CN104280705B - magnetic resonance image reconstruction method and device based on compressed sensing - Google Patents
magnetic resonance image reconstruction method and device based on compressed sensing Download PDFInfo
- Publication number
- CN104280705B CN104280705B CN201410526251.3A CN201410526251A CN104280705B CN 104280705 B CN104280705 B CN 104280705B CN 201410526251 A CN201410526251 A CN 201410526251A CN 104280705 B CN104280705 B CN 104280705B
- Authority
- CN
- China
- Prior art keywords
- image
- variance
- module
- amplitude
- reconstruction
- 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
- Magnetic Resonance Imaging Apparatus (AREA)
- Image Processing (AREA)
Abstract
The invention provides a magnetic resonance image reconstruction method and device based on compressed sensing. The method includes the steps that under-sampled original K space data are acquired through magnetic resonance scanning; image reconstruction is conducted on the original K space data through a compressed sensing method; after an image reconstruction result tends to converge, disturbance is added to the original K space data, variances of the reconstruction result obtained after disturbance is added are counted, and then a first variogram is acquired; a region of interest in the first variogram is extracted; image mask processing is conducted on the region of interest, and then a second variogram is acquired; the second variogram is added to the image reconstruction result through a non-linear function, and then a reconstructed image is acquired. By the adoption of the method and device, the accuracy of the magnetic resonance image reconstruction result is improved.
Description
Technical field
The present invention relates to mr imaging technique field, particularly relate to a kind of magnetic resonance based on compressed sensing
Image rebuilding method and device.
Background technology
Compressive sensing theory utilizes the openness of signal, and only need to gather a small amount of sample can reconstruct former by high-quality
Beginning data.In recent years, compressed sensing (CS) theory has obtained quick development in magnetic resonance fast imaging
And application, utilize this theory, the k-space adopted can reconstruct original image from owing, thus reduce k-space
Gathering line number, reduce sweep time, reach the purpose of fast imaging.
Traditional MR image reconstruction method based on compressed sensing in image reconstruction process, low contrast
The detailed information of image is easily lost, and can be more serious with the increase information dropout accelerating multiple, impact
The precision of MR image reconstruction result.
Summary of the invention
Based on this, it is necessary to provide a kind of can improve MR image reconstruction result precision based on compressed sensing
MR image reconstruction method and apparatus.
A kind of MR image reconstruction method based on compressed sensing, described method includes:
The raw k-space data of lack sampling are obtained by magnetic resonance imaging；
Utilize compression sensing method that described raw k-space data are carried out image reconstruction；
After the reconstructed results of described image reconstruction tends to convergence, add in described raw k-space data and disturb
Dynamic, and add up add disturbance after the variance of reconstructed results obtain first variance figure；
Extract the area-of-interest in described first variance figure；
Described area-of-interest does image masks process, obtain second variance figure；
Utilize nonlinear function to be joined in described image reconstruction result by second variance figure, obtain rebuilding image.
Wherein in an embodiment, described in described raw k-space data add disturbance, and add up add
After entering disturbance, the variance of reconstructed results obtains the step of first variance figure, including:
In described raw k-space data, randomly select the sampled point of predetermined number, described sampled point is set to
Zero makes raw k-space be converted to new K space data；
Utilize compression sensing method that described new K space data is carried out image reconstruction, obtain reconstructed results figure
Picture；
Repeat the above-mentioned two step of preset times, obtain the reconstruction identical with described preset times quantity
Result images；
Add up the variance in described reconstructed results image, it is thus achieved that first variance figure.
Wherein in an embodiment, the step of the area-of-interest in described extraction described first variance figure,
Including:
Extract the foreground area in first variance figure and background area, using described foreground area as region of interest
Territory.
Wherein in an embodiment, described utilize nonlinear function that second variance figure is joined described image
In reconstructed results, obtain rebuilding the step of image, including:
The amplitude of described image reconstruction result is done normalized, the reconstruction amplitude figure after being processed；
Utilize nonlinear function that each pixel in described reconstruction amplitude figure is done nonlinear transformation, become
Change image；
Described second variance figure is added with described changing image, obtains being added image；
Each pixel in described addition image is done nonlinear inversion transformation, it is thus achieved that rebuild image.
