CN106491131B - A kind of dynamic imaging methods and device of magnetic resonance - Google Patents
A kind of dynamic imaging methods and device of magnetic resonance Download PDFInfo
- Publication number
- CN106491131B CN106491131B CN201611260606.4A CN201611260606A CN106491131B CN 106491131 B CN106491131 B CN 106491131B CN 201611260606 A CN201611260606 A CN 201611260606A CN 106491131 B CN106491131 B CN 106491131B
- Authority
- CN
- China
- Prior art keywords
- data
- lack sampling
- space
- sampling
- acs
- 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
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
- A61B5/055—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
Landscapes
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- High Energy & Nuclear Physics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Biophysics (AREA)
- Heart & Thoracic Surgery (AREA)
- Radiology & Medical Imaging (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- Signal Processing (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- General Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
A kind of dynamic imaging methods of magnetic resonance include: to carry out lack sampling to dynamic data by TGRAPPA technology；To the obtained data of lack sampling, sample mode is accelerated to carry out second of lack sampling using variable density；Second of undersampled signal is rebuild；The K space data obtained after reconstruction is carried out to simulate parallel lack sampling by TGRAPPA mode, according to the fully sampled automatic calibration data ACS of the data configuration of lack sampling；The automatic calibration data ACS is mapped into space more higher than the dimension of current spatial, calculates the weight coefficient of each coil in new space；The K space data of parallel lack sampling is simulated according to weight coefficient filling, second of reconstruction generates final image.This method does not need the automatic calibration signal for additionally acquiring K space center, and under higher accelerations multiple, can preferably inhibit in parallel imaging as accelerate multiple to become larger and caused by noise amplify, the acquisition higher image of signal-to-noise ratio.
Description
Technical field
The invention belongs to magnetic resonance arts more particularly to the dynamic imaging methods and device of a kind of magnetic resonance.
Background technique
Magnetic resonance is a kind of technology that tissue is imaged using magnetostatic field and RF magnetic field.Magnetic resonance not only mentions
Contrast in tissue abundant is supplied, and harmless, therefore has become a kind of strong tool of medical clinic applications.
Currently, the dynamic imaging speed of magnetic resonance is always the big bottleneck for restricting mr techniques fast development slowly.Such as
Where guarantee under the premise of clinically-acceptable image quality, scanning speed is improved, to reduce sweep time, it appears especially
It is important.
For in magnetic resonance dynamic imaging technology, commercial fast imaging techniques are mainly parallel imaging at present, including it is dynamic
The automatic calibrated section parallel acquisition (English abbreviation TGRAPPA) of state broad sense is compiled using the adaptive susceptibility of termporal filter
Code (English abbreviation TSENSE), automatic calibrated section parallel acquisition (the English abbreviation Nonlinear of Nonlinear Generalized
GRAPPA) etc., such method requires the spatial information of receiving coil, to fill the K space data for owing to adopt.
When accelerating multiple to accelerate image imaging by improving, TGRAPPA method will lead to the increase of aliasing artifact, and
With the raising for accelerating multiple, picture noise can also amplify；The factors such as the inhomogeneities due to magnetic field, the sensitivity of TSENSE method
Degree mapping (Sensitivity Map) is difficult to be precisely calculated out；For Nonlinear GRAPPA method, there is presently no
For dynamic imaging.Therefore, current dynamic imaging techniques cannot effectively meet wanting for dynamic imaging speed and image quality
It asks.
Summary of the invention
The purpose of the present invention is to provide a kind of dynamic imaging methods of magnetic resonance, to solve existing dynamic imaging techniques
The problem of requirement of dynamic imaging speed and image quality cannot effectively be met.
In a first aspect, the embodiment of the invention provides a kind of dynamic imaging methods of magnetic resonance, which comprises
By the automatic calibrated section parallel acquisition TGRAPPA technology of DYNAMIC GENERALIZED, dynamic data is carried out owing to adopt for the first time
Sample；
To the obtained data of first time lack sampling, sample mode is accelerated to be owed to adopt for the second time using variable density
Sample；
Second of undersampled signal is rebuild according to compression sensing method, obtains the figure with volume pleat artifact
Picture；
The K space data obtained after reconstruction is carried out to simulate parallel lack sampling by TGRAPPA mode, according to the number of lack sampling
According to the fully sampled automatic calibration data ACS of construction；
The automatic calibration data ACS is mapped into space more higher than the dimension of current spatial, is calculated in new space
The weight coefficient of each coil out；
The K space data of parallel lack sampling is simulated according to weight coefficient filling, second of reconstruction generates final image.
