Compressed sensing using regional sparsity

Patent metadata

Field	Value
Publication number	US-9275294-B2
Application number	US-201213569430-A
Country	US
Kind code	B2
Filing date	Aug 8, 2012
Priority date	Nov 3, 2011
Publication date	Mar 1, 2016
Grant date	Mar 1, 2016

Abstract

A method for reconstructing an image includes acquiring raw image data during a scan of an area, estimating an image from the raw image data, separating the estimated image into a region of interest (ROI) and a background region, and applying compressed sensing to iteratively update only the ROI and maintain the background region to reconstruct an image.

First claim

What is claimed is:

1. A method for reconstructing an image comprises:
estimating an image from acquired raw image data;
separating the estimated image into a region of interest (ROI) and a background region; and
applying compressed sensing to the ROI while excluding application of the compressed sensing with respect to the background region to iteratively update only the ROI while maintaining the background region, to reconstruct an image,
wherein the applying comprises performing the compressed sensing on a sum of a fidelity term and a sparsity term,
wherein the fidelity term includes a difference between a transform of the estimated image and the raw image data, and
wherein the sparsity term includes a mask multiplied by the estimated image that masks out the background region and retains the ROI.

2. The method of claim 1 , wherein the background entirely surrounds the ROI.

3. The method of claim 1 , wherein the separating further separates the estimated image into a second ROI distinct from the first ROI, and the applying comprises applying the compressed sensing to iteratively update the first ROI using a first sparsity constraint and iteratively update the second ROI using a second other sparsity constraint.

4. The method of claim 3 , wherein one of the two sparsity constraints is image domain sparsity and the other is spatial total variance sparsity.

5. The method of claim 1 , wherein prior to the estimating, the method comprises:
acquiring un-gridded image data using a radial trajectory acquisition and oversampling; and
gridding the un-gridded image data to generate the raw image data.

6. The method of claim 1 , wherein estimating the image comprises performing an inverse Fourier transform on the raw image data.

7. The method of claim 1 , wherein the sparsity term applies a sparsity transform to a result of the mask multiplied by the estimated image.

8. The method of claim 7 , wherein the sparsity transform generates a sparse result.

9. The method of claim 1 , wherein the mask is a matrix of 1s and 0s, the 0s corresponding to a location of the background region and the 1s corresponding to a location of the ROI.

10. The method of claim 1 , wherein a weighting factor is multiplied by the sparsity term.

11. A non-transitory computer readable storage medium embodying instructions executable by a processor to perform method steps for reconstructing an image, the method steps comprising instructions for:
estimating an image from acquired raw image data;
separating the estimated image into a region of interest (ROI) and a background region; and
applying compressed sensing to the ROI while excluding application of the compressed sensing with respect to the background region to iteratively update only the ROI while maintaining the background region, to reconstruct an image,
wherein the applying comprises performing the compressed sensing on a sum of a fidelity term and a sparsity term,
wherein the fidelity term includes a difference between a transform of the estimated image and the raw image data, and
wherein the sparsity term includes a mask multiplied by the estimated image that masks out the background region and retains the ROI.

12. The computer readable medium of claim 11 , wherein the background region entirely surrounds the ROI.

13. The computer readable storage medium of claim 11 , wherein the separating further separates the estimated image into a second ROI distinct from the first ROI, and the applying comprises applying the compressed sensing to iteratively update the ROI using a first sparsity constraint and iteratively update the second ROI using a second other sparsity constraint.

14. The computer readable storage of claim 11 , wherein the instructions further include instructions for:
acquiring un-gridded image data using a radial trajectory acquisition and oversampling; and
gridding the un-gridded image data to generate the raw image data.

15. A method for reconstructing an image comprises:
generating a cost function including a fidelity term and a sparsity term; and
minimizing the generated cost function to generate a reconstructed image,
wherein the fidelity term includes a difference between a transform of an estimated image and raw image data, and
wherein the sparsity term includes a mask multiplied by the estimated image that retains a region of interest (ROI) of the estimated image and masks out a background of the estimated image.

16. The method of claim 15 , wherein the minimizing comprises performing compressed sensing on the cost function.

17. The method of claim 15 , wherein prior to the generating of the cost function, the method comprises:
acquiring un-gridded image data using a radial trajectory acquisition and oversampling; and
gridding the un-gridded image data to generate the raw image data.

18. The method of claim 5 wherein prior to the generating of the cost function, the method comprises performing an inverse Fourier transform on the raw image data to generate the estimated image.

19. The method of claim 15 , wherein the sparsity term applies a sparsity transform to a result of the mask multiplied by the estimated image.

20. The method of claim 15 , wherein the background region entirely surrounds the ROI.

Assignees

• Chen Xiao
• Pan Li
• Siemens Ag

Inventors

• Chen Xiao
• Pan Li

Classifications

• G06V10/7715

  Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods · CPC title

• G06V10/25

  Determination of region of interest [ROI] or a volume of interest [VOI] · CPC title

• G06V10/22

  by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition · CPC title

• 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

• G06F18/2136

  based on sparsity criteria, e.g. with an overcomplete basis · CPC title