Wherein in an embodiment, the described amplitude to described reconstructed results image does normalized,
The step of the reconstruction amplitude figure after process, including:
To each pixel delivery value in described reconstructed results image, obtain amplitude figure；
Obtain the maximum gradation value in described amplitude figure, by each pixel in described amplitude figure divided by maximum
Gray value, obtains rebuilding amplitude figure.
A kind of MR image reconstruction device based on compressed sensing, described device includes:
Spatial data capture module, for obtaining the raw k-space data of lack sampling by magnetic resonance imaging；
First image reconstruction module, is used for utilizing compressed sensing device that described raw k-space data are carried out figure
As rebuilding；
First variance figure acquisition module, after tending to convergence when the reconstructed results of described image reconstruction, in institute
State in raw k-space data addition disturbance, and add up add disturbance after the variance of reconstructed results obtain first party
Difference figure；
Region of interesting extraction module, for extracting the area-of-interest in described first variance figure；
Second variance figure acquisition module, processes for described area-of-interest does image masks, obtains second
Variogram；
Second image reconstruction module, is used for utilizing nonlinear function that second variance figure joins described image weight
Build in result, obtain rebuilding image.
Wherein in an embodiment, described first variance figure acquisition module includes:
Spatial data more new module, for randomly selecting adopting of predetermined number in described raw k-space data
Sampling point, is set to described sampled point zero and makes raw k-space be converted to new K space data；
3rd image reconstruction module, for utilizing compressed sensing device that described new K space data is carried out figure
As rebuilding, obtain reconstructed results image；
Repeat module, for repeating the above-mentioned two step of preset times, obtain presetting with described
The reconstructed results image that number of times quantity is identical；
Variance statistic module, for adding up the variance in described reconstructed results image, it is thus achieved that first variance figure.
Wherein in an embodiment, described region of interesting extraction module is additionally operable to extract in first variance figure
Foreground area and background area, using described foreground area as area-of-interest.
Wherein in an embodiment, described second image reconstruction module includes:
Normalized module, for the amplitude of described image reconstruction result is done normalized, obtains everywhere
Reconstruction amplitude figure after reason；
Nonlinear transformation module, for utilizing nonlinear function to each pixel in described reconstruction amplitude figure
Do nonlinear transformation, obtain changing image；
Image addition module, for being added with described changing image by described second variance figure, obtains addition figure
Picture；
Inverse transform module, for doing nonlinear inversion transformation to each pixel in described addition image, it is thus achieved that
Rebuild image.
Wherein in an embodiment, described normalized module includes:
Delivery value module, for each pixel delivery value in described reconstructed results image, obtains amplitude
Figure；
Rebuild amplitude figure acquisition module, for obtaining the maximum gradation value in described amplitude figure, by described amplitude
Each pixel in figure, divided by maximum gradation value, obtains rebuilding amplitude figure.
Above-mentioned MR image reconstruction method and apparatus based on compressed sensing, carries out magnetic at compression sensing method
In resonance image process of reconstruction, raw k-space data add disturbance so that reconstructed results occurs trickle
Change, by the variogram of reconstructed results after statistics addition disturbance, utilizes the information in variogram, to reconstruction
The information lost in result is supplemented, and improves the precision of MR image reconstruction result.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of MR image reconstruction method based on compressed sensing in an embodiment；
Fig. 2 is to add disturbance in an embodiment in described raw k-space data, and adds up addition disturbance
The variance of rear reconstructed results obtains the schematic flow sheet of first variance figure step；
Fig. 3 is to utilize nonlinear function to join in reconstructed results by second variance figure in an embodiment,
To the schematic flow sheet rebuilding image step；
Fig. 4 is that in an embodiment, MR image reconstruction method based on compressed sensing carries out image reconstruction mistake
In journey, each step generates the displaying figure of image；
Fig. 5 is the structural representation of MR image reconstruction device based on compressed sensing in an embodiment；
Fig. 6 is the structural representation of first variance figure acquisition module in an embodiment；
Fig. 7 is the structural representation of second variance figure acquisition module in an embodiment；
Fig. 8 is the structural representation of normalized module in an embodiment.
Detailed description of the invention
In order to make the purpose of the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawing and reality
Execute example, the present invention is further elaborated.Only should be appreciated that specific embodiment described herein
Only in order to explain the present invention, it is not intended to limit the present invention.