With reference to first aspect, described to pass through the automatic school of DYNAMIC GENERALIZED in the first possible implementation of first aspect
Quasi- part parallel acquires TGRAPPA technology, carries out first time lack sampling to dynamic data specifically:
It is R according to presetting drop to adopt rate1, N is acquired altogethertFrame data, to the t on time orientationi(i=1,2......,
Nt) frame data, frequency coding direction is fully sampled, and phase-encoding direction is since first line, every (R1- 1) line acquires one
Line, and the n-th R1+ R frame data are acquired since the R bars line, until NtAll acquisition finishes frame data, wherein 1≤R≤R1,
With reference to first aspect, described to the first time lack sampling in second of possible implementation of first aspect
Obtained data accelerate sample mode to carry out second of lack sampling step using variable density specifically:
Each frame data obtained to first time lack sampling, phase-encoding direction fully sampled by frequency coding direction,
According to the stochastical sampling mode of compressed sensing, variable density sampling is carried out.
With reference to first aspect, described according to compression sensing method pair in the third possible implementation of first aspect
Second of undersampled signal is rebuild, and the image step with volume pleat artifact is obtained specifically:
It according to compression sensing method, is rebuild according to lack sampling data of the Reconstructed equation F ρ=y to channel, obtains band volume
The image of pleat artifact, in which: F indicates that Fourier owes to adopt operator, and ρ is the image to be rebuild, and y is the K space data of lack sampling；
It is solved by convex optimization method, so that 1 Norm minimum of each coil channel, obtains the solution item of Reconstructed equation
Part:The image with volume pleat artifact of each coil is obtained according to the solving condition,
Wherein, | | ρ | |1It is 1 norm, | | ρ | |2It is 2 norms, y is the K space data for owing to adopt, and ε is less than the threshold parameter of level of noise.
With reference to first aspect, in the 4th kind of possible implementation of first aspect, the described pair of space K obtained after rebuilding
Data are carried out simulating parallel lack sampling by TGRAPPA mode, according to the fully sampled automatic calibration data of the data configuration of lack sampling
ACS step specifically:
The K space data obtained after reconstruction is carried out to simulate parallel lack sampling by TGRAPPA mode, lack sampling is obtained
NtFrame data, by each frame data t on time orientationi(i=1,2......, Nt) be averaging after addition again, it is calculated complete
The ACS data adopted.
With reference to first aspect, described by the automatic calibration data in the 5th kind of possible implementation of first aspect
ACS maps to space more higher than the dimension of current spatial, calculates the weight coefficient step packet of each coil in new space
It includes:
The automatic calibration data ACS is mapped into space more higher than the dimension of current spatial according to mapping kernel function K,
It solves to obtain weight coefficient according to weight coefficient calculation formula Aw=b, in which:
The mapping function K is defined as: K (at,i,at,j)=< Ψ (at,i),Ψ(at,j) >, (t=1,2......, Nt), <
> indicate inner product, at,iIndicate t (t=1,2......, Nt) any two element in the corresponding matrix A of frame, matrix A is by ACS
The point composition that all lack samplings obtain in data, w indicate that weight coefficient, b indicate the point not sampled in ACS, and the size of A is
M × N, M indicate to remove the total size of the automatic calibration data ACS on boundary, consecutive points in all coils used in N expression reconstruction
Quantity.
The 5th kind of possible implementation with reference to first aspect, in the 6th kind of possible implementation of first aspect, institute
State mapping function K is indicated using Polynomial kernel function, and: K (at,i,at,j)=(β aT t,iat,j+c)d, (t=1,2......,
Nt), in which: β and c is scalar, and the order of d representative polynomial kernel function solves and obtains the weight coefficient are as follows:
W=(ΨH(A)Ψ(A))-1ΨH(A) b, wherein Ψ H (A) indicates the conjugate transposition of Ψ (A), (ΨH(A)Ψ(A)
)-1Indicate (ΨH(A) Ψ (A)) it is inverse.
