CN104778667B - The bearing calibration of cupping artifact in a kind of Cone-Beam CT based on level set - Google Patents
The bearing calibration of cupping artifact in a kind of Cone-Beam CT based on level set Download PDFInfo
- Publication number
- CN104778667B CN104778667B CN201510176350.8A CN201510176350A CN104778667B CN 104778667 B CN104778667 B CN 104778667B CN 201510176350 A CN201510176350 A CN 201510176350A CN 104778667 B CN104778667 B CN 104778667B
- Authority
- CN
- China
- Prior art keywords
- cone
- artifact
- level set
- correction
- sectioning 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
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses a kind of bearing calibration of cupping artifact in Cone-Beam CT based on level set, this method is corrected applied to Cone-Beam CT sectioning image.This method can be adaptive the cupping artifact of carries out Cone-Beam CT correct, correction can be automatically performed without manual intervention.This method does not need multiple scanning testee；The complexity of cone-beam CT system is not increased；For the sectioning image after reconstruction, it can be directly facing user, do not need to make any change to the existing equipment of original Cone-Beam CT, so that it may complete correction work, this method can be carried out efficiently the cupping artifact correction of Cone-Beam CT, while can also improve the contrast of image.
Description
Technical field
The present invention relates to a kind of bearing calibrations of cupping artifact in Cone-Beam CT sectioning image using level set algorithm, belong to
Technical field of image processing.
Background technique
Cone-Beam CT has good scanning speed and radiation utilization rate, can effectively reduce the load output of X-ray tube, drop
Low scanning cost, can also be quickly obtained high-resolution three-dimension faultage image data.
Currently, influencing many because being known as of cone-beam CT reconstruction picture quality, such as: x-ray scattering, noise, geometric error, energy
Spectrum, probe unit non_uniform response etc..But since cone-beam plate CT uses large-scale X-ray flat panel detector, this makes into
Image quality amount is more vulnerable to X-ray scattering and the influence of beam hardening compared with traditional CT.It is formed because of scattering and beam hardening
Artifact, the inaccuracy of CT number etc. seriously affect analysis and judgement to reconstruction image.In the cone-beam CT reconstruction figure for being directed to human body
As in, these artifacts are mainly shown as CT value inhomogeneities artifact, the artifact for showing as cup-shaped mostly of the artifact.These artifacts
In terms of for the visualization display based on threshold value and the segmentation aspect influence of the pyramidal CT image based on threshold value is very serious.Therefore,
Correction for Cone-Beam CT cupping artifact seems particularly necessary.
In order to reduce the influence of cupping artifact (that is: CT value inhomogeneities artifact), currently, most of bearing calibration is main
It is cupping artifact caused by considering because of scattering, and the research of the prior art or documents and materials is concentrated mainly on in projected image
Scatter correction.These methods can be divided into two classes: one kind is based on software, and another kind of is based on hardware.It is most of with soft
Method based on part is based on Monte-Carlo Simulation, it is that a kind of pair of cone beam computed tomography (CT) scattering corrects very effective method.But it
It is especially time-consuming.In recent years, it has been proposed that some improved Monte Carlo simulation algorithms, such as: the method based on GPU.
But even if heavy calculating still hinders its actual application using fast algorithm.Many hardware based bearing calibrations
It is also good, such as: part ray blocks, primary is modulated etc..
The shortcomings that prior art, specifically includes that
(1) prior art is primarily directed to projection image correction, not directly against the correction of sectioning image after reconstruction.
(2) most of prior art concentrates in the method for artifact correction caused by due to scattering, and needs to add mostly
Add hardware device, such as: patent 200710019084 and 201310039298, the two patents require to set in expensive Cone-Beam CT
Standby upper addition hardware device, increases the complexity of operation and causes potential security risk to equipment.Especially patent
200710019084 need twice sweep testee, so the undoubtedly amount of radiation of increased measured object.
