Efficient method for performing K-space channel combination for non-cartesian MRI acquisitions

Patent metadata

Field	Value
Publication number	US-9389292-B2
Application number	US-201414501323-A
Country	US
Kind code	B2
Filing date	Sep 30, 2014
Priority date	Sep 30, 2013
Publication date	Jul 12, 2016
Grant date	Jul 12, 2016

Abstract

A system and method for combining k-space data acquired on multiple different receiver channels in a multichannel receiver is provided. One or more convolution kernels are used to combine the k-space data. Each convolution kernel is designed as the combination of one or more channel combination kernels and an alias-suppressing kernel. The channel combination kernels are designed to have a smaller sample spacing than the acquired data, and the alias-suppressing kernel is designed to suppress aliasing artifacts in stopbands defined by the sample spacing of the channel combination kernels.

First claim

The invention claimed is:

1. A method for combining k-space measurements made on a plurality of different receiver channels in a multichannel receiver that forms a part of a magnetic resonance imaging (MRI) system, the steps of the method comprising:
a) directing the MRI system to acquire a k-space data set for a plurality of different receiver channels in the multichannel receiver;
b) forming at least one convolution kernel by combining:
i) at least one coil combination kernel having a sample spacing smaller than a sample spacing in the acquired k-space data sets; and
ii) an alias-suppressing kernel that is designed based on the sample spacing of the at least one coil combination kernel;

c) generating channel combined data by applying the at least one convolution kernel formed in step b) to the k-space data sets acquired in step a); and
d) reconstructing an image from the channel combined data generated in step c).

2. The method as recited in claim 1 wherein the at least one coil combination kernel includes a discretely defined non-separable channel combination kernel.

3. The method as recited in claim 1 wherein the at least one coil combination kernel includes a direct virtual coil (DVC) combination kernel.

4. The method as recited in claim 1 wherein the alias-suppressing kernel is a continuously defined separable aliasing-suppressing kernel.

5. The method as recited in claim 1 wherein the alias-suppressing kernel is a Kaiser-Bessel kernel.

6. The method as recited in claim 1 wherein step b) includes convolving the at least one coil combination kernel with the alias-suppressing kernel.

7. The method as recited in claim 1 wherein step a) includes acquiring the k-space data sets by sampling k-space along non-Cartesian k-space trajectories.

8. The method as recited in claim 1 wherein step c) includes processing the k-space data sets acquired in step a) before applying the at least one convolution kernel.

9. The method as recited in claim 8 wherein the processing in step c) includes forming synthesized k-space data sets from the acquired k-space data sets, and wherein step c) includes applying the at least one convolution kernel at least to the synthesized k-space data sets.

10. The method as recited in claim 8 wherein the acquired k-space data sets are N-dimensional k-space data sets and the processing in step c) includes Fourier transforming the acquired k-space data sets along less than N dimensions.

11. The method as recited in claim 1 wherein the plurality of different receiver channels comprises a subset of an available number of receiver channels in the multichannel receiver.

12. A method for reconstructing an image of a subject using a magnetic resonance imaging (MRI) system, the steps of the method comprising:
a) directing the MRI system to acquire k-space data on a plurality of different receiver channels in a multichannel receiver;
b) generating combined k-space data by convolving the acquired k-space data with at least one kernel that includes a first kernel component that resamples the k-space data to a finer sample spacing than in the acquired data and a second kernel component that suppresses aliasing in stopbands that are defined by the sample spacing of the first kernel component; and
c) reconstructing an image from the combined k-space data.

13. The method as recited in claim 12 wherein the first kernel component is a discretely defined non-separable channel combination kernel.

14. The method as recited in claim 12 wherein the first kernel component is a direct virtual coil (DVC) combination kernel.

15. The method as recited in claim 12 wherein the second kernel component is a continuously defined separable aliasing-suppressing kernel.

16. The method as recited in claim 12 wherein the second kernel component is a Kaiser-Bessel kernel.

17. The method as recited in claim 12 wherein step a) includes acquiring the k-space data by sampling k-space along non-Cartesian k-space trajectories.

18. The method as recited in claim 12 wherein the first kernel component comprises a plurality of kernel components that each resample the k-space data to a finer sample spacing than in the acquired data.

19. The method as recited in claim 12 wherein the plurality of different receiver channels comprises a subset of an available number of receiver channels in the multichannel receiver.

Assignees

• Beatty Philip James
• Sunnybrook Res Inst

Inventors

• Beatty Philip James

Classifications

• G01R33/5611 Primary

  Parallel magnetic resonance imaging, e.g. sensitivity encoding [SENSE], simultaneous acquisition of spatial harmonics [SMASH], unaliasing by Fourier encoding of the overlaps using the temporal dimension [UNFOLD], k-t-broad-use linear acquisition speed-up technique [k-t-BLAST], k-t-SENSE (structural details of arrays of sub-coils G01R33/3415) · CPC title

• G01R33/4824

  using a non-Cartesian trajectory · CPC title