Second aspect, the embodiment of the invention provides a kind of dynamic imaging device of magnetic resonance, described device includes:
First lack sampling unit, for passing through the automatic calibrated section parallel acquisition TGRAPPA technology of DYNAMIC GENERALIZED, to dynamic
Data carry out first time lack sampling；
Second lack sampling unit, for accelerating to sample using variable density to the obtained data of first time lack sampling
Mode carries out second of lack sampling；
First reconstruction unit is obtained for being rebuild according to compression sensing method to second of undersampled signal
Image with volume pleat artifact；
Data configuration unit is owed to adopt parallel for simulate by TGRAPPA mode to the K space data obtained after reconstruction
Sample, according to the fully sampled automatic calibration data ACS of the data configuration of lack sampling；
Weight-coefficient calculating unit is higher than the dimension of current spatial for mapping to the automatic calibration data ACS
Space, calculate the weight coefficient of each coil in new space；
Second reconstruction unit, for simulating the K space data of parallel lack sampling according to weight coefficient filling, for the second time
It rebuilds and generates final image.
In conjunction with second aspect, in the first possible implementation of second aspect, first sampling unit is specifically used
In:
It is R according to presetting drop to adopt rate1, N is acquired altogethertFrame data, to the t on time orientationi(i=1,2......,
Nt) frame data, frequency coding direction is fully sampled, and phase-encoding direction is since first line, every (R1- 1) line acquires one
Line, and the n-th R1+ R frame data are acquired since the R bars line, until NtAll acquisition finishes frame data, wherein 1≤R≤R1,
In conjunction with second aspect, in second of possible implementation of second aspect, second sampling unit is specifically used
In:
Each frame data obtained to first time lack sampling, phase-encoding direction fully sampled by frequency coding direction,
It is theoretical according to the stochastical sampling of compressed sensing, carry out variable density sampling.
After the present invention carries out parallel first time lack sampling to frame image, then the data after sampling are accelerated by variable density
Second of lack sampling, then secondary undersampled signal is rebuild, obtain with volume pleat artifact image, after reconstruction
K space data is carried out simulating parallel lack sampling by TGRAPPA mode, constructs fully sampled automatic calibration data ACS, by described in certainly
Dynamic calibration data ACS maps to space more higher than the dimension of current spatial, calculates the weight of each coil in new space
Coefficient, the K space data of parallel lack sampling is simulated according to weight coefficient filling, and second of reconstruction generates final image.This
Method does not need the automatic calibration signal for additionally acquiring K space center, and under higher speed multiple, can preferably press down
In parallel imaging processed as accelerate multiple become larger and caused by noise amplify, obtain the higher image of signal-to-noise ratio.
Detailed description of the invention
Fig. 1 is the implementation flow chart of the dynamic imaging methods of magnetic resonance provided in an embodiment of the present invention；
Fig. 2 is the structural schematic diagram of the dynamic imaging device of magnetic resonance provided in an embodiment of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
The embodiment of the present invention is designed to provide a kind of MR imaging method, to solve fast short-term training in the prior art
As the parallel imaging in technology, including when the automatic calibrated section parallel acquisition (English abbreviation TGRAPPA) of DYNAMIC GENERALIZED, utilization
Between filter adaptive susceptibility encode (English abbreviation TSENSE), the automatic calibrated section parallel acquisition of Nonlinear Generalized
(English abbreviation is Nonlinear GRAPPA) etc., such method requires the spatial information of receiving coil, to fill the K for owing to adopt
Spatial data.When accelerating multiple to accelerate image imaging by improving, TGRAPPA method will lead to the increase of aliasing artifact, and
And with the raising for accelerating multiple, picture noise can also amplify；The factors such as the inhomogeneities due to magnetic field, TSENSE method it is quick
Sensitivity mapping (Sensitivity Map) is difficult to be precisely calculated out；For Nonlinear GRAPPA method, do not have also at present
Have for dynamic imaging.Therefore, current dynamic imaging techniques cannot effectively meet dynamic imaging speed and image quality
It is required that.The present invention will be further described below with reference to the drawings.
Fig. 1 shows the implementation process of the dynamic imaging methods of the magnetic resonance of first embodiment of the invention offer, is described in detail such as
Under:
In step s101, by the automatic calibrated section parallel acquisition TGRAPPA technology of DYNAMIC GENERALIZED, to dynamic data into
Row first time lack sampling.Specifically,
It is described by the automatic calibrated section parallel acquisition TGRAPPA technology of DYNAMIC GENERALIZED, dynamic data is carried out for the first time
Lack sampling step, is specifically as follows:
It is R according to presetting drop to adopt rate1, N is acquired altogethertFrame data, to the t on time orientationi(i=1,2......,
Nt) frame data, frequency coding direction is fully sampled, and phase-encoding direction is since first line, every (R1- 1) line acquires one
Line, and the n-th R1+ R frame data are acquired since the R bars line, until NtAll acquisition finishes frame data, wherein 1≤R≤R1,
In step s 102, to the obtained data of first time lack sampling, using variable density accelerate sample mode into
Second of lack sampling of row；
To the data of lack sampling are acquired in step S101, phase-encoding direction fully sampled in frequency coding direction
It is acquired for variable density, and the acquisition of phase code follows the stochastical sampling theory of compressed sensing.If the drop that variable density drop is adopted adopts rate
For R2, then total acceleration multiple is R1×R2。
In step s 103, second of undersampled signal is rebuild according to compression sensing method, is had
Roll up the image of pleat artifact.