As it is shown in figure 1, in one embodiment, it is provided that a kind of magnetic resonance image (MRI) weight based on compressed sensing
Construction method, the method comprises the steps:
Step 101, obtains the raw k-space data of lack sampling by magnetic resonance imaging.
Step 102, utilizes compression sensing method that raw k-space data are carried out image reconstruction.
In the present embodiment, compression sensing method carries out image reconstruction equation and is: min | | ψ x | |1,s.t.||Fpx-y||2≤ ε, its
In, Ψ is referred to as fixing sparse transformation, and y is the K spacing wave recorded, FpFor the fourier descriptor of lack sampling,
ε is the parameter relevant with the noise grade of signal.Conventional sparse transformation has a wavelet transformation, principal component analysis,
Finite difference conversion etc..X is reconstruction image to be solved, uses conjugate gradient decent to solve Reconstructed equation,
Obtain reconstructed results x
Step 103, after the reconstructed results of described image reconstruction tends to convergence, at described raw k-space number
According to middle addition disturbance, and add up add disturbance after the variance of reconstructed results obtain first variance figure.
| | the ψ x | | in the present embodiment, by min1,s.t.||Fpx-y||2≤ ε is converted to lagrangian representation:Wherein, λ is regularization coefficient, rule of thumb chooses.Use Conjugate gradient descent
Method solves above formula so that reconstructed results tends to convergence.
In one embodiment, step 103, described raw k-space data add disturbance, and adds up
Add the variance of reconstructed results after disturbance to obtain first variance figure and include:
Step 201, randomly selects the sampled point of predetermined number in raw k-space data, is put by sampled point
It is zero to make raw k-space be converted to new K space data.The span of predetermined number is: be more than
In 1 and less than or equal to 20, preferential, predetermined number is 10.
Step 202, utilizes compression sensing method that new K space data is carried out image reconstruction, is rebuild
Result images.
Step 203, repeats the above-mentioned two step of preset times, obtains identical with preset times quantity
Reconstructed results image.The times N of repeated execution of steps 201 and step 202 time obtains N number of reconstructed results figure
As { x1,x2…xN, wherein N is more than 10, it is preferred that preset times N is 20.
Step 204, the variance in statistics reconstructed results image, it is thus achieved that first variance figure.
In the present embodiment, it is thus achieved that N number of reconstructed results image there is identical locus, i.e. reconstructed results figure
Any one spatial point of picture has N number of reconstructed value, to each spatial dot statistics variance rebuild in result images,
After obtain first variance figure.
Step 104, extracts the area-of-interest in first variance figure.
Concrete, extract the foreground area in first variance figure and background area, foreground area is emerging as sense
Interest region.In one embodiment, use threshold method by the foreground area in first variance figure and background area
Make a distinction, by selecting a threshold value that the pixel being more than threshold value in image is labeled as foreground area, little
Pixel in threshold value is labeled as background area.
Step 105, does image masks and processes, obtain second variance figure area-of-interest.
Step 106, utilizes nonlinear function to join in reconstructed results by second variance figure, obtains rebuilding image.
In the present embodiment, the low contrast lost during recovering MR image reconstruction by second variance figure
Information, utilizes nonlinear function to be joined in reconstructed results by second variance figure and carries out extensive to low contrast information
Multiple.Above-mentioned MR image reconstruction method based on compressed sensing, carries out magnetic resonance figure at compression sensing method
As, in process of reconstruction, raw k-space data adding disturbance so that reconstructed results generation slight change,
By the variogram of reconstructed results after statistics addition disturbance, utilize the information in variogram, in reconstructed results
The information lost is supplemented, and improves the precision of MR image reconstruction result.
As it is shown on figure 3, in one embodiment, step 106, utilize nonlinear function to be added by second variance figure
Enter in reconstructed results, obtain rebuilding image and include:
Step 301, does normalized to the amplitude of image reconstruction result, the reconstruction amplitude figure after being processed.
Concrete, to each pixel delivery value rebuild in result images, obtain amplitude figure；Obtain in amplitude figure
Maximum gradation value, by each pixel in amplitude figure divided by maximum gradation value, obtain rebuild amplitude figure.