In conclusion Monte-carlo Simulation Method expends the time very much, just in the method for the prior art or documents and materials
Correction result is limited to the structure of modulation panel itself in grade ray modulation method, is based on partial dispersion radionetric survey method, some
It needs to increase exposure dose, some methods are not high to the accuracy of estimation of scatter distributions.And the present invention can well solve
The problem of face.
Summary of the invention
Present invention aims at propose a kind of bearing calibration of cupping artifact in the Cone-Beam CT based on level set, this method
It is corrected applied to Cone-Beam CT sectioning image.This method can be adaptive carries out Cone-Beam CT cupping artifact correction, without manually
Intervention can be automatically performed correction.This method does not need multiple scanning testee；The complexity of cone-beam CT system is not increased；
For the sectioning image after reconstruction, it can be directly facing user, do not need to carry out any change to the existing equipment of original Cone-Beam CT
It is dynamic, so that it may to complete correction work, this method can be carried out efficiently the cupping artifact correction of Cone-Beam CT, while can also improve
The contrast of image.
The technical scheme adopted by the invention to solve the technical problem is that: cup-shaped is pseudo- in a kind of Cone-Beam CT based on level set
The bearing calibration of shadow, this method comprises the following steps:
Step 1: obtaining the Cone-Beam CT slice of data for having artifact；
Step 2: overall situation flow function ε (φ, f are calculated according to formula (7) and formula (8)s, p) and level set constraint equation F
(φ,p,fs)；
Step 3: to the Cone-Beam CT slice of data for having artifact, according to formula (10), fixed p and fs, use finite difference
Method iteration develops out
Step 4: according to formula (11), fixed φ and fs, calculate the estimated value of p variate-value
Step 5: according to formula (12), fixed φ and p calculates fsThe estimated value of variate-value
Step 6: ifIt does not restrain or up to the number of iterations is not arrived, then enablesAnd return to step
3；
Step 7: the Cone-Beam CT sectioning image according to formula (13), after calculating correction.
Effective effect:
1, the present invention can be corrected directly against the cupping artifact of the sectioning image after reconstruction.
2, calculation amount of the present invention is relatively small, can be carried out efficiently the same of Cone-Beam CT sectioning image cupping artifact correction
When, the contrast of image can be also improved well.
3, the present invention can be directly facing CT slice demand user, not need to carry out original Cone-Beam CT existing equipment any
Change can complete correction work.
Detailed description of the invention
Fig. 1 is flow chart of the method for the present invention.
Fig. 2 is the schematic diagram of image slice.
Identifier declaration: CTP486 original slice figure (a) is indicated；(b) it indicates using the CTP486 after the method for the present invention correction
Slice map；(c) CTP Top original slice figure is indicated；(d) it indicates using the CTP Top slice map after the method for the present invention correction.
Fig. 3 is the horizontal sectional drawing of image slice.
Identifier declaration: (a), (b) indicate center row horizontal sectional drawing.
Fig. 4 is CTP486 cupping artifact index τcupThe schematic diagram of zoning.
Fig. 5 is CTP Top cupping artifact index τcupZoning.
Fig. 6 is the schematic diagram of image (right side) after image (left side) and gray scale inhomogeneity correction before gray scale inhomogeneity correction.
Fig. 7 is 1 dimension horizontal sectional drawing of the head Fig. 6 Cone-Beam CT slice.
Fig. 8 is the experiment of breast Cone-Beam CT sectioning image gray scale inhomogeneity correction, and left column is image (a, c) before correcting, right column
For the schematic diagram of image (b, d) after correction.
Specific embodiment
The invention is described in further detail with reference to the accompanying drawings of the specification.