Firstly, being based on compressive sensing theory, the data of the lack sampling all to each channel are rebuild, and obtain band volume pleat
It is as follows to solve equation for the image of artifact:
F ρ=y [1]
Wherein F indicates that Fourier owes to adopt operator, and ρ is the image to be rebuild, and y is the K space data for owing to adopt.In magnetic resonance
In dynamic imaging, F is actually by along phase code KyLack sampling Fourier's operator F in directionuyAnd along Fu in time t direction
Leaf operator FtIt codetermines, i.e. F=FuyFt。
[1] formula can be solved by convex optimization method, so that 1 Norm minimum of each coil channel, to obtain each line
The image with volume pleat artifact of circle, as shown in [2] formula:
Wherein, | | ρ | |1It is 1 norm, | | ρ | |2It is 2 norms.Y is the K space data for owing to adopt, and ε is less than level of noise
Threshold parameter.
In step S104, the K space data obtained after reconstruction is carried out to simulate parallel lack sampling by TGRAPPA mode,
According to the fully sampled automatic calibration data ACS of the data configuration of lack sampling.
The parallel lack sampling of simulation, can be further to the image after reconstruction with the sampling step in simulation steps S101
Carry out lack sampling operation.Specifically: by the automatic calibrated section parallel acquisition TGRAPPA technology of DYNAMIC GENERALIZED, to reconstruction image
K space data carry out lack sampling.
The N that lack sampling obtains is carried out according to the image of reconstructiontFrame data construct automatic calibration data ACS data, and method is such as
Under: by each frame data t on time orientationi(i=1,2......, Nt) be added, then be averaging, can acquire one it is fully sampled
ACS data.
In step s105, the automatic calibration data ACS is mapped into space more higher than the dimension of current spatial,
New space calculates the weight coefficient of each coil.
According to each frame t (t=1,2......, the N in the data for carrying out lack sampling in step S104 to reconstruction imageT)
The corresponding deficient K space data adopted and fully sampled automatic calibration data ACS, by the automatic calibration data ACS data
It is mapped to the space of a more higher-dimension, and calculates the weight coefficient of each coil in this new space, and fills and owes to adopt
The K space data of sample, the final image obtained without rolling up pleat artifact, detailed process is as follows:
In GRAPPA method, we mainly calculate the weight coefficient of each coil by solving following formula,
Aw=b [3]
Wherein matrix A is made of the point of all lack samplings in ACS data, and w indicates that weight coefficient, b indicate automatic calibration number
According to the point that do not adopted in ACS.The size of A is M × N, and M indicates to remove the total size of the ACS data on boundary, and N is indicated in reconstruction
The quantity of consecutive points in all coils used.Weight coefficient w can be found out by solving equation [3].
In the present invention, when ACS data are mapped to the space of a more higher-dimension, if the kernel function of mapping is K, it is as follows to define K:
K(at,i,at,j)=< Ψ (at,i),Ψ(at,j) >, (t=1,2......, Nt) [4]
<>indicates inner product, at,iIndicate t (t=1,2......, Nt) any two element in the corresponding matrix A of frame, K
Can be there are many representation method, such as Polynomial kernel function and gaussian kernel function.When using Polynomial kernel function, the kernel function
It can indicate are as follows:
K(at,i,at,j)=(β aT t,iat,j+c)d, (t=1,2......, Nt) [5]
The advantages of wherein β and c is scalar, the order of d representative polynomial kernel function, Polynomial kernel function is in kernel function
Ψ(at,i) there is deterministic expression, d=2 is taken, then
Wherein at,1,...at,NIt is t (t=1,2......, Nt) matrix A corresponding to frame element, it is multiple due to calculating
Miscellaneous the reason of spending, we have taken Ψ (at,i) in a part, i.e.,
(at,i,at,j) it is in the space K of t frame along kthxThe most adjacent point in direction, (at,p,at,q) it is secondary adjacent point.
As β=1, c=1,
Under the mapping of above-mentioned kernel function, formula [3] just becomes Ψ (A) w=b [7]
I.e. to each frame t (t=1,2......, Nt), Ψ (A)=[Ψ (a1),...Ψ(ai)...,Ψ(aM)]T,aiIt is
[3] in formula matrix A a line, the size of Ψ (A) is M × Nk, NkIt is the dimension in new space, and is usually significantly larger than former space
Dimension N.Formula [7] can be solved by least square method come calculated result are as follows:
W=(ΨH(A)Ψ(A))-1ΨH(A)b [8]
In step s 106, the K space data of parallel lack sampling, second of reconstruction are simulated according to weight coefficient filling
Generate final image.
After the weight coefficient w of each frame is solved, as traditional GRAPPA method, using w and to reconstruction image
The data of lack sampling can reconstruct final image.
After this method carries out parallel first time lack sampling to frame image, then the data after sampling are accelerated by variable density
Second of lack sampling, then secondary undersampled signal is rebuild, obtain with volume pleat artifact image, after reconstruction
K space data carries out simulating parallel lack sampling, constructs fully sampled automatic calibration data ACS, by the automatic calibration data ACS
Space more higher than the dimension of current spatial is mapped to, the weight coefficient of each coil is calculated in new space, according to described
The K space data of parallel lack sampling is simulated in weight coefficient filling, and second of reconstruction generates final image.This method does not need additionally
The automatic calibration signal for acquiring K space center, and under higher speed multiple, can preferably inhibit in parallel imaging by
In accelerate multiple become larger and caused by noise amplify, obtain the image of good quality.
Fig. 2 is the structural schematic diagram of the dynamic imaging device of magnetic resonance provided in an embodiment of the present invention, and details are as follows:
The dynamic imaging device of magnetic resonance described in the embodiment of the present invention, comprising:
First lack sampling unit 201, it is right for passing through the automatic calibrated section parallel acquisition TGRAPPA technology of DYNAMIC GENERALIZED
Dynamic data carries out first time lack sampling；
Second lack sampling unit 202, for being adopted using variable density acceleration to the obtained data of first time lack sampling
Sample loading mode carries out second of lack sampling；
First reconstruction unit 203 is obtained for being rebuild according to compression sensing method to second of undersampled signal
To the image with volume pleat artifact；
Data configuration unit 204, it is parallel for simulate by TGRAPPA mode to the K space data obtained after reconstruction
Lack sampling, according to the fully sampled automatic calibration data ACS of the data configuration of lack sampling；
Weight-coefficient calculating unit 205, for by the automatic calibration data ACS map to than current spatial dimension more
High space calculates the weight coefficient of each coil in new space；
Second reconstruction unit 206, for simulating the K space data of parallel lack sampling according to weight coefficient filling, the
Secondary reconstruction generates final image.
Preferably, first sampling unit is specifically used for:
It is R according to presetting drop to adopt rate1, N is acquired altogethertFrame data, to the t on time orientationi(i=1,2......,
Nt) frame data, frequency coding direction is fully sampled, and phase-encoding direction is since first line, every (R1- 1) line acquires one
Line, and the n-th R1+ R frame data are acquired since the R bars line, until NtAll acquisition finishes frame data, wherein 1≤R≤R1,
Preferably, second sampling unit is specifically used for:
Each frame data obtained to first time lack sampling, phase-encoding direction fully sampled by frequency coding direction,
It is theoretical according to the stochastical sampling of compressed sensing, carry out variable density sampling.
The dynamic imaging device of magnetic resonance described in Fig. 2, it is corresponding with the dynamic imaging methods of magnetic resonance described in Fig. 1, herein not
It repeats.
In several embodiments provided by the present invention, it should be understood that disclosed device and method can pass through it
Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the unit, only
Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be tied
Another system is closed or is desirably integrated into, or some features can be ignored or not executed.Another point, it is shown or discussed
Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or logical of device or unit
Letter connection can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes the complete of each embodiment the method for the present invention
Portion or part.And storage medium above-mentioned include: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory),
Random access memory (RAM, Random Access Memory), magnetic or disk etc. be various to can store program code
Medium.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (10)
1. a kind of dynamic imaging methods of magnetic resonance, which is characterized in that the described method includes:
By the automatic calibrated section parallel acquisition TGRAPPA technology of DYNAMIC GENERALIZED, first time lack sampling is carried out to dynamic data；
To the obtained data of first time lack sampling, sample mode is accelerated to carry out second of lack sampling using variable density；
Second of undersampled signal is rebuild according to compression sensing method, obtains the image with volume pleat artifact；
The K space data obtained after reconstruction is carried out to simulate parallel lack sampling by TGRAPPA mode, according to the data structure of lack sampling
Make fully sampled automatic calibration data ACS；
The automatic calibration data ACS is mapped into space more higher than the dimension of current spatial, is calculated often in new space
The weight coefficient of a coil；
The K space data of parallel lack sampling is simulated according to weight coefficient filling, second of reconstruction generates final image.
2. method according to claim 1, which is characterized in that described to pass through the automatic calibrated section parallel acquisition of DYNAMIC GENERALIZED
TGRAPPA technology carries out first time lack sampling to dynamic data, specifically:
It is R according to presetting drop to adopt rate1, N is acquired altogethertFrame data, to the t on time orientationi(i=1,2......, Nt) frame
Data, frequency coding direction is fully sampled, and phase-encoding direction is since first line, every (R1- 1) line acquires a line,
And the n-th R1+ R frame data are acquired since the R bars line, until NtAll acquisition finishes frame data, wherein 1≤R≤R1,
3. method according to claim 1, which is characterized in that the obtained data of first time lack sampling, using change
Density accelerates sample mode to carry out second of lack sampling, step specifically:
Each frame data obtained to first time lack sampling, phase-encoding direction fully sampled by frequency coding direction, according to
The stochastical sampling of compressed sensing is theoretical, carries out variable density sampling.
4. method according to claim 1, which is characterized in that it is described according to compression sensing method to second of lack sampling
Signal is rebuild, and the image with volume pleat artifact, step are obtained specifically:
According to compression sensing method, is rebuild, obtained according to lack sampling data of the Reconstructed equation F ρ=y to each coil channel
Image with volume pleat artifact, in which: F indicates that Fourier owes to adopt operator, and ρ is the image to be rebuild, and y is the space the K number of lack sampling
According to；
It is solved by convex optimization method, so that 1 Norm minimum of each coil channel, obtains the solving condition of Reconstructed equation:The image with volume pleat artifact of each coil is obtained according to the solving condition,
In, | | ρ | |1It is 1 norm, | | ρ | |2It is 2 norms, y is the K space data for owing to adopt, and ε is less than the threshold parameter of level of noise.
5. method according to claim 1, which is characterized in that the described pair of K space data obtained after rebuilding is by the side TGRAPPA
Formula carries out simulating parallel lack sampling, according to the fully sampled automatic calibration data ACS step of the data configuration of lack sampling specifically:
The K space data obtained after reconstruction is carried out to simulate parallel lack sampling by TGRAPPA mode, the N for obtaining lack samplingtFrame
Data, by each frame data t on time orientationi(i=1,2......, Nt) be averaging after addition again, it is calculated and adopts entirely
ACS data.
6. method according to claim 1, which is characterized in that map to the automatic calibration data ACS and compare current spatial
The higher space of dimension, calculate the weight coefficient of each coil in new space, step includes:
The automatic calibration data ACS is mapped into space more higher than the dimension of current spatial according to mapping kernel function K, according to
Weight coefficient calculation formula Aw=b solves to obtain weight coefficient, in which:
The mapping function K is defined as: K (at,i,at,j)=< Ψ (at,i),Ψ(at,j) >, (t=1,2......, Nt),<>indicates
Inner product, at,iIndicate t (t=1,2......, Nt) any two element in the corresponding matrix A of frame, matrix A is by ACS data
In the obtained point composition of all lack samplings, w indicates weight coefficient, and b indicates the point that does not sample in ACS, the size of A be M ×
N, M indicate the total size of the automatic calibration data ACS on removing boundary, and N indicates consecutive points in all coils used in reconstruction
Quantity.
7. method according to claim 6, which is characterized in that the mapping function K using Polynomial kernel function indicate, and: K
(at,i,at,j)=(β aT t,iat,j+c)d, (t=1,2......, Nt), in which: β and c is scalar, d representative polynomial kernel function
Order, solution obtain the weight coefficient are as follows: w=(ΨH(A)Ψ(A))-1ΨH(A) b, wherein ΨH(A) being total to for Ψ (A) is indicated
Yoke transposition, (ΨH(A)Ψ(A))-1Indicate (ΨH(A) Ψ (A)) it is inverse.
