CN101533098A - Method for reducing noise in hardening correction of CT beam - Google Patents
Method for reducing noise in hardening correction of CT beam Download PDFInfo
- Publication number
- CN101533098A CN101533098A CN200910021911A CN200910021911A CN101533098A CN 101533098 A CN101533098 A CN 101533098A CN 200910021911 A CN200910021911 A CN 200910021911A CN 200910021911 A CN200910021911 A CN 200910021911A CN 101533098 A CN101533098 A CN 101533098A
- Authority
- CN
- China
- Prior art keywords
- sub
- data
- range
- hardening
- correction
- 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
Images
Abstract
The invention discloses a method for reducing noise in hardening correction of CT beam. The method comprises the following steps that: a group of hardening data which are used for scanning a testing object by an used CT system is acquired, wherein the object penetrating length of each ray corresponds to a multicolor projection gray scale; the original hardening data are simplified by adopting a statistical method based on an object penetrating length histogram; the simplified hardening data undergo least squares fitting to obtain a hardening curve by a newly constructed exponential function; a tangent line of the exponential function on the origin is calculated, and the tangent line is used as a correction straight line; and the beam hardening correction calculation is carried out by adopting the improved beam hardening correction method. The method has the advantages of reducing errors and noise in the original hardening data, further improving the anti-noise capability of hardening curve fitting, improving the calculation method of the beam hardening correction, and basically controlling picture noise to the original level while basically eliminating cup-shaped artifact.
Description
Technical field
The present invention relates to the noise suppressing method in a kind of CT beam hardening correction, belong to the CT technical field.
Background technology
CT (Computed Tomography) technology is utilized ray (being generally X ray) transmission object to be detected, and obtains one group of projected image on detector, cooperates corresponding reconstruction algorithm to obtain the sectioning image of object again.Research at present and the CT that uses can be divided into two-dimensional ct and three dimensional CT two big classes, can segment accordingly again with the different of reconstruction mode according to scanning.The CT technology has the ability that object integral body (outside and inner) is detected, and is used widely in fields such as medical science and industry.
In the CT system, the X-ray beam that radiographic source sends has the energy distribution of certain limit, and this ray is called the polychrome ray.When polychrome beam and matter interaction, the damping capacity of energy photons wherein is greater than high-energy photons, therefore the average energy of beam raises along with the increase of transmission thickness, cause the attenuation ratio of beam more and more littler, this phenomenon is called beam hardening (Beam Hardening).It is the monoenergetic hypothesis that the CT reconstruction algorithm is based on X-ray beam, if directly the projection of adopting the polychrome beam to produce is rebuild, to cause occurring on the sectioning image cup-shaped or strip artifact, image can produce distortion when serious, makes structure, size, density, one-tenth in the section grade physicochemical property interpretation and metering exactly.
Beam hardening is a significant problem that must solve in the CT practical application.At present, the beam hardening correction method of CT mainly is divided into monoenergetic method and dual intensity method two big classes.Because operational complicacy, the dual intensity method seldom is used in engineering practice.The monoenergetic method is easy to realize that practical application effect is also relatively good, therefore is widely studied.Yang Min, Lu Hongnian, people such as great distance are in " optical technology " (2003,29 (2): the method that proposes in article 177-182) " ray hardened correction is studied in the CT reconstruct " is exactly a kind of typical monoenergetic correction method, it proofreaies and correct thinking: utilize the wedge shape die body to obtain ray and pass through one group of corresponding relation data (being called the data of hardening) between object length and the polychrome projection gray level, again these group data are carried out fitting of a polynomial, from true origin this curve is done tangent line then, set up the funtcional relationship of polychrome data and monochromatic data with this tangent line, thereby reach the purpose of hardening correcting.This method implements simply, but requires to have the die body with the identical material of object to be detected, and this has just influenced the range of application and the dirigibility of this method.
Because the Quantum Properties of the X ray that sends of radiographic source itself, and the various noises in the digital imagery hardware, make to have much noise inevitably in the projected image of CT scan gained.In addition, no matter what method of employing is obtained the sclerosis data, all can not avoid detecting error and noise fully.In this case, if directly adopt said method to carry out beam hardening correction, the curve of fitting of a polynomial may produce the vibration of non-expectation in some interval, proofreaies and correct this moment and can't carry out, and the original noise of polychrome projection can be exaggerated, thereby reduces the signal to noise ratio (S/N ratio) of CT image.