Step 302, utilizes nonlinear function that each pixel rebuild in amplitude figure is done nonlinear transformation,
To changing image.In one embodiment, nonlinear function is power function, f (x)=xa, x is for rebuilding amplitude
Pixel in figure, a is default numerical value, it is preferred that a=5.
Step 303, is added second variance figure with changing image, obtains being added image.By in second variance figure
Pixel point value carry out being added with the pixel point value of changing image obtain be added image.
Step 304, does nonlinear inversion transformation to each pixel being added in image, it is thus achieved that rebuild image.
By nonlinear function f (x)=x1/aThe each pixel being added in image is done nonlinear inversion transformation.Its
In, x is the pixel being added in image, and a is default numerical value, it is preferred that a=5.
In one embodiment, it is that MR image reconstruction method based on compressed sensing carries out image such as Fig. 4
In process of reconstruction, each step generates the displaying figure of image.As shown in Figure 4,4a is original image, and 4b is logical
The image that overcompression cognitive method obtains after carrying out image reconstruction, wherein 4a white circle marked region 40 He
The low contrast information of 42 disappears in 4b after compression sensing method is rebuild；
4c is the first variance figure that the variance in statistics 4b obtains, and can be seen that in figure and contain 4a from 4c
In 40 and 42 part low contrast information.After 4d is for carrying out image masks process to area-of-interest in 4c
Obtain second variance figure.4e is the reconstruction amplitude figure obtained after the amplitude to 4b does normalized.4f is right
The changing image that each pixel in 4e obtains after doing nonlinear transformation.After 4g is 4f Yu 4d image addition
The addition image obtained.4h be 4g is done inverse transformation after the reconstruction image that obtains, it can clearly be seen that in 4b
The low contrast information 40 and 42 lost occurs in that again in 4h, and it is based on compressed sensing that the present invention provides
Low contrast information can be recovered by MR image reconstruction method, improves the essence of MR image reconstruction
Degree.
As it is shown in figure 5, in one embodiment, it is provided that a kind of based on compressed sensing magnetic resonance image (MRI) weight
Building device, this device includes:
Spatial data capture module 50, for obtaining the raw k-space data of lack sampling by magnetic resonance imaging.
First image reconstruction module 51, is used for utilizing compressed sensing device that raw k-space data are carried out image
Rebuild.
First variance figure acquisition module 52, after tending to convergence when the reconstructed results of image reconstruction, original
K space data adds disturbance, and adds up and add the variance of reconstructed results after disturbance and obtain first variance figure.
Region of interesting extraction module 53, for extracting the area-of-interest in first variance figure.
In one embodiment, region of interesting extraction module 53 is additionally operable to extract the prospect in first variance figure
Region and background area, using foreground area as area-of-interest.
Second variance figure acquisition module 54, processes for area-of-interest does image masks, obtains second party
Difference figure.
Second image reconstruction module 55, is used for utilizing nonlinear function that second variance figure is joined image reconstruction
In result, obtain rebuilding image.
As shown in Figure 6, in one embodiment, first variance figure acquisition module 52 includes:
Spatial data more new module 520, for randomly selecting adopting of predetermined number in raw k-space data
Sampling point, is set to sampled point zero and makes raw k-space be converted to new K space data.
3rd image reconstruction module 521, for utilizing compressed sensing device that new K space data is carried out figure
As rebuilding, obtain reconstructed results image.
Repeat module 523, for repeating the above-mentioned two step of preset times, obtain and default time
The reconstructed results image that number quantity is identical.
Variance statistic module 524, for adding up the variance in reconstructed results image, it is thus achieved that first variance figure.
As it is shown in fig. 7, in one embodiment, the second image reconstruction module 55 includes:
Normalized module 550, for the amplitude of image reconstruction result is done normalized, is processed
After reconstruction amplitude figure.
Nonlinear transformation module 551, for utilizing nonlinear function to do each pixel rebuild in amplitude figure
Nonlinear transformation, obtains changing image.
Image addition module 552, for being added with changing image by second variance figure, obtains being added image.
Inverse transform module 553, for doing nonlinear inversion transformation to each pixel being added in image, it is thus achieved that weight
Build image.
As shown in Figure 8, in one embodiment, normalized module 550 includes:
Delivery value module 5500, for each pixel delivery value rebuild in result images, obtaining amplitude
Figure.