8. a kind of dynamic imaging device of magnetic resonance, which is characterized in that described device includes:
First lack sampling unit, for passing through the automatic calibrated section parallel acquisition TGRAPPA technology of DYNAMIC GENERALIZED, to dynamic data
Carry out first time lack sampling；
Second lack sampling unit, for accelerating sample mode using variable density to the obtained data of first time lack sampling
Carry out second of lack sampling；
First reconstruction unit is had for being rebuild according to compression sensing method to second of undersampled signal
Roll up the image of pleat artifact；
Data configuration unit simulates parallel lack sampling for carrying out to the K space data obtained after reconstruction by TGRAPPA mode,
According to the fully sampled automatic calibration data ACS of the data configuration of lack sampling；
Weight-coefficient calculating unit, for the automatic calibration data ACS to be mapped to sky more higher than the dimension of current spatial
Between, the weight coefficient of each coil is calculated in new space；
Second reconstruction unit, for simulating the K space data of parallel lack sampling, second of reconstruction according to weight coefficient filling
Generate final image.
9. device according to claim 8, which is characterized in that first sampling unit is specifically used for:
It is R according to presetting drop to adopt rate1, N is acquired altogethertFrame data, to the t on time orientationi(i=1,2......, Nt) frame
Data, frequency coding direction is fully sampled, and phase-encoding direction is since first line, every (R1- 1) line acquires a line,
And the n-th R1+ R frame data are acquired since the R bars line, until NtAll acquisition finishes frame data, wherein 1≤R≤R1,
10. device according to claim 8, which is characterized in that second sampling unit is specifically used for:
Each frame data obtained to first time lack sampling, phase-encoding direction fully sampled by frequency coding direction, according to
The stochastical sampling of compressed sensing is theoretical, carries out variable density sampling.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611260606.4A CN106491131B (en) | 2016-12-30 | 2016-12-30 | A kind of dynamic imaging methods and device of magnetic resonance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611260606.4A CN106491131B (en) | 2016-12-30 | 2016-12-30 | A kind of dynamic imaging methods and device of magnetic resonance |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106491131A CN106491131A (en) | 2017-03-15 |
| CN106491131B true CN106491131B (en) | 2019-05-07 |
Family
ID=58334749
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611260606.4A Active CN106491131B (en) | 2016-12-30 | 2016-12-30 | A kind of dynamic imaging methods and device of magnetic resonance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106491131B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107958471B (en) * | 2017-10-30 | 2020-12-18 | 深圳先进技术研究院 | CT imaging method and device based on undersampled data, CT equipment and storage medium |
| CN107958472B (en) * | 2017-10-30 | 2020-12-25 | 深圳先进技术研究院 | PET imaging method, device and equipment based on sparse projection data and storage medium |
| CN108825205B (en) * | 2018-04-09 | 2020-09-22 | 中国石油大学(北京) | Method and device for compressed sensing acquisition of underground nuclear magnetic resonance spectrum signals |
| CN109782202B (en) * | 2018-12-14 | 2020-05-15 | 天津大学 | Static magnetic resonance test body model system |
| CN109782203B (en) * | 2018-12-14 | 2020-07-31 | 天津大学 | Dynamic magnetic resonance test body model system |
| CN109633504B (en) * | 2018-12-14 | 2020-05-15 | 天津大学 | Dynamic-static composite magnetic resonance test body model system |
| CN110547799B (en) * | 2019-08-19 | 2022-10-28 | 上海联影医疗科技股份有限公司 | Magnetic resonance imaging method, computer device and computer-readable storage medium |
| CN111696165B (en) * | 2020-05-21 | 2023-09-15 | 深圳安科高技术股份有限公司 | Magnetic resonance image generation method and computer equipment |
| CN112557981B (en) * | 2020-12-03 | 2023-06-13 | 川北医学院 | Improved algorithm of parallel magnetic resonance imaging |
| CN113009398B (en) * | 2021-04-08 | 2021-12-17 | 浙江大学 | Imaging method and apparatus combining k-space and image space reconstruction |
| CN113298901B (en) * | 2021-05-13 | 2022-12-06 | 中国科学院深圳先进技术研究院 | Method for reconstructing magnetic resonance image in convoluted field of view, computer device and storage medium |
| CN113533408A (en) * | 2021-07-21 | 2021-10-22 | 杭州电子科技大学 | Variable density data sampling method for improving quality of parallel magnetic resonance reconstruction image |