Summary of the invention
In order to overcome the various noises of prior art there is the deficiency of adverse effect in the CT beam hardening correction, the invention provides the noise suppressing method in a kind of CT beam hardening correction, can strengthen the robustness of beam hardening correction method.
The technical solution adopted for the present invention to solve the technical problems comprises the following steps:
(1) obtain one group of sclerosis data that employed CT system scans object to be detected, wherein each ray passes through the corresponding polychrome projection gray level of object length, and acquisition methods can carry out according to known technology;
(2) adopt and based on the method for passing through the object length statistics with histogram this to be organized original sclerosis data and simplify, reducing the sum of errors noise in the original sclerosis data, and reduce the sclerosis data volume that is used for follow-up hardening curve match, save computing time;
(3) adopt an exponential function of neotectonics that the sclerosis data after simplifying are carried out least square fitting and obtain hardening curve, this function shape is stable, can not cause the curve shape of match to produce radical change, further suppress the influence of noise beam hardening correction because of containing noise in the sclerosis data;
(4) calculate the tangent line of this exponential function at initial point, with this tangent line as proofreading and correct straight line;
(5) adopt improved beam hardening correction method to carry out beam hardening correction and calculate, avoided the noise scale-up problem that causes in the conventional method calculating.
Noise suppressing method in the above-mentioned CT beam hardening correction not only can be applied to two-dimensional ct, can also be applied to three dimensional CT.
In above-mentioned steps (2), as follows based on the concrete steps of the sclerosis data simplifying method that passes through the object length statistics with histogram:
1) the original sclerosis data of traversal find and pass through object length maximal value (passing through the object length minimum value default is 0);
2) the data number of expectation after simplifying is set, for ease of handling, establish the data number after statistics with histogram sub-range number equals to expect to simplify, it is identical to get each sub-range size, and the sub-range size equals to pass through the object length maximal value divided by the sub-range number, calculates the start-stop position in each sub-range;
3) be that key word carries out conventional statistics with histogram to original sclerosis data to pass through object length, obtain whole data point distribution situation of passing through the object length interval;
4) respectively at each statistics sub-range, calculate the data point gray average in the sub-range, get this sub-range gray average 5%～20% as the valid data discrimination threshold, it is effective gray scale lower limit=sub-range gray average-discrimination threshold, the effective gray scale upper limit=sub-range gray average+discrimination threshold, the data point outside will being positioned between effective gray area is considered as bad point and deletion;
5) respectively at each statistics sub-range, calculate the gray scale of the valid data gray average in sub-range as this sub-range output data point, the length of this data point is the midrange in this sub-range, for significant figure strong point number is 0 sub-range, and the gray scale of its output data point obtains by the output data point grey value interpolation in two sub-ranges about adjacent with this sub-range.
In above-mentioned steps (3), be an essential characteristic that is positioned at first quartile, monotone increasing and crosses the non-linear concave curve of initial point according to hardening curve, construct a new index function that strengthens hardening curve match stability:
L＝f(P
p)＝a·P
p·exp(b·P
p c) (1)
P wherein
pBe the polychrome projection gray level value in the sclerosis data, a, b, c are fitting coefficients.By the physical property of sclerosis data and curve itself as can be known, can not be less than or equal to 0 by the data of hardening being carried out function coefficients a, b, c that least square fitting obtains, i.e. a〉0, b 0, c 0.By to the simple analysis of 1,2 order derivatives of (1) formula as can be known, the mathematical property of this function is: for P
p∈ (0 ,+∞), f (P
p) be the monotonically increasing concave function, rate of curve is with P
pIncrease and increase, and do not have stationary point and flex point.Obviously, this function shape of neotectonics is stable, meet the essential characteristic of hardening curve, and form is fairly simple, is convenient to The Fitting Calculation and follow-up correction calculation.
In above-mentioned steps (4), can obtain proofreading and correct straight line by simple computation and be (1) formula
L＝f′(0)P
m＝a·P
m (2)
P wherein
mBe the monochromatic projection gray level value after proofreading and correct, a is the fitting coefficient of (1) formula.
In above-mentioned steps (5), be the noise scale-up problem of avoiding causing in the conventional method calculating, improved beam hardening correction computing method step is as follows:
1) adopt known filtering method and parameter that original polychrome projected image is carried out filtering;
2) establishing filtered polychrome projection gray level value is
, will
Estimated value as the noise size;
3) will
Substitution hardening curve equation (1) formula calculates corresponding crossing length value L;
4) straight-line equation (2) formula is proofreaied and correct in the L substitution and calculate the monochromatic projection gray level value after obtaining proofreading and correct
5) correction of a final proof result
The invention has the beneficial effects as follows: because by based on the original sclerosis data compaction of passing through the object length statistics with histogram, reduced the sum of errors noise in the original sclerosis data, and constructed a new fitting function with good curve shape stability, further strengthened the anti-noise ability of hardening curve match, the beam hardening correction computing method have been improved at last, basic eliminate the pseudo-shadow of cup-shaped in, with the picture noise basic controlling in previous level.
The present invention is further described below in conjunction with drawings and Examples.
Description of drawings
Fig. 1 is a beam hardening correction process flow diagram of the present invention;
Fig. 2 is that sectioning image same position gray scale compares before and after the beam hardening correction.
Embodiment
To the object to be detected that a material is an iron, use the inventive method and proofread and correct the pseudo-shadow of its CT beam hardening, carry out following steps:
(1) obtain one group of sclerosis data that employed CT system scans object to be detected, wherein each ray passes through the corresponding polychrome projection gray level of object length, and acquisition methods is a staircase method;
(2) adopt and based on the method for passing through the object length statistics with histogram this to be organized original sclerosis data and simplify, to reduce the sum of errors noise in the original sclerosis data, and minimizing is used for the sclerosis data volume of follow-up hardening curve match, save computing time, as follows based on the concrete steps of the sclerosis data simplifying method that passes through the object length statistics with histogram:
1) the original sclerosis data of traversal find and pass through object length maximal value (passing through the object length minimum value default is 0);
2) the data number that is provided with after expectation is simplified is 10, for ease of handling, the data number after if statistics with histogram sub-range number equals to expect to simplify, it is identical and equal to pass through the object length maximal value divided by the sub-range number to get each sub-range size, calculates the start-stop position in each sub-range;
3) be that key word carries out conventional statistics with histogram to original sclerosis data to pass through object length, obtain whole data point distribution situation of passing through the object length interval;
4) respectively at each statistics sub-range, calculate the data point gray average in the sub-range, get this sub-range gray average 10% as the valid data discrimination threshold, it is effective gray scale lower limit=sub-range gray average-discrimination threshold, the effective gray scale upper limit=sub-range gray average+discrimination threshold, the data point outside will being positioned between effective gray area is considered as bad point and deletion;
5) respectively at each statistics sub-range, calculate the gray scale of the valid data gray average in sub-range as this sub-range output data point, the length of this data point is the midrange in this sub-range, for significant figure strong point number is 0 sub-range, and the gray scale of its output data point obtains by the output data point grey value interpolation in two sub-ranges about adjacent with this sub-range.
(3) an exponential function L=f (P of employing neotectonics
p)=aP
pExp (bP
p c) the sclerosis data after simplifying are carried out least square fitting obtain hardening curve, wherein a=6.458, b=0.4336, c=1.923, this function shape is stable, can not cause the curve shape of match to produce radical change, further suppress the influence of noise beam hardening correction because of containing noise in the sclerosis data;
(4) calculate the tangent line of this exponential function at initial point, with this tangent line as proofreading and correct straight line, i.e. L=f ' (0) P
m=aP
m
(5) adopt improved beam hardening correction method to carry out beam hardening correction and calculate, the noise scale-up problem of having avoided conventional method to cause in calculating, improved beam hardening correction computing method step is as follows:
1) adopt Gauss's weighted filtering method that original polychrome projected image is carried out filtering;
2) establishing filtered polychrome projection gray level value is
Will
Estimated value as the noise size;
3) will
Substitution hardening curve equation (1) formula calculates corresponding crossing length value L;
4) straight-line equation (2) formula is proofreaied and correct in the L substitution and calculate the monochromatic projection gray level value after obtaining proofreading and correct
5) correction of a final proof result
Fig. 2 is that sectioning image same position gray scale compares before and after the beam hardening correction, therefrom can obviously find out, though general beam hardening correction has been eliminated the pseudo-shadow of cup-shaped on the whole substantially, but original noise is obviously amplified, and the inventive method is in the pseudo-shadow of basic elimination cup-shaped, make the picture noise after the correction keep original level substantially, show that the inventive method is practicable.