Rebuild amplitude figure acquisition module 5501, for obtaining the maximum gradation value in amplitude figure, by amplitude figure
Each pixel divided by maximum gradation value, obtain rebuild amplitude figure.
Embodiment described above only have expressed the several embodiments of the present invention, and it describes more concrete and detailed,
But therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that, for this area
Those of ordinary skill for, without departing from the inventive concept of the premise, it is also possible to make some deformation and
Improving, these broadly fall into protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be with appended
Claim is as the criterion.
Claims (10)
1. a MR image reconstruction method based on compressed sensing, described method includes:
The raw k-space data of lack sampling are obtained by magnetic resonance imaging；
Utilize compression sensing method that described raw k-space data are carried out image reconstruction；
After the reconstructed results of described image reconstruction tends to convergence, add in described raw k-space data and disturb
Dynamic, and add up add disturbance after the variance of reconstructed results obtain first variance figure；
Extract the area-of-interest in described first variance figure；
Described area-of-interest does image masks process, obtain second variance figure；
Utilize nonlinear function to be joined in the reconstructed results of described image reconstruction by second variance figure, obtain weight
Build image.
Method the most according to claim 1, it is characterised in that described in described raw k-space data
Middle addition disturbance, and add up and add the variance of reconstructed results after disturbance and obtain the step of first variance figure, including:
In described raw k-space data, randomly select the sampled point of predetermined number, described sampled point is set to
Zero makes raw k-space be converted to new K space data；
Utilize compression sensing method that described new K space data is carried out image reconstruction, obtain reconstructed results figure
Picture；
Repeat the above-mentioned two step of preset times, obtain the reconstruction identical with described preset times quantity
Result images；
Add up the variance in described reconstructed results image, it is thus achieved that first variance figure.
Method the most according to claim 1, it is characterised in that in described extraction described first variance figure
The step of area-of-interest, including:
Extract the foreground area in first variance figure and background area, using described foreground area as region of interest
Territory.
Method the most according to claim 1, it is characterised in that described utilize nonlinear function by second
Variogram joins in the reconstructed results of described image reconstruction, obtains rebuilding the step of image, including:
The amplitude of the reconstructed results of described image reconstruction is done normalized, the reconstruction amplitude after being processed
Figure；
Utilize nonlinear function that each pixel in described reconstruction amplitude figure is done nonlinear transformation, become
Change image；
Described second variance figure is added with described changing image, obtains being added image；
Each pixel in described addition image is done nonlinear inversion transformation, it is thus achieved that rebuild image.
Method the most according to claim 4, it is characterised in that the described weight to described image reconstruction
The amplitude building result does normalized, the step of the reconstruction amplitude figure after being processed, including:
To each pixel delivery value in the reconstructed results of described image reconstruction, obtain amplitude figure；
Obtain the maximum gradation value in described amplitude figure, by each pixel in described amplitude figure divided by maximum
Gray value, obtains rebuilding amplitude figure.
6. a MR image reconstruction device based on compressed sensing, it is characterised in that described device includes:
Spatial data capture module, for obtaining the raw k-space data of lack sampling by magnetic resonance imaging；
First image reconstruction module, is used for utilizing compressed sensing device that described raw k-space data are carried out figure
As rebuilding；
First variance figure acquisition module, after tending to convergence when the reconstructed results of described image reconstruction, in institute
State in raw k-space data addition disturbance, and add up add disturbance after the variance of reconstructed results obtain first party
Difference figure；
Region of interesting extraction module, for extracting the area-of-interest in described first variance figure；
Second variance figure acquisition module, processes for described area-of-interest does image masks, obtains second
Variogram；
Second image reconstruction module, is used for utilizing nonlinear function that second variance figure joins described image weight
In the reconstructed results built, obtain rebuilding image.
Device the most according to claim 6, it is characterised in that described first variance figure acquisition module bag
Include:
Spatial data more new module, for randomly selecting adopting of predetermined number in described raw k-space data
Sampling point, is set to described sampled point zero and makes raw k-space be converted to new K space data；
3rd image reconstruction module, for utilizing compressed sensing device that described new K space data is carried out figure
As rebuilding, obtain reconstructed results image；
Repeat module, for repeating above-mentioned spatial data more new module and the 3rd figure of preset times
As rebuilding the step that module performs, obtain the reconstructed results image identical with described preset times quantity；
Variance statistic module, for adding up the variance in described reconstructed results image, it is thus achieved that first variance figure.
Device the most according to claim 6, it is characterised in that described region of interesting extraction module is also
For extracting the foreground area in first variance figure and background area, using described foreground area as region of interest
Territory.
Device the most according to claim 6, it is characterised in that described second image reconstruction module obtains
Module includes:
Normalized module, for the amplitude of the reconstructed results of described image reconstruction is done normalized,
Reconstruction amplitude figure after being processed；
Nonlinear transformation module, for utilizing nonlinear function to each pixel in described reconstruction amplitude figure
Do nonlinear transformation, obtain changing image；
Image addition module, for being added with described changing image by described second variance figure, obtains addition figure
Picture；
Inverse transform module, for doing nonlinear inversion transformation to each pixel in described addition image, it is thus achieved that
Rebuild image.
Device the most according to claim 9, it is characterised in that described normalized module includes:
Delivery value module, for each pixel delivery value in the reconstructed results of described image reconstruction, obtains
To amplitude figure；
Rebuild amplitude figure acquisition module, for obtaining the maximum gradation value in described amplitude figure, by described amplitude
Each pixel in figure, divided by maximum gradation value, obtains rebuilding amplitude figure.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410526251.3A CN104280705B (en) | 2014-09-30 | 2014-09-30 | magnetic resonance image reconstruction method and device based on compressed sensing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410526251.3A CN104280705B (en) | 2014-09-30 | 2014-09-30 | magnetic resonance image reconstruction method and device based on compressed sensing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104280705A CN104280705A (en) | 2015-01-14 |
| CN104280705B true CN104280705B (en) | 2017-01-11 |
Family
ID=52255774
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410526251.3A Active CN104280705B (en) | 2014-09-30 | 2014-09-30 | magnetic resonance image reconstruction method and device based on compressed sensing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104280705B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106546454A (en) * | 2015-09-16 | 2017-03-29 | 中国辐射防护研究院 | The Technique of Electron Paramagnetic Resonance is used for the method for reducing background when irradiation processing is detected |
| CN106023116B (en) * | 2016-05-30 | 2018-09-28 | 中国科学院深圳先进技术研究院 | A kind of compressed sensing image rebuilding method and device based on block Weighted Constraint |
| CN108280862B (en) * | 2018-01-31 | 2021-07-23 | 安徽福晴医疗科技有限公司 | Reconstruction method and device of magnetic resonance image |
| CN109949257B (en) * | 2019-03-06 | 2021-09-10 | 西安电子科技大学 | Region-of-interest compressed sensing image reconstruction method based on deep learning |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102309328A (en) * | 2011-10-19 | 2012-01-11 | 中国科学院深圳先进技术研究院 | Diffusion-tensor imaging method and system |
| CN103033783A (en) * | 2012-12-10 | 2013-04-10 | 深圳先进技术研究院 | Magnetic resonance imaging fast reconstruction system and method thereof |
| CN103048632A (en) * | 2012-12-19 | 2013-04-17 | 北京大学 | Compressive sensing-based quick fine angular resolution diffusion imaging method |
| CN103064046A (en) * | 2012-12-25 | 2013-04-24 | 深圳先进技术研究院 | Image processing method based on sparse sampling magnetic resonance imaging |
| CN103142228A (en) * | 2012-12-14 | 2013-06-12 | 中国科学院深圳先进技术研究院 | Compressed sensing magnetic resonance fast imaging method |
| CN103356193A (en) * | 2013-07-19 | 2013-10-23 | 哈尔滨工业大学深圳研究生院 | Method and system for performing rapid diffusion tensor imaging under compression sensing framework |
| EP2660618A1 (en) * | 2012-05-04 | 2013-11-06 | Esaote S.p.A. | Biomedical image reconstruction method |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8879852B2 (en) * | 2010-11-10 | 2014-11-04 | Siemens Aktiengesellschaft | Non-contrast-enhanced 4D MRA using compressed sensing reconstruction |
| WO2013188957A1 (en) * | 2012-06-18 | 2013-12-27 | University Health Network | Method and system for compressed sensing image reconstruction |
