ACCELERATED PSEUDO-RANDOM DATA MAGENTIC RESONANCE IMAGING SYSTEM AND METHOD

• US 20110241677A1
• Filed: 03/18/2011
• Published: 10/06/2011
• Est. Priority Date: 04/02/2010
• Status: Active Grant

• Current Assignee: General Electric Company, Wisconsin Alumni Research Foundation (University of Wisconsin System)
• Original Assignee: General Electric Company
• Inventors: Wang, Kang, Korosec, Frank, Holmes, James, Beatty, Philip James, Busse, Reed Frederick
• Granted Patent: US 8,653,817 B2
• Time in Patent Office: Days
• US Class Current: 324/309
• CPC Class Codes: G01R 33/4818 MR characterised by data acquisition along a specific k-space trajectory or by the temporal order of k-space coverage, e.g. centric or segmented coverage of k-space
• CPC Class Codes: G01R 33/5611 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 Class Codes: G01R 33/56308 Characterization of motion or flow; Dynamic imaging

Abstract

The present disclosure is intended to describe embodiments for improving image data acquisition and processing in accelerated dynamic magnetic resonance imaging sequences. One embodiment is described where a method includes an acquisition sequence configured to acquire an undersampled set of magnetic resonance data. The undersampled set of magnetic resonance data has a pseudo-random sampling pattern within a data space acquired at a first time, the pseudo-random sampling pattern being influenced by other pseudo-random sampling patterns within the data space arising from the acquisition of additional undersampled sets of magnetic resonance data at respective times. In some embodiments, the pseudo-random sampling patterns of the undersampled sets of magnetic resonance data interleave to yield a desired sampling pattern.

Claims

1. A magnetic resonance imaging method, comprising:

performing parallel imaging reconstruction on a set of magnetic resonance data to generate an initial image, the set of magnetic resonance data comprising data acquired in a time frame of interest and having a non-uniform sampling pattern on a predefined area of a data space.

2. The method of claim 1, wherein the predefined area of the data space comprises a uniform pattern of locations that are to be sampled and locations that are not to be sampled.

3. The method of claim 2, wherein the non-uniform sampling pattern comprises a pseudo-random sampling pattern having locations towards the center of the data space that are sampled with greater frequency than those further from the center of the data space.

4. The method of claim 1, wherein the patterns are along the phase encoding axes of k-space or hybrid space.

5. The method of claim 1, wherein performing parallel imaging reconstruction comprises performing parallel imaging reconstruction on a coil-by-coil basis to suppress aliasing artifacts, and the parallel imaging reconstruction is calibrated for the uniform sampling pattern.

6. The method of claim 5, comprising:

performing one or more additional acquisition sequences at respective time frames to acquire one or more additional sets of magnetic resonance data having respective pseudo-random data undersampling patterns on the predefined area; and

performing back projection reconstruction to generate a reconstructed image for the time frame of interest using the set of magnetic resonance data and the additional sets of magnetic resonance data.

7. The method of claim 6, wherein the initial image is used for processing during the back projection reconstruction.

8. The method of claim 7, wherein performing the back projection reconstruction comprises:

generating a temporally blurred image substantially free of spatial blurring from a combined set of magnetic resonance data, wherein the combined set of magnetic resonance data is generated using a combination function to combine the set of magnetic resonance data acquired in the timeframe of interest and the additional sets of magnetic resonance data;

generating a location-corrected image from the combined set of magnetic resonance data and the sampling pattern of the set of magnetic resonance data; and

generating the reconstructed image using the image using the initial, the temporally blurred, and the location-corrected images.

9. The method of claim 8, wherein generating the temporally blurred image comprises correcting the combined set of magnetic resonance data for the sampling density of the combined set of data.

10. The method of claim 9, wherein generating the location-corrected image comprises sampling the temporally blurred image at the pseudo-random locations of the set of magnetic resonance data acquired in the timeframe of interest.

11. The method of claim 6, wherein performing the back projection reconstruction comprises suppressing incoherent image artifacts by using the temporally blurred image as a multiplicative constraint.

12. A magnetic resonance imaging method, comprising:

performing a magnetic resonance acquisition sequence comprising;

defining a uniform undersampling pattern comprising locations that are to be sampled and locations that are to be left unsampled on at least a portion of a phase encoding plane of a data space; and

acquiring an undersampled set of magnetic resonance data representative of a feature of a subject of interest using a non-uniform undersampling pattern on the defined uniform undersampling pattern.

13. The method of claim 12, wherein the non-uniform sampling pattern comprises a pseudo-random undersampling pattern.

14. The method of claim 13, wherein the data space is k-space or hybrid space, and the pseudo-random sampling pattern comprises locations towards the center of the data space that are sampled with greater frequency than those further from the center of the data space.

15. A magnetic resonance imaging method comprising:

performing multiplicative constrained reconstruction on an undersampled set of magnetic resonance data acquired within a timeframe of interest to generate a desired image;

wherein the data comprises an undersampling pattern having locations that are sampled and locations that are not sampled along a phase encoding plane of a data space.

16. The method of claim 15, wherein the undersampling pattern is generated using a TRIPPS acquisition, a TRICKS acquisition, a CAPR acquisition, or any combination thereof.

17. The method of claim 15, wherein undersampling pattern comprises a pseudo-random sampling pattern.

18. The method of claim 17, wherein the pseudo-random sampling pattern is dependent upon at least one other pseudo-random sampling pattern from one or more additional undersampled sets of magnetic resonance data.

19. The method of claim 17, wherein the data comprises a predefined uniform sampling pattern on which the pseudo-random sampling pattern is disposed, the predefined uniform sampling pattern comprising a uniform pattern of sampling locations within the data space that are able to be sampled using the pseudo-random sampling pattern and locations that are not able to be sampled using the pseudo-random sampling pattern.

20. The method of claim 19, comprising performing parallel imaging reconstruction prior to performing multiplicative constrained reconstruction.

21. The method of claim 15, wherein the multiplicative constrained reconstruction comprises:

generating a first image from the undersampled magnetic resonance data set acquired in the time frame of interest, the first image being substantially free of temporal blurring;

generating a second image substantially free of spatial blurring generated from a combined set of magnetic resonance data, the combined set of magnetic resonance data comprising the undersampled magnetic resonance data set acquired in the time frame of interest and at least one additional undersampled magnetic resonance data set acquired in at least one additional time frame; and

generating a third image from location-corrected data, and the location corrected data is generated by sampling the combined set of magnetic resonance data using the undersampling pattern of the undersampled magnetic resonance data set acquired in the time frame of interest.

22. The method of claim 21, comprising:

generating the desired image by dividing the first image by the third image and applying a multiplicative constraint, the multiplicative constraint comprising the second image.