Systems and methods for motion-compensated reconstruction of magnetic resonance images
Systems and methods for reconstructing a motion-compensated magnetic resonance image are presented. In certain implementations, a computer-implemented method is provided. The method may include a plurality of operations, including receiving a set of k-space data from a magnetic resonance imaging device, dividing the set of k-space data into a plurality of groups, performing a plurality of initialization operations, performing a first iterative process until a first criteria for the first iterative process is achieved for a current scale of motion estimation, performing a second iterative process until a second criteria for the second iterative process is achieved, and outputting a motion-compensated magnetic resonance image reconstructed in accordance with a predetermined scale of motion estimation.
1. A computer-implemented method for reconstructing a motion-compensated MR image, the method comprising:
receiving a set of k-space data from an MRI device;
dividing the set of k-space data into a plurality of groups;
performing a plurality of initialization operations;
performing a first iterative process until a first criteria for the first iterative process is achieved, the first iterative process outputting, for a current scale of motion estimation,
a set of deformation coefficients of a motion estimation model for each group of k-space data; and
a motion-compensated MR image reconstructed based on all the sets of the deformation coefficients and all the groups of k-space data; and
performing a second iterative process until a second criteria for the second iterative process is achieved, the second iterative process outputting
a final motion-compensated MR image reconstructed based on the output of the first iterative process and a predetermined scale of motion estimation.
2. The computer-implemented method of claim 1 , wherein the set of k-space data comprises a plurality of subsets of k-space data simultaneously acquired using a plurality of coil arrays.
3. The computer-implemented method of claim 1 , wherein the set of k-space data was acquired according to a sampling pattern selected for parallel imaging, wherein the sampling pattern is a radial trajectory or a spiral trajectory.
4. The computer-implemented method of claim 3 , wherein the sampling pattern is a variable-density sampling pattern.
5. The computer-implemented method of claim 1 , wherein each iteration of the first iterative process comprises
estimating a set of deformation coefficients of the motion estimation model for each group of k-space data; and
reconstructing the motion-compensated MR image based on all the sets of the deformation coefficients and all the groups of k-space data.
6. The computer-implemented method of claim 1 , wherein each iteration of the second iterative process comprises
updating the current scale of motion estimation; and
performing one or more iterations of the first iterative process.
7. The computer-implemented method of claim 1 , wherein the plurality of initialization operations comprises
set estimated motion during acquisition of a first group of k-space data to zero;
obtaining an initial coarse image; and setting an initial scale of motion estimation.
8. The computer-implemented method of claim 1 , further comprising estimating the set of deformation coefficients of the motion estimation model for each group of k-space data via non-linear conjugate gradient optimization.
9. The computer-implemented method of claim 1 , wherein the motion estimation model comprises a b-spline function and one or more regularization terms.
10. The computer-implemented method of claim 9 , wherein the one or more regularization terms comprise a quadratic penalty function for constraining an extent of estimated motion.
11. The computer-implemented method of claim 9 , wherein the one or more regularization terms comprise a quadratic penalty function for implementing a constraint selected for a parallel MR imaging technique.
12. A non-transitory computer readable medium that stores a set of instructions that is executable by at least one processor to cause the processor to perform a method for reconstructing a motion-compensated MR image, the method comprising:
receiving a set of k-space data from an MRI device;
dividing the set of k-space data into a plurality of groups;
performing a plurality of initialization operations;
performing a first iterative process until a first criteria for the first iterative process is achieved, the first iterative process outputting, for a current scale of motion estimation,
a set of deformation coefficients of a motion estimation model for each group of k-space data; and
a motion-compensated MR image reconstructed based on all the sets of the deformation coefficients and all the groups of k-space data; and
performing a second iterative process until a second criteria for the second iterative process is achieved, the second iterative process outputting
a final motion-compensated MR image reconstructed based on the output of the first iterative process and a predetermined scale of motion estimation.
13. The non-transitory computer readable medium of claim 12 , wherein the set of k-space data comprises a plurality of subsets of k-space data simultaneously acquired using a plurality of coil arrays.
14. The non-transitory computer readable medium of claim 12 , wherein the set of k-space data was acquired according to a sampling pattern selected for parallel imaging, wherein the sampling pattern is a radial trajectory or a spiral trajectory.
15. The non-transitory computer readable medium of claim 14 , wherein the sampling pattern is a variable-density sampling pattern.
16. The non-transitory computer readable medium of claim 12 , wherein each iteration of the first iterative process comprises
estimating a set of deformation coefficients of the motion estimation model for each group of k-space data; and
reconstructing the motion-compensated MR image based on all the sets of the deformation coefficients and all the groups of k-space data.
17. The non-transitory computer readable medium of claim 12 , wherein each iteration of the second iterative process comprises
updating the current scale of motion estimation; and
performing one or more iterations of the first iterative process.
18. The non-transitory computer readable medium of claim 12 , wherein the plurality of initialization operations comprises
set estimated motion during acquisition of a first group of k-space data to zero;
obtaining an initial coarse image; and setting an initial scale of motion estimation.
19. The non-transitory computer readable medium of claim 12 , wherein the set of instructions that is executable by the at least one processor to cause the server to further perform
estimating the set of deformation coefficients of the motion estimation model for each group of k-space data via non-linear conjugate gradient optimization.
20. A method for performing MRI, the method comprising:
acquiring a set of k-space data;
dividing the set of k-space data into a plurality of groups;
performing a plurality of initialization operations;
performing a first iterative process until a first criteria for the first iterative process is achieved, the first iterative process outputting, for a current scale of motion estimation,
a set of deformation coefficients of a motion estimation model for each group of k-space data; and
a motion-compensated MR image reconstructed based on all the sets of the deformation coefficients and all the groups of k-space data; and
performing a second iterative process until a second criteria for the second iterative process is achieved, the second iterative process outputting
a final motion-compensated MR image reconstructed based on the output of the first iterative process and a predetermined scale of motion estimation.