As shown in Figure 1, in a kind of Cone-Beam CT based on level set cupping artifact bearing calibration, this method includes following step
It is rapid:
Step 1: obtaining the Cone-Beam CT slice of data for having artifact；
Step 2: overall situation flow function ε (φ, f are calculated according to formula (7) and formula (8)s, p) and level set constraint equation F
(φ,p,fs)；
Step 3: to the Cone-Beam CT slice of data for having artifact, according to formula (10), fixed p and fs, use finite difference
Method iteration develops out
Step 4: according to formula (11), fixed φ and fs, calculate the estimated value of p variate-value
Step 5: according to formula (12), fixed φ and p calculates fsThe estimated value of variate-value
Step 6: ifIt does not restrain or up to the number of iterations is not arrived, then enablesAnd return to step
3；
Step 7: the Cone-Beam CT sectioning image according to formula (13), after calculating correction.
Bearing calibration of the invention specifically includes:
Rebuilding sectioning image decomposition includes:
Algorithm for reconstructing is based on FDK algorithm in Cone-Beam CT, and reconstruction image collection f can be write as
Wherein dsoDistance of the expression source to rotary shaft, I3DThe sequence of (t, z (r), φ) expression projected image.Here, it projects
Image I3DIt is decomposed into as follows:
I3D=P3D+S3D+n (2)
Wherein P3DIt is the true picture of never artifact, S3DIt is the artifact sections as caused by scattering and beam hardening, n is
Noise.P3DIt is the intrinsic physical attribute of object slice object, if there is substance in N, it can be assumed that it is divided into N number of gray scale
(that is: CT value) constant region.It is observed that the present invention can have found that Cone-Beam CT is sliced artifact S3DIt is that slowly varying cup-shaped is pseudo-
Shadow.It is as follows to rewrite reconstruction image formula,
Wherein
That is reconstruction image can be expressed as three independent elements additions.
Defining neighborhood self-energy function includes:
The present invention considers the circular region Θ of a radius ρy∈ Ω, and define Θy=x:| x-y | and≤ρ }, y ∈ Ω.
Entire domain sectioning image is denoted as Ω, and Ω can be divided into N number of subregion, respectivelyDue to fsIt is equally one slowly varying
Image, so in circular domain ΘyIn, the value f of all Xs(x) close to value fs(y).Therefore, in each subregion Θy∩Ωi,
Intensity fp(x)+fs(x) close to coefficient pi+fs(y)。
Therefore, for region Θy∩ΩiThe present invention has:
f(x)≈fs(y)+pi, (4)
To neighborhood ΘyInterior, the present invention defines energy function and is
Whereinα is a normaliztion constant, so that ∫ K (u)=1, δ are Gaussian functions
Standard variance (or scale parameter), ρ is ΘyThe radius of circular domain.
It defines global energy function and level set function includes:
For whole image region, energy function can be indicated are as follows:
The present invention indicates energy function using level set, and by taking the level set of N=2 as an example, other multilevel collection situations can
It is deduced accordingly.For convenience, present invention vector p=(p1,…pN) represent constant p1,…pN.Therefore, level set function φ,
Vector p and cup-shaped fsIt is the variable of energy ε, therefore ε can be write as ε (φ, fs, p) following form:
Wherein, M1(φ)=H (φ) and M2(φ)=1-H (φ), H is Heaviside function, and φ is a level set
Function, zero level integrate as C0={ x: φ (x)=0 }, its component image field are two disjoint Ω1={ x: φ (x) > 0 } and
Ω2The region={ x: φ (x) < 0 }.For 2 level sets, C0={ x: φ (x)=0 } and Ck={ x: φk(x)=0 } then figure
As being divided into three regions, i.e. N=3.
In order to constrain zero level set function, the present invention adds bound term to entire level set function.New height collection equation is fixed
Justice are as follows:
F(φ,p,fs)=ε (φ, fs,p)+vL(φ)+μRq(φ) (8)
WhereinFor constraining the slickness of zero level contour line, H is Heaviside function,
For maintaining symbolic distance attribute, wherein
7.2.4 predictor φ, p, fs
Fixed p and fs, minimize F (φ, p, fs) normal gradients descending method is used, solve gradient flow equation
It obtains:
Here ei=∫ K (y-x) | f (x)-(fs(y)+pi)|2Dy,It is terraced operator, div () is divergence operator, and δ is Di
Clarke δ function.
We are developed φ out using finite difference method iteration, are as a result usedIt indicates.Method is described as follows: being setFor φ
The value of discretization,For the right formula approximation of formula (9).Then haveAnd then obtain iteration
Formula:
Wherein enablec0For the constant greater than 0.
Fixed φ and fs, with being worth after the optimization for indicating p.Minimize ε (φ, fs, p), thenIt obtains
Fixed φ and p.Minimize ε (φ, fs, p), thenIt obtains
By iterative method, F (φ, p, f are minimized each times).In current iteration using last estimated result into
Row iteration is carried out until Data Convergence or until reaching the number of iterations.Finally, the present invention corrected after image are as follows:
Experimentation and result of the invention specifically includes:
Quantitative analysis index definition includes:
The present invention defines cupping artifact τcup=100 (uM,edge-uM,center)/uM,edge, wherein uM,centerAnd uM,edgeIt is mould
The CT value (that is: gray value) of body center and peripheral.
Root mean square contrast is expressed asWherein IijIt is two dimensional image (i, j)
Position pixel value.
Contrast signal-to-noise ratio degree CNR calculation formula is CNR=| uM,1-uM,2|/σM, wherein uM,1, uM,2For two contrast districts
The mean value in domain, σM,1,σM,2For the standard deviation of contrast district, pixel noise σMIt is defined as σM=(σM,1+σM,2)/2。
500 die body Cone-Beam CT of Catphan is sliced cupping artifact correction experiment.
Experiment of the invention is tested using Catphan500 die body, using at the top of CTP486 in die body and die body (i.e.:
CTP Top).It can find out from Fig. 2, the present invention can eliminate cupping artifact and be difficult to perceive level to human eye；It is indicated in Fig. 3
It is 1 dimension horizontal profile out in corresponding diagram 2, can therefrom finds out that the present invention can significantly eliminate the cupping artifact of sectioning image.It is right
Tables 1 and 2 is shown in the analysis of scatter correction effect quantitatively.It can be obtained from table, method proposed by the present invention makes cupping artifact τcupIt is flat
Decline about 91.8%.
The correction front and back quantitative analysis of table 1:CTP486 sectioning image, the chart of cutting before correction is shown as PI_NONE, after correction
Slice map is expressed as PI_SC
The correction front and back quantitative analysis of table 2:CTP Top sectioning image, the chart of cutting before correction is shown as PI_NONE, after correction
Slice map be expressed as PI_SC.
The Cone-Beam CT slice cupping artifact correction experiment of human skull
It is as shown in Figure 6 for the Cone-Beam CT cupping artifact correction front and back slice map of human skull.Fig. 7 is the centre of Fig. 6 image
Row sectional view.There is gray scale unevenness (on the left of Fig. 7) in cerebral tissue before not correcting, shows as the raising of fringe region brightness.School
Uniform gray level after just, has reached ideal effect.
The correction experiment of breast Cone-Beam CT cupping artifact
As shown in figure 8, the experiment of breast Cone-Beam CT shows that gray scale inhomogeneities artifact can be reduced to people by method of the invention
The imperceptible state of eye, and the present invention improves the contrast of image.RMSC after correction is about 1.3 times before correction.
Claims (3)
1. the bearing calibration of cupping artifact in a kind of Cone-Beam CT based on level set, which is characterized in that the method includes as follows
Step:
Step 1: obtaining the Cone-Beam CT slice of data for having artifact；
Step 2: according toWith F (φ,
p,fs)=ε (φ, fs,p)+vL(φ)+μRq(φ) calculates overall situation flow function ε (φ, fs, p) and level set constraint equation F (φ,
p,fs)；Wherein, f (x) is the reconstruction sectioning image containing scatter artefacts, fsIt (y) is the sectioning image of scatter artefacts, piIt is one
Constant,α is a normaliztion constant, so that ∫ K (y-x)=1, δ is Gaussian function
Several standard variances, ρ are ΘyThe radius of circular domain, M1(φ)=H (φ) and M2(φ)=1-H (φ), H are Heaviside letters
Number, and φ is a level set function, p=(p1,…pN) indicate not containing the sectioning image of scatter artefacts, p1,…pNIt is normal
Number indicates the CT value of a certain substance, and L (φ)=∫ ▽ H (φ) dx is used to constrain the slickness of zero level contour line, and H is
Heaviside function, Rq(φ)=∫ q (▽ φ) dx is used to maintain symbolic distance attribute, wherein
Step 3: to have artifact Cone-Beam CT slice of data, according toFixed p and fs, use is limited
Difference method iteration develops outStep 4: according toGu
Determine φ and fs, calculate the estimated value of p variate-value
Step 5: according toFixed φ and p, calculates fsVariable
The estimated value of value
Step 6: ifIt does not restrain or up to the number of iterations is not arrived, then enablesAnd return to above-mentioned steps 3；
Step 7: according toCone-Beam CT sectioning image after calculating correction.
2. the bearing calibration of cupping artifact, feature exist in a kind of Cone-Beam CT based on level set according to claim 1
In the cupping artifact correction of the adaptive carry out Cone-Beam CT of the method；The method does not need multiple scanning testee；No
Increase the complexity of cone-beam CT system；For the sectioning image after reconstruction, it can be directly facing user, do not needed to original cone-beam
The existing equipment of CT makes any change, and can complete correction work.
3. the bearing calibration of cupping artifact, feature exist in a kind of Cone-Beam CT based on level set according to claim 1
In the method is corrected applied to Cone-Beam CT sectioning image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510176350.8A CN104778667B (en) | 2015-04-14 | 2015-04-14 | The bearing calibration of cupping artifact in a kind of Cone-Beam CT based on level set |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510176350.8A CN104778667B (en) | 2015-04-14 | 2015-04-14 | The bearing calibration of cupping artifact in a kind of Cone-Beam CT based on level set |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104778667A CN104778667A (en) | 2015-07-15 |
| CN104778667B true CN104778667B (en) | 2019-05-03 |
Family
ID=53620115
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510176350.8A Active CN104778667B (en) | 2015-04-14 | 2015-04-14 | The bearing calibration of cupping artifact in a kind of Cone-Beam CT based on level set |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104778667B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105528771B (en) * | 2016-01-19 | 2018-09-18 | 南京邮电大学 | The bearing calibration of cupping artifact in a kind of Cone-Beam CT using energy function method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1640361A (en) * | 2005-01-06 | 2005-07-20 | 东南大学 | Positive computerized tomography restoration method for multi-phase horizontal set |
| CN101158653A (en) * | 2007-11-16 | 2008-04-09 | 西北工业大学 | Diffuse transmission measuring and correcting method of cone-beam CT system |
| CN101510298A (en) * | 2009-03-17 | 2009-08-19 | 西北工业大学 | Synthesis correction method for CT pseudo-shadow |
| CN103020928A (en) * | 2012-11-21 | 2013-04-03 | 深圳先进技术研究院 | Metal artifact correcting method of cone-beam CT (computed tomography) system |
| US9265475B2 (en) * | 2011-07-01 | 2016-02-23 | Carestream Health, Inc. | Methods and apparatus for scatter correction for CBCT system and cone-beam image reconstruction |
-
2015
- 2015-04-14 CN CN201510176350.8A patent/CN104778667B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1640361A (en) * | 2005-01-06 | 2005-07-20 | 东南大学 | Positive computerized tomography restoration method for multi-phase horizontal set |
| CN101158653A (en) * | 2007-11-16 | 2008-04-09 | 西北工业大学 | Diffuse transmission measuring and correcting method of cone-beam CT system |
| CN101510298A (en) * | 2009-03-17 | 2009-08-19 | 西北工业大学 | Synthesis correction method for CT pseudo-shadow |
| US9265475B2 (en) * | 2011-07-01 | 2016-02-23 | Carestream Health, Inc. | Methods and apparatus for scatter correction for CBCT system and cone-beam image reconstruction |
| CN103020928A (en) * | 2012-11-21 | 2013-04-03 | 深圳先进技术研究院 | Metal artifact correcting method of cone-beam CT (computed tomography) system |
Non-Patent Citations (2)
| Title |
|---|
| 《Cupping artifact correction and automated classification for high-resolution dedicated breast CT images》;XiaofengYang et al.;《Med.Phys.》;20121031;第39卷(第10期);全文 |
| 圆轨迹锥束CT的伪影成因和校正方法综述;谷建伟 等;《CT和三维成像学术年会论文集》;20041031;全文 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104778667A (en) | 2015-07-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11341613B2 (en) | System and method for image reconstruction | |
| US10621756B2 (en) | Apparatus and method for correcting bias in low-count computed tomography projection data | |
| US10111638B2 (en) | Apparatus and method for registration and reprojection-based material decomposition for spectrally resolved computed tomography | |
| US10255696B2 (en) | System and method for image reconstruction | |
| US20130202080A1 (en) | System and Method for Denoising Medical Images Adaptive to Local Noise | |
| Marchant et al. | Shading correction algorithm for improvement of cone-beam CT images in radiotherapy | |
| Bai et al. | Z-index parameterization for volumetric CT image reconstruction via 3-D dictionary learning | |
| US10605933B2 (en) | X-ray spectral calibration technique for cone-beam CT | |
| CN104408758A (en) | Low-dose processing method of energy spectrum CT image | |
| US11816764B2 (en) | Partial volume correction in multi-modality emission tomography | |
| Wu et al. | Iterative CT shading correction with no prior information | |
| US9311740B2 (en) | Method for enhancing reconstructed 3-D tomosynthesis volume image | |
| US20080144904A1 (en) | Apparatus and Method for the Processing of Sectional Images | |
| Snoeren et al. | Thickness correction of mammographic images by means of a global parameter model of the compressed breast | |
| CN105528771B (en) | The bearing calibration of cupping artifact in a kind of Cone-Beam CT using energy function method | |
| CN105979875B (en) | Medical imaging device, method, apparatus and computer readable medium for generating a breast parameter map | |
| CN110914868A (en) | System and method for scatter calibration | |
| CN104778667B (en) | The bearing calibration of cupping artifact in a kind of Cone-Beam CT based on level set | |
| Rong et al. | CT reconstruction from few-views with anisotropic edge-guided total variance | |
| Fu et al. | PWLS-PR: low-dose computed tomography image reconstruction using a patch-based regularization method based on the penalized weighted least squares total variation approach | |
| Aootaphao et al. | Experiment-based scatter correction for cone-beam computed tomography using the statistical method | |
| Madhav et al. | Development of in vivo characterization of breast tissues through absolute attenuation coefficients using dedicated cone-beam CT | |
| Kim et al. | Model-based noise reduction in scatter correction using a deep convolutional neural network for radiography | |
| Rong et al. | Prior image based anisotropic edge guided TV minimization for few-view CT reconstruction | |
| Haase et al. | Estimation of statistical weights for model-based iterative CT reconstruction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 210003 new model road 66, Gulou District, Nanjing, Jiangsu Applicant after: Nanjing Post & Telecommunication Univ. Address before: 210023 9 Wen Yuan Road, Qixia District, Nanjing, Jiangsu. Applicant before: Nanjing Post & Telecommunication Univ. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |