CN1305298C - Image reconstructing method and X-ray CT device - Google Patents
Image reconstructing method and X-ray CT device Download PDFInfo
- Publication number
- CN1305298C CN1305298C CNB200310120365XA CN200310120365A CN1305298C CN 1305298 C CN1305298 C CN 1305298C CN B200310120365X A CNB200310120365X A CN B200310120365XA CN 200310120365 A CN200310120365 A CN 200310120365A CN 1305298 C CN1305298 C CN 1305298C
- Authority
- CN
- China
- Prior art keywords
- projection
- data
- ray
- pass component
- convolved
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Image Processing (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The present invention relates to a method for reconstructing images with little noise and an X-ray CT device for image reconstruction. The solving method comprises the steps that projection data is convolved to solve convolved projection data (601); high-pass components (603, 605) are extracted from the convolved projection data, and marginal information (607) is extracted from the high-pass components; the marginal information is removed from the high-pass components so as to solve noise (609); the noise is removed from the projection data to solve corrected projection data (611); reverse projection (613) is carried out on the corrected projection data.
Description
Technical field
The present invention relates to a kind of image reconstructing method and X ray CT (Computed Tomography) (computer tomography) device, method is come the method for reconstructed image, and the X ray CT device that carries out this image reconstruction by convolution back projection (convoluted back projection) to be particularly related to a kind of data for projection of collecting based on X ray CT device (projection data).
Background technology
In X ray CT device, carry out image reconstruction based on the data for projection that obtains by scanning (scan) object.In image reconstruction, use convolution back projection method.Utilize filtering (filtering) to be equivalent under the situation of convolution algorithm in Fourier space, also convolution back projection method is called filter back projection (filtered back projection) method.
If the intensity that reduces X ray is to reduce the x-ray bombardment amount of object, the SNR of data for projection (signal-to-noise ratio) reduces, and the noise in the reconstructed image (noise) increases.Improve image quality in order to remove noise, reconstructed image is carried out low-pass filtering (low-pass filtering).
Summary of the invention
To the low-pass filtering of reconstructed image, be effectively still owing to having reduced the spatial resolution of image, so have the shortcoming that is difficult to find out minute structure for removing denoising.
Therefore, the method for the image that the objective of the invention is to realize that reconstructed noise is few, and the X ray CT device that carries out this image reconstruction.
(1) the present invention for a viewpoint addressing the above problem is an image reconstructing method, it passes through convolution back projection method reconstructed image based on data for projection, this data for projection utilizes the X-ray scanning object to collect by X ray CT device, it is characterized in that:, obtain the data for projection of convolution the data for projection convolution; From described convolved projection data, extract high pass component; From described high pass component, extract marginal information; From described high pass component, remove marginal information, obtain noise; From described data for projection, remove described noise, obtain the data for projection of correction; Data for projection to described correction carries out back projection.
(2) the present invention for other viewpoints of addressing the above problem is an X ray CT device, and it has: collecting part, and it collects data for projection by utilizing the X-ray scanning object; Reconstruction means, by convolution back projection method reconstructed image, it is characterized in that: described reconstruction means comprises based on described data for projection for it:
To the data for projection convolution, obtain the parts of the data for projection of convolution; From described convolved projection data, extract the parts of high pass component; From described high pass component, extract the parts of marginal information; From described high pass component, remove marginal information, obtain the parts of noise; From described data for projection, remove described noise, obtain the parts of the data for projection of correction; The data for projection of described correction is carried out the parts of back projection.
Carry out by filtering by carrying out in Fourier space for described convolution, but its benefit is high-speed computation.The extraction of described high pass component is the data for projection by the described convolution of low-pass filtering, extracts low-pass component, removes from the data for projection of described convolution that low-pass component carries out, and its benefit is to extract high pass component exactly.The extraction of described marginal information is undertaken by the rim detection computing, and its benefit is to carry out exactly the extraction of marginal information.
In the invention of above-mentioned each viewpoint, owing to, obtain the data for projection of convolution to the data for projection convolution, from convolved projection data, extract high pass component, from high pass component, extract marginal information, from high pass component, remove marginal information, obtain noise, from data for projection, remove described noise, obtain the data for projection of correction, the data for projection of described correction is carried out back projection, so, can realize the method for the image that reconstructed noise is few, and the X ray CT device that carries out this image reconstruction.
Description of drawings
Fig. 1 is the block diagram of X ray CT device.
Fig. 2 is the pie graph of expression X-ray detector.
Fig. 3 is the pie graph of expression x-ray bombardment, checkout gear.
Fig. 4 is the figure of the relation of expression x-ray bombardment, checkout gear and object.
Fig. 5 is the flow chart of the operation of expression X ray CT device.
Fig. 6 is the flow chart of image reconstruction.
Fig. 7 is the schematic diagram of the collection of expression convolved projection data.
Embodiment
Below, be used to implement most preferred embodiment of the present invention with reference to description of drawings.And, the invention is not restricted to the best form of implementation that is used to carry out an invention.Fig. 1 is the block diagram of expression X ray CT device.This device is an example that is used to implement best mode of the present invention.Formation by this device is shown, realizes a example about the optimised form of the present invention of X ray CT device.Operation by this device is shown, implements a example about the optimised form of the present invention of image reconstructing method.
As shown in Figure 1, this device has: scanning stand (gantry) 2, photography platform (table) 4 and operating console (console) 6.Scanning stand 2 has X-ray tube 20.(collimate: collimation), shine on the X-ray detector 24 by collimator (collimator) 22 shapings from the not shown X ray of X-ray tube 20 emissions for the X ray light beam of circular cone (cone) shape is cone beam (cone beam) X ray.X-ray detector 24 has a plurality of detecting elements that are arranged in matrix (array) shape that the width with the X ray light beam is complementary.The formation of X-ray detector 24 is described in the back separately.
In the space between X-ray tube 20 and X-ray detector 24, photography target is loaded on the photography platform 4 and is written into.X-ray tube 20, collimator 22 and X-ray detector 24 constitute x-ray bombardment, checkout gear.X-ray bombardment, checkout gear are described in the back separately.
Control by X ray controller (controller) 28 from X-ray tube 20 irradiation X ray.And, omitted diagram to annexation between X-ray tube 20 and the X ray controller 28.Collimator 22 is by 30 controls of collimator controller.And, omitted the diagram of annexation between collimator 22 and the collimator controller 30.
More than parts from X-ray tube 20 to collimator controller 30 be loaded on the rotating part 34 of scanning stand 2.The rotation of rotating part 34 is by Rotation Controllers 36 controls.And, omitted diagram to the annexation between rotating part 34 and the rotating part controller 36.
Operating console 6 has data processing equipment 60.Data processing equipment 60 is by for example computer formations such as (computer).Control interface (interface) 62 is connected on the data processing equipment 60.Scanning stand 2 and photography platform 4 are connected on the control interface 62.Data processing equipment 60 is by control interface 62 gated sweep stands 2 and photography platform 4.
Fig. 2 represents the mode configuration of X-ray detector 24.As shown in Figure 2, X-ray detector 24 is the X-ray detectors that a plurality of x-ray detection devices 24 (ik) are arranged in the multichannel (channel) of 2 dimension arrays.The whole X ray photosurface that forms bending or cylinder concave shape of a plurality of x-ray detection devices 24 (ik).
I is a channel number, i=1 for example, and 2 ..., 1000.K is a column number, k=1 for example, and 2 ..., 32.Column number k components identical constitutes the detecting element row respectively in the x-ray detection device 24 (ik).And the detecting element of X-ray detector 24 row are not limited to 32 row, also can be suitable a plurality of or single.
X-ray detection device 24 (ik) is made of combination scintillator (scintillato) and photodiode (photodiode).And, be not limited thereto, also can be the semiconductor x-ray detection device that for example utilizes cadmium tellurium (CdTe) etc., or utilize xenon (Xe) ionisation of gas device type x-ray detection device.
Fig. 3 represents the correlation of X-ray tube 20 and collimator 22 and X-ray detector 24 in x-ray bombardment, the checkout gear.And, the state that Fig. 3 (a) expression is seen from the front of scanning stand 2, (b) state seen from the side of expression.Shown in same figure, after being configured as conical X-ray beam 400 by collimator 22, the X ray of X-ray tube 20 emission shines on the X-ray detector 24.
The expansion of a direction of the conical X-ray beam 400 of expression in Fig. 3 (a).Below this direction is called Width.The Width of X-ray beam 400 is consistent with the orientation of the passage of X-ray detector 24.(b) expansion of other direction of expression X-ray beam 400.Below, this direction is called the thickness direction of X-ray beam 400.The thickness direction of X-ray beam 400 is consistent with the cloth set direction of a plurality of detecting element row of X-ray detector 24.2 propagation directions of X-ray beam 400 are vertical mutually.
Body axes and this X-ray beam 400 are intersected, for example as shown in Figure 4, the object 8 that is loaded on the photography platform 4 is loaded in the x-ray bombardment space.Scanning stand 2 forms the structure that comprises the tubular of x-ray bombardment, checkout gear in inside.
The x-ray bombardment space forms in the inner space of the barrel-shaped structure of scanning stand 2.The image projection of the object 8 by X-ray beam 400 section is to X-ray detector 24.Utilize X-ray detector 24, the X ray that sees through object 8 detects on each detector columns.The thickness t h that shines the X-ray beam 400 on the object 8 regulates by the stretching degree in the aperture of collimator 22.
Parallel with the rotation of x-ray bombardment, checkout gear, mobile continuously along the body axes direction of object 8 by the platform 4 of will photographing like that as shown in arrow 42, x-ray bombardment, checkout gear rotate along the helical track of surrounding object 8 with respect to object 8.Carry out so-called helical scanning (helical scan) like this.At the state of platform 4 that stops to photograph, if axial scan (axialscan) is just carried out in rotational x-ray irradiation, checkout gear.
The direction of rotating shaft of scanning is made as the z direction, the direction that connects pivot and X-ray tube 20 is made as the y direction, will be made as the x direction with the z direction direction vertical with the y direction.Form rotating coordinate system xyz like this.
Every scanning one circle can be collected the data for projection of the image (view) of a plurality of (for example about 1000).The collection of data for projection is by X-ray detector 24, data collection portion 26, data collection buffer 64 a series of carrying out.
The operation of this device is described.Fig. 5 represents operating process (flow) figure of this device.As shown in the drawing, carry out the setting of photography conditions in step (stage) 501.The setting of photography conditions is undertaken by operating means 70 by the user.Like this, set various sweep parameters (scan parameter) such as photography position, tube voltage, tube current, slice thickness.
Next scan in step 503.According to the photography conditions of setting in step 501, scan by scanning stand 2 and photography platform 4.Like this, come the data for projection of intelligence-collecting object 8.Scanning stand 2 and photography platform 4 are examples of gathering-device of the present invention.
Then, at step 507 reconstructed image.Image reconstruction is undertaken by data processing equipment 60.Data processing equipment 60 is examples of reconfiguration device of the present invention.Data processing equipment 60 as hereinafter described, utilizes filter back projection's method reconstructed image based on the data for projection that utilizes scanning to collect.The image of reconstruct is shown and stores in step 507.The demonstration of image is undertaken by display unit 60.The storage of image is undertaken by storage device 66.
The following describes the reconstruct of image, Fig. 6 represents the flow chart by the image reconstruction of data processing equipment 60.As shown in the drawing, in step 601, carry out filtering to data for projection.Filtering is the filtering in Fourier space, is equivalent to convolution.Data for projection reads from storage device 66.
Like this, obtain convolved projection data (convoluted projection data).As shown in Figure 7, convolved projection data becomes 2 dimension data CPij in image, channel space.Convolved projection data is stored in the storage device 66.The data processing equipment 60 that carries out filtering in step 601 is examples obtaining the device of convolved projection data.
Next, carry out low-pass filtering in step 603 to convolved projection data.Low-pass filtering is undertaken by following formula.
[formula 1]
Here, LFk1 is 2 dimension filter factors.Like this, obtain low-pass component (low-passcomponent) LCPij of the data for projection of convolution.LCPij is stored in the storage device 66.
Then, in step 605, low-pass component is deducted from convolved projection data (subtraction).Subtraction is undertaken by following formula.
[formula 2]
HCP
ij=CP
ij-LCP
ij (2)
Like this, obtain high pass component (high-pass component) HCPij of convolved projection data.HCPij is stored in the storage device 66.
By processing, from convolved projection data, extract high pass component in step 603 and step 605.The data processing equipment 60 that carries out the processing of step 603 and 605 is examples that extract the device of high pass component.
High pass component HCPij also contains the information of representing the edge based on noise.Thereby, next, carry out the detection at edge in step 607.Rim detection is undertaken by following formula.
[formula 3]
El
ij=f(HCP
kl)
where:
f is?an?edge?detection?function
k?∈{-c,-c+1,…,0,…,c-1,c}
l?∈{-d,-d+1,…,0,…,d-1,d} (3)
Like this, extract marginal information Elij.Elij is stored in the storage device 66.The benefit that marginal information Elij is also carried out processing as the following formula has been to improve the reliability of marginal information.
[formula 4]
The data processing equipment 60 that carries out rim detection in step 607 is examples that extract the device of marginal information.
Then, from high pass component, deduct marginal information in step 609.Subtraction is undertaken by following formula.
[formula 5]
Noise
ij=HCP
ij-El
ij’ (5)
Obtain noise like this.The data processing equipment 60 of obtaining noise in step 609 is examples obtaining the device of noise.Noise is stored in the storage device 66.
Next, in step 611, from convolved projection data, deduct noise.Subtraction is undertaken by following formula.
[formula 6]
nrCP
ij=CP
ij-Noise
ij (6)
Obtain like this and revise data for projection (amended projection data).Revising data for projection is stored in the storage device 66.Obtaining the data processing equipment 60 of revising data for projection in step 611 is to obtain an example revising the data for projection device.
Then, use the correction data for projection to carry out back projection in step 613.The data processing equipment 60 that carries out back projection in step 613 is examples of back projection device.
Obtain reconstructed image by back projection.Reconstructed image is stored in the storage device 66.Because the correction data for projection becomes the data for projection except that denoising, so reconstructed image does not contain noise.
Resemble the prior art the reconstructed image low-pass filtering owing to remove denoising and be not, but to convolved projection data, difference noise and marginal information are removed denoising, so, can only remove denoising and do not damage the spatial resolution of reconstructed image.
Claims (6)
1. an image reconstructing method passes through convolution back projection method reconstructed image based on data for projection, and this data for projection utilizes the X-ray scanning object to collect by X ray CT device, it is characterized in that described method comprises the following steps:
Data for projection is carried out convolution, obtain convolved projection data;
Described convolution is to be undertaken by the filtering in Fourier space;
From described convolved projection data, extract high pass component;
From described high pass component, extract marginal information;
From described high pass component, remove described marginal information, obtain noise;
From described data for projection, remove described noise, obtain the correction data for projection;
Described correction data for projection is carried out back projection.
2. image reconstructing method according to claim 1 is characterized in that: the extraction of described high pass component is undertaken by following step: to described convolved projection data low-pass filtering, extract low-pass component; From described convolved projection data, remove this low-pass component.
3. image reconstructing method according to claim 1 and 2 is characterized in that: the extraction of described marginal information is to be undertaken by the rim detection computing.
4. X ray CT device has: collecting part, collect data for projection by utilizing the X-ray scanning object; Reconstruction means, is characterized in that by convolution back projection method reconstructed image based on described data for projection:
Described reconstruction means comprises:
Data for projection is carried out convolution, obtain the parts of convolved projection data;
Described convolution is to be undertaken by the filtering in Fourier space;
From described convolved projection data, extract the parts of high pass component;
From described high pass component, extract the parts of marginal information;
From described high pass component, remove described marginal information, obtain the parts of noise;
From described data for projection, remove described noise, obtain the parts of revising data for projection;
Described correction data for projection is carried out the parts of back projection.
5. X ray CT device according to claim 4 is characterized in that: the extraction of described high pass component is undertaken by following step: to described convolved projection data low-pass filtering, extract low-pass component; From described convolved projection data, remove this low-pass component.
6. according to claim 4 or 5 described X ray CT devices, it is characterized in that: the extraction of described marginal information is to be undertaken by the rim detection computing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB200310120365XA CN1305298C (en) | 2003-11-06 | 2003-11-06 | Image reconstructing method and X-ray CT device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB200310120365XA CN1305298C (en) | 2003-11-06 | 2003-11-06 | Image reconstructing method and X-ray CT device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN1614636A CN1614636A (en) | 2005-05-11 |
CN1305298C true CN1305298C (en) | 2007-03-14 |
Family
ID=34761520
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB200310120365XA Expired - Fee Related CN1305298C (en) | 2003-11-06 | 2003-11-06 | Image reconstructing method and X-ray CT device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN1305298C (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103186888B (en) * | 2011-12-30 | 2017-11-21 | Ge医疗***环球技术有限公司 | A kind of method and device of removal CT picture noises |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH07303197A (en) * | 1994-05-10 | 1995-11-14 | Ge Yokogawa Medical Syst Ltd | Image processing method and image processor |
US5473654A (en) * | 1994-06-24 | 1995-12-05 | General Electric Company | Backprojection for x-ray CT system |
US6219441B1 (en) * | 1993-06-22 | 2001-04-17 | General Electric Company | Reconstruction of images from three-dimensional cone beam data |
CN1321884A (en) * | 2001-06-08 | 2001-11-14 | 清华同方核技术股份有限公司 | Taxonomic real-time reconstruction checking method of large-size workpiece, defect |
US6587539B2 (en) * | 2001-03-13 | 2003-07-01 | Shimadzu Corporation | Radiographic apparatus |
-
2003
- 2003-11-06 CN CNB200310120365XA patent/CN1305298C/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6219441B1 (en) * | 1993-06-22 | 2001-04-17 | General Electric Company | Reconstruction of images from three-dimensional cone beam data |
JPH07303197A (en) * | 1994-05-10 | 1995-11-14 | Ge Yokogawa Medical Syst Ltd | Image processing method and image processor |
US5473654A (en) * | 1994-06-24 | 1995-12-05 | General Electric Company | Backprojection for x-ray CT system |
US6587539B2 (en) * | 2001-03-13 | 2003-07-01 | Shimadzu Corporation | Radiographic apparatus |
CN1321884A (en) * | 2001-06-08 | 2001-11-14 | 清华同方核技术股份有限公司 | Taxonomic real-time reconstruction checking method of large-size workpiece, defect |
Also Published As
Publication number | Publication date |
---|---|
CN1614636A (en) | 2005-05-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN1292712C (en) | Method and apparatus for multi-modality imaging | |
US7327822B2 (en) | Methods, apparatus, and software for reconstructing an image | |
CN1289035C (en) | X-ray CT device | |
US7706497B2 (en) | Methods and apparatus for noise estimation for multi-resolution anisotropic diffusion filtering | |
US6574297B2 (en) | System and method for image reconstruction in a cone beam imaging system | |
CN101061504A (en) | Restoration of the nuclear medicine 2D planar image by iterative constrained deconvolution | |
CN1672637A (en) | Method and system for imaging using multiple offset X-ray emission points | |
CN101036164A (en) | Image reconstruction with voxel dependent interpolation | |
US6654440B1 (en) | Methods and apparatus for computed tomography scanning using a two-dimensional radiation source | |
JP2003199737A (en) | Reconstruction method for tomosynthesis | |
US20080253502A1 (en) | Method and Device for the Iterative Reconstruction of Cardiac Images | |
CN1765323A (en) | Within radius image-dependent detector row filtering method, X-ray CT apparatus and program | |
JP6505513B2 (en) | X-ray computed tomography imaging apparatus and medical image processing apparatus | |
CN1675656A (en) | Reconstruction method for tilted-gantry computed tomography | |
CN1903127A (en) | Method for producing computer tomographic image by computer tomographic equipment | |
US8798350B2 (en) | Method and system for reconstruction algorithm in cone beam CT with differentiation in one direction on detector | |
CN1742296A (en) | Voxel-driven spiral reconstruction for cone-beam computer tomography | |
JP2007151806A (en) | X-ray ct imaging method and x-ray ct apparatus | |
US20060262893A1 (en) | Methods and systems to facilitate reducing cone beam artifacts in images | |
CN1294879C (en) | Apparatus and method for x-ray computed tomography | |
JP4474304B2 (en) | Ring artifact removal method and X-ray CT apparatus | |
JP5214110B2 (en) | X-ray CT system | |
CN1305298C (en) | Image reconstructing method and X-ray CT device | |
CN1593343A (en) | Radiation tomographic imaging apparatus and radiation tomographic imaging method, and image producing apparatus | |
JP5461803B2 (en) | X-ray CT system |
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: 20070314 Termination date: 20211106 |
|
CF01 | Termination of patent right due to non-payment of annual fee |