| WO2023115339A1 (en) * | 2021-12-21 | 2023-06-29 | 深圳先进技术研究院 | Physics-based data augmentation contrastive learning representation imaging method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102540116A (en) * | 2011-12-08 | 2012-07-04 | 中国科学院深圳先进技术研究院 | Magnetic resonance imaging method and system |
| CN104122520A (en) * | 2013-04-26 | 2014-10-29 | 上海联影医疗科技有限公司 | Magnetic resonance image reconstruction method and device |
| CN104181486A (en) * | 2013-07-05 | 2014-12-03 | 上海联影医疗科技有限公司 | Magnetic resonance image reconstruction method and apparatus |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9153060B2 (en) * | 2012-04-19 | 2015-10-06 | Ohio State Innovation Foundation | Method for estimating a GRAPPA reconstruction kernel |
| US9964621B2 (en) * | 2013-10-01 | 2018-05-08 | Beth Israel Deaconess Medical Center, Inc. | Methods and apparatus for reducing scan time of phase contrast MRI |
-
2016
- 2016-12-30 CN CN201611260606.4A patent/CN106491131B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102540116A (en) * | 2011-12-08 | 2012-07-04 | 中国科学院深圳先进技术研究院 | Magnetic resonance imaging method and system |
| CN104122520A (en) * | 2013-04-26 | 2014-10-29 | 上海联影医疗科技有限公司 | Magnetic resonance image reconstruction method and device |
| CN104181486A (en) * | 2013-07-05 | 2014-12-03 | 上海联影医疗科技有限公司 | Magnetic resonance image reconstruction method and apparatus |
Non-Patent Citations (1)
| Title |
|---|
| Sparsity-Promoting Calibration for GRAPPA Accelerated Parallel MRI Reconstruction;Daniel S. Weller, et. al.;《IEEE Transactions on Medical Imaging》;20130731;第32卷(第7期);1325-1335 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106491131A (en) | 2017-03-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106491131B (en) | A kind of dynamic imaging methods and device of magnetic resonance | |
| CN111513716B (en) | Method and system for magnetic resonance image reconstruction using an extended sensitivity model and a deep neural network | |
| CN106970343B (en) | Magnetic resonance imaging method and device | |
| Ahmad et al. | Variable density incoherent spatiotemporal acquisition (VISTA) for highly accelerated cardiac MRI | |
| US8587307B2 (en) | Systems and methods for accelerating the acquisition and reconstruction of magnetic resonance images with randomly undersampled and uniformly undersampled data | |
| CN105232045B (en) | Single sweep Quantitative MRI Measurement diffusion imaging method based on double echo | |
| CN111951344B (en) | Magnetic resonance image reconstruction method based on cascade parallel convolution network | |
| CN108717717A (en) | The method rebuild based on the sparse MRI that convolutional neural networks and alternative manner are combined | |
| CN107064845A (en) | One-dimensional division Fourier's parallel MR imaging method based on depth convolution net | |
| Gramfort et al. | Denoising and fast diffusion imaging with physically constrained sparse dictionary learning | |
| US11170543B2 (en) | MRI image reconstruction from undersampled data using adversarially trained generative neural network | |
| CN103218795B (en) | Based on the part K spatial sequence image reconstructing method of self-adaptation doubledictionary study | |
| CN106471389A (en) | The reduction of the artifact causing of moving between exciting in MRI due to exciting more | |
| US9939509B2 (en) | Variable density incoherent spatiotemporal acquisition (VISTA) for highly accelerated magnetic resonance imaging | |
| CN104569880B (en) | A kind of magnetic resonance fast imaging method and system | |
| CN103077510B (en) | Multivariate compressive sensing reconstruction method based on wavelet HMT (Hidden Markov Tree) model | |
| CN108447102A (en) | A kind of dynamic magnetic resonance imaging method of low-rank and sparse matrix decomposition | |
| CN108010094A (en) | A kind of MR image reconstruction method and apparatus | |
| CN105118078A (en) | Undersampled CT image reconstruction method | |
| CN103027681A (en) | System used for reconstructing and parallelly obtaining mri image | |
| CN107367703A (en) | Magnetic resonance scanning method, system, device and storage medium | |
| CN115294229A (en) | Method and apparatus for reconstructing Magnetic Resonance Imaging (MRI) images | |
| CN109188327B (en) | Magnetic resonance image fast reconstruction method based on tensor product complex small compact framework | |
| KR102163337B1 (en) | Method for accelerating multiple-acquisition magnetic resonance imaging by varying undersampling-dimension and device for the same | |
| CN105184755A (en) | Parallel magnetic resonance imaging high quality reconstruction method based on self-consistency and containing combined total variation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |