Non-contrast-enhanced 4D MRA using compressed sensing reconstruction

• US 8,879,852 B2
• Filed: 10/28/2011
• Issued: 11/04/2014
• Est. Priority Date: 11/10/2010
• Status: Active Grant

• Current Assignee: Siemens Healthineers AG (Siemens AG)
• Original Assignee: Siemens AG
• Inventors: Chang, Ti-chiun, Nadar, Mariappan S., Ghring, Jens, Zenge, Michael, Block, Kai Tobias, Schmitt, Peter, Mueller, Edgar
• Primary Examiner(s): Patel, Nirav G
• Assistant Examiner(s): Mistry, Oneal R
• Application Number:
  US13/284,143
• Publication Number:
  US 20130121550A1
• Time in Patent Office: 1,103 Days
• Field of Search: None
• US Class Current: 382/206
• CPC Class Codes: A61B 5/055 involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
• CPC Class Codes: G01R 33/561 by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
• CPC Class Codes: G01R 33/5635 Angiography, e.g. contrast-enhanced angiography [CE-MRA] or time-of-flight angiography [TOF-MRA]
• CPC Class Codes: G01R 33/56545 caused by finite or discrete sampling, e.g. Gibbs ringing, truncation artefacts, phase aliasing artefacts
• CPC Class Codes: G06T 11/003 Reconstruction from projections, e.g. tomography

Abstract

A reconstructed image is rendered of a patient by a processor from a set of undersampled MRI data by first subtracting two repetitions of the acquired data in k-space to create a third dataset. The processor reconstructs the image by minimizing an objective function under a constraint related to the third dataset, wherein the objective function includes applying a Karhunen-Loeve Transform (KLT) to a temporal dimension of data. The objective function under the constraint is expressed as arg min_f{∥φ(f)∥₁ subject to ∥Af−y∥₂≦ε}. The reconstructed image is an angiogram which may be a 4D angiogram. The angiogram is used to diagnose a vascular disease.

Claims

1. A method for reconstructing an image of an object from image data, comprising:

a processor receiving non-contrast-enhanced magnetic resonance image data of the object containing a first and a second dataset in k-space obtained with a medical imaging device which are acquