Systems and methods for fast reconstruction for quantitative susceptibility mapping using magnetic resonance imaging

Patent metadata

Field	Value
Publication number	US-9542763-B2
Application number	US-201514693525-A
Country	US
Kind code	B2
Filing date	Apr 22, 2015
Priority date	Apr 25, 2014
Publication date	Jan 10, 2017
Grant date	Jan 10, 2017

Abstract

Described here are systems and methods for quantitative susceptibility mapping (“QSM”) using magnetic resonance imaging (“MRI”). Susceptibility maps are reconstructed from phase images using an automatic regularization technique based in part on variable splitting. Two different regularization parameters are used, one, λ, that controls the smoothness of the final susceptibility map and one, μ, that controls the convergence speed of the reconstruction. For instance, the regularization parameters can be determined using an L-curve heuristic to find the parameters that yield the maximum curvature on the L-curve. The μ parameter can be determined based on an l 2 -regularization and the λ parameter can be determined based on the iterative l 1 -regularization used to reconstruct the susceptibility map.

First claim

The invention claimed is:

1. A method for performing quantitative susceptibility mapping using a magnetic resonance imaging (MRI) system, the steps of the method comprising:
a) acquiring data from a subject using the MRI system;
b) reconstructing phase images from the acquired data;
c) computing at least one initial estimate of a susceptibility map based on the reconstructed phase images;
d) determining a first regularization parameter based on the at least one initial estimate of the susceptibility map; and
e) reconstructing a susceptibility map using an l 1 -minimization that comprises iteratively minimizing an objective function that includes the first regularization parameter and a second regularization parameter, wherein the second regularization parameter is undated in each iteration based on iterative updates to the susceptibility map in that iteration.

2. The method as recited in claim 1 , wherein the objective function introduces an auxiliary variable that replaces an image gradient term, such that the l 1 -minimization has a closed form.

3. The method as recited in claim 1 , wherein step c) includes computing the at least one initial estimate of the susceptibility map using an l 2 -minimization that comprises minimizing an objective function that includes the first regularization parameter.

4. The method as recited in claim 3 , wherein the objective function in the l 2 -minimization includes a magnitude prior.

5. The method as recited in claim 3 , wherein the first regularization parameter is determined in step d) based on an L-curve computed using a consistency condition enforced in the l 2 -minimization and a regularization used in the l 2 -minimization.

6. The method as recited in claim 5 , wherein the first regularization parameter is determined in step d) based on a curvature of the L-curve.

7. The method as recited in claim 6 , wherein the first regularization parameter is determined in step d) as a point of maximum curvature in the curvature of the L-curve.

8. The method as recited in claim 1 , wherein step e) includes determining the second regularization parameter based on an L-curve computed using a consistency condition enforced in the l 1 -minimization and a regularization used in the l 1 -minimization.

9. The method as recited in claim 8 , wherein the second regularization parameter is determined in step e) based on a curvature of the L-curve.

10. The method as recited in claim 9 , wherein the second regularization parameter is determined in step e) as a point of maximum curvature in the curvature of the L-curve.

11. The method as recited in claim 1 , wherein step b) further comprises processing the reconstructed phase images to remove phase wrapping.

12. The method as recited in claim 1 , wherein step b) further comprises removing a relative phase offset from the acquired data before reconstructing the phase images.

13. The method as recited in claim 1 , wherein the objective function in the l 1 -minimization includes a magnitude prior.

14. The method as recited in claim 1 , wherein the data acquired in step a) is susceptibility-weighted data.

Assignees

• Bilgic Berkin
• Setsompop Kawin
• Massachusetts Gen Hospital

Inventors

• Bilgic Berkin
• Setsompop Kawin

Classifications

• G06T11/26 Primary

  Drawing of charts or graphs · CPC title

• A61B5/4848

  Monitoring or testing the effects of treatment, e.g. of medication · CPC title

• G01R33/56536

  due to magnetic susceptibility variations · CPC title

• G01R33/5602

  by filtering or weighting based on different relaxation times within the sample, e.g. T1 weighting using an inversion pulse · CPC title

• G01R33/443

  Assessment of an electric or a magnetic field, e.g. spatial mapping, determination of a B0 drift or dosimetry · CPC title