| US9081074B2 (en) * | 2012-08-21 | 2015-07-14 | Beth Israel Deaconess Medical Center, Inc. (Bidmc, Inc.) | Method and apparatus for accelerated phase contrast magnetic resonance angiography and blood flow imaging |
-
2014
- 2014-09-30 CN CN201410526251.3A patent/CN104280705B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102309328A (en) * | 2011-10-19 | 2012-01-11 | 中国科学院深圳先进技术研究院 | Diffusion-tensor imaging method and system |
| EP2660618A1 (en) * | 2012-05-04 | 2013-11-06 | Esaote S.p.A. | Biomedical image reconstruction method |
| CN103033783A (en) * | 2012-12-10 | 2013-04-10 | 深圳先进技术研究院 | Magnetic resonance imaging fast reconstruction system and method thereof |
| CN103142228A (en) * | 2012-12-14 | 2013-06-12 | 中国科学院深圳先进技术研究院 | Compressed sensing magnetic resonance fast imaging method |
| CN103048632A (en) * | 2012-12-19 | 2013-04-17 | 北京大学 | Compressive sensing-based quick fine angular resolution diffusion imaging method |
| CN103064046A (en) * | 2012-12-25 | 2013-04-24 | 深圳先进技术研究院 | Image processing method based on sparse sampling magnetic resonance imaging |
| CN103356193A (en) * | 2013-07-19 | 2013-10-23 | 哈尔滨工业大学深圳研究生院 | Method and system for performing rapid diffusion tensor imaging under compression sensing framework |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104280705A (en) | 2015-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109557489B (en) | Magnetic resonance imaging method and device | |
| CN102708576B (en) | Method for reconstructing partitioned images by compressive sensing on the basis of structural dictionaries | |
| CN105513026B (en) | One kind being based on the non local similar compressed sensing reconstructing method of image | |
| US8587307B2 (en) | Systems and methods for accelerating the acquisition and reconstruction of magnetic resonance images with randomly undersampled and uniformly undersampled data | |
| CN104280705B (en) | magnetic resonance image reconstruction method and device based on compressed sensing | |
| CN102332153B (en) | Kernel regression-based image compression sensing reconstruction method | |
| CN104569880B (en) | A kind of magnetic resonance fast imaging method and system | |
| CN103854268B (en) | The Image Super-resolution Reconstruction method returned based on multinuclear Gaussian process | |
| CN104217448B (en) | Magnetic resonance fast imaging method and system based on iterative characteristic amendment | |
| RU2013103063A (en) | DYNAMIC CONTRAST IMPROVED MR VISUALIZATION WITH RECONSTRUCTION OF COMPRESSED MEASUREMENT | |
| CN102217934A (en) | Magnetic resonance imaging method and system | |
| US10867388B2 (en) | System and method for multi-contrast magnetic resonance imaging | |
| CN103654789A (en) | Fast magnetic resonance parametric imaging method and system | |
| CN105046672A (en) | Method for image super-resolution reconstruction | |
| CN105232045A (en) | Single-scanning quantitative magnetic resonance diffusion imaging method based on dual echoes | |
| CN103116873B (en) | Image denoising method | |
| CN102024266A (en) | Image structure model-based compressed sensing image reconstruction method | |
| CN107833182A (en) | The infrared image super resolution ratio reconstruction method of feature based extraction | |
| Zou et al. | Joint wavelet sub-bands guided network for single image super-resolution | |
| Wang et al. | Iterative feature refinement for accurate undersampled MR image reconstruction | |
| Hui et al. | Multi-channel adaptive partitioning network for block-based image compressive sensing | |
| CN112508786B (en) | Satellite image-oriented arbitrary-scale super-resolution reconstruction method and system | |
| CN105241813B (en) | Compression sampling opto-acoustic microscopic imaging method and device | |
| CN117333365A (en) | Image super-resolution method based on hybrid transducer super-resolution network | |
| CN111681297A (en) | Image reconstruction method, computer device, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230223 Address after: 201807 2258 Chengbei Road, Jiading District, Shanghai Patentee after: Shanghai Lianying Medical Technology Co.,Ltd. Address before: 1068 No. 518055 Guangdong city in Shenzhen Province, Nanshan District City Xili University School Avenue Patentee before: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY |
|
| TR01 | Transfer of patent right |