and method based on slit-hole hybrid collimator |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113205462A (en) * | 2021-04-06 | 2021-08-03 | 武汉大学 | Photon reflectivity image denoising method based on neural network learning prior |
| CN113205462B (en) * | 2021-04-06 | 2022-07-19 | 武汉大学 | Photon reflectivity image denoising method based on neural network learning prior |
| CN113671557A (en) * | 2021-08-16 | 2021-11-19 | 中国科学院近代物理研究所 | Positive electron nuclide activity distribution calculation method, system, device and storage medium |
| CN113670961A (en) * | 2021-08-19 | 2021-11-19 | 南京航空航天大学 | Gamma photon multiple scattering correction method based on space distribution fitting |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8000513B2 (en) | System and method for 3D time of flight PET forward projection based on an exact axial inverse rebinning relation in fourier space | |
| Xu et al. | 4-pi Compton imaging with single 3D position-sensitive CdZnTe detector | |
| CN103052959B (en) | Efficient calculating computed tomography photography with Optimal Recursive | |
| Scheins et al. | Analytical calculation of volumes-of-intersection for iterative, fully 3-D PET reconstruction | |
| CN109472841B (en) | CBCT three-dimensional reconstruction method based on Gaussian mixture/Poisson maximum likelihood function | |
| CN110244299B (en) | Distributed method for SAR image recovery based on ADMM | |
| CN115082446B (en) | Method for measuring aircraft skin rivet based on image boundary extraction | |
| Goli et al. | Bayes' Rays: Uncertainty Quantification for Neural Radiance Fields | |
| Wang et al. | An FPGA-based real-time maximum likelihood 3D position estimation for a continuous crystal PET detector | |
| CN112419434A (en) | Gamma photon 3D imaging noise suppression method and application | |
| CN112419434B (en) | Gamma photon 3D imaging noise suppression method and application | |
| Kang et al. | Point cloud smooth sampling and surface reconstruction based on moving least squares | |
| CN114283235B (en) | Three-dimensional magnetic layer reconstruction method and system based on limited angle projection data | |
| CN112946601B (en) | Gauss-Seidel-based efficient distributed target phase optimization method | |
| Caucci et al. | List-mode MLEM image reconstruction from 3D ML position estimates | |
| Yao et al. | Reconstruction algorithm for regions of interest in γ-photon images | |
| CN115754007A (en) | Damage detection method based on acoustic emission technology and tomography technology | |
| CN114581547A (en) | Method and device for realizing Compton camera reconstruction by constructing sparse source set matrix | |
| Küstner et al. | Parallel MLEM on multicore architectures | |
| Wang et al. | Study of large axial space rebinning algorithm in γ-photon industrial tomography image reconstruction | |
| KR101356881B1 (en) | Method and apparatus for restructuring image for parallel processing of positron emission tomography with high resolution | |
| Tang et al. | Joint Regularized-based Image Reconstruction by Combining Super-Resolution Sinogram for Computed Tomography Imaging | |
| Brard et al. | Image reconstruction strategy for an axial PET | |
| Tanaka et al. | A fast statistical image reconstruction algorithm for TOF-PET using DRAMA (Dynamic RAMLA) strategy | |
| CN118091732A (en) | Compton mobile imaging method based on priori scene model |
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 |