Claims (5)
1, the noise suppressing method in the CT beam hardening correction is characterized in that comprising the steps:
(1) obtain one group of sclerosis data that employed CT system scans object to be detected, wherein each ray passes through the corresponding polychrome projection gray level of object length;
(2) adopt and based on the method for passing through the object length statistics with histogram this to be organized original sclerosis data and simplify;
(3) adopt an exponential function of neotectonics that the sclerosis data after simplifying are carried out least square fitting and obtain hardening curve;
(4) calculate the tangent line of this exponential function at initial point, with this tangent line as proofreading and correct straight line;
(5) adopting improved beam hardening correction method to carry out beam hardening correction calculates.
2, the noise suppressing method in the CT beam hardening correction according to claim 1 is characterized in that:
In the described step (2), comprise the steps: based on the sclerosis data simplifying method that passes through the object length statistics with histogram
1) the original sclerosis data of traversal find and pass through the object length maximal value;
2) the data number of expectation after simplifying is set, the data number after if statistics with histogram sub-range number equals to expect to simplify, it is identical to get each sub-range size, and the sub-range size equals to pass through the object length maximal value divided by the sub-range number, calculates the start-stop position in each sub-range;
3) be that key word carries out conventional statistics with histogram to original sclerosis data to pass through object length, obtain whole data point distribution situation of passing through the object length interval;
4) respectively at each statistics sub-range, calculate the data point gray average in the sub-range, get this sub-range gray average 5%～20% as the valid data discrimination threshold, it is effective gray scale lower limit=sub-range gray average-discrimination threshold, the effective gray scale upper limit=sub-range gray average+discrimination threshold, the data point outside will being positioned between effective gray area is considered as bad point and deletion;
5) respectively at each statistics sub-range, calculate the gray scale of the valid data gray average in sub-range as this sub-range output data point, the length of this data point is the midrange in this sub-range, for significant figure strong point number is 0 sub-range, and the gray scale of its output data point obtains by the output data point grey value interpolation in two sub-ranges about adjacent with this sub-range.
3, the noise suppressing method in the CT beam hardening correction according to claim 1 is characterized in that:
In the described step (3), be an essential characteristic that is positioned at first quartile, monotone increasing and crosses the non-linear concave curve of initial point, construct a new index function that strengthens hardening curve match stability according to hardening curve
P wherein
pBe the polychrome projection gray level value of sclerosis in the data, a, b, c are fitting coefficients, obtain a by the sclerosis data are carried out least square fitting〉0, b 0, c 0.
4, the noise suppressing method in the CT beam hardening correction according to claim 1 is characterized in that:
In the described step (4), the correction straight line is L=f ' (0) P
m=aP
m, P wherein
mIt is the monochromatic projection gray level value after proofreading and correct.
5, the noise suppressing method in the CT beam hardening correction according to claim 1 is characterized in that:
In the described step (5), improved beam hardening correction computing method step is as follows:
1) original polychrome projected image is carried out filtering;
2) establishing filtered polychrome projection gray level value is
Will
Estimated value as the noise size;
3) will
Substitution
Calculate corresponding crossing length value L;
4) with L substitution L=hundred million ' (0) P
m=aP
mCalculate the monochromatic projection gray level value after obtaining proofreading and correct
5) correction of a final proof result
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100219111A CN101533098B (en) | 2009-04-08 | 2009-04-08 | Method for reducing noise in hardening correction of CT beam |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100219111A CN101533098B (en) | 2009-04-08 | 2009-04-08 | Method for reducing noise in hardening correction of CT beam |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101533098A true CN101533098A (en) | 2009-09-16 |
| CN101533098B CN101533098B (en) | 2011-08-03 |
Family
ID=41103819
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009100219111A Expired - Fee Related CN101533098B (en) | 2009-04-08 | 2009-04-08 | Method for reducing noise in hardening correction of CT beam |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101533098B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102662189A (en) * | 2012-04-13 | 2012-09-12 | 北京华泰诺安科技有限公司 | Method for radiation detection and analysis based on counter |
| CN103445803A (en) * | 2013-09-09 | 2013-12-18 | 深圳先进技术研究院 | CT system beam hardening elimination method and CT system beam hardening elimination system based on sonogram |
| CN109919868A (en) * | 2019-02-27 | 2019-06-21 | 西北工业大学 | A kind of detecting of cone-beam CT beam hardening curve and projection weighted correction method |
| CN114240970A (en) * | 2021-12-21 | 2022-03-25 | 北京适创科技有限公司 | Automatic rectifying and automatic strengthening method for CT data |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7215732B2 (en) * | 2004-09-30 | 2007-05-08 | General Electric Company | Method and system for CT reconstruction with pre-correction |
| CN101126723B (en) * | 2007-09-30 | 2011-06-22 | 西北工业大学 | Cone-beam CT beam hardening calibration method based on section profile reprojection |
-
2009
- 2009-04-08 CN CN2009100219111A patent/CN101533098B/en not_active Expired - Fee Related
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102662189A (en) * | 2012-04-13 | 2012-09-12 | 北京华泰诺安科技有限公司 | Method for radiation detection and analysis based on counter |
| CN102662189B (en) * | 2012-04-13 | 2013-11-27 | 北京华泰诺安科技有限公司 | Method for radiation detection and analysis based on counter |
| CN103445803A (en) * | 2013-09-09 | 2013-12-18 | 深圳先进技术研究院 | CT system beam hardening elimination method and CT system beam hardening elimination system based on sonogram |
| CN103445803B (en) * | 2013-09-09 | 2015-09-30 | 深圳先进技术研究院 | Based on CT system beam hardening removing method and the system thereof of sinogram |
| CN109919868A (en) * | 2019-02-27 | 2019-06-21 | 西北工业大学 | A kind of detecting of cone-beam CT beam hardening curve and projection weighted correction method |
| CN109919868B (en) * | 2019-02-27 | 2022-10-04 | 西北工业大学 | Method for detecting and projection weighting correction of cone beam CT beam hardening curve |
| CN114240970A (en) * | 2021-12-21 | 2022-03-25 | 北京适创科技有限公司 | Automatic rectifying and automatic strengthening method for CT data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101533098B (en) | 2011-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101533098B (en) | Method for reducing noise in hardening correction of CT beam | |
| La Riviere | Penalized‐likelihood sinogram smoothing for low‐dose CT | |
| JP3269960B2 (en) | Post-processing apparatus and method for removing blocking phenomenon | |
| EA017302B1 (en) | Method of noise reduction of digital x-ray image series | |
| US7379575B2 (en) | Method for post- reconstructive correction of images of a computer tomograph | |
| CN103134823B (en) | Convolution-based X-ray CT system beam hardening correction method | |
| CN101510298A (en) | Synthesis correction method for CT pseudo-shadow | |
| CN101639936A (en) | X-ray image enhancing method and system thereof | |
| CN111899188B (en) | Neural network learning cone beam CT noise estimation and suppression method | |
| CN104318536A (en) | Method and device for CT image correction | |
| CN104161536A (en) | Cone beam CT scatter correction method and device based on complementary gratings | |
| CN104166971A (en) | CT image reconstruction method | |
| Podgorsak et al. | CT artifact correction for sparse and truncated projection data using generative adversarial networks | |
| CN113205461B (en) | Low-dose CT image denoising model training method, denoising method and device | |
| CN105374014A (en) | Image correction method, image correction device, medical image generation method and medical image generation device | |
| US20200342640A1 (en) | System and method for beam hardening correction (bhc) in computed tomography (ct) image reconstruction | |
| Werner | On convergence rates for iteratively regularized Newton-type methods under a Lipschitz-type nonlinearity condition | |
| CN101416881B (en) | CT beam hardening correction method based sectionalized hardening curve | |
| Jian et al. | Beam-hardening correction method based on original sinogram for X-CT | |
| Yang et al. | Fast iterative beam hardening correction based on frequency splitting in computed tomography | |
| US8792747B2 (en) | Image processing method and device, and medical imaging instrument | |
| CN111652949B (en) | Terahertz focusing tomography image reconstruction method based on AdaFM-Net | |
| Min et al. | A weighted rebinned backprojection‐filtration algorithm from partially beam‐blocked data for a single‐scan cone‐beam CT with hybrid type scatter correction | |
| Aootaphao et al. | Penalized-likelihood reconstruction for metal artifact reduction in cone-beam CT | |
| CN112884862B (en) | Cone beam CT temperature drift correction method and system based on centroid projection trajectory fitting |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20110803 Termination date: 20150408 |
|
| EXPY | Termination of patent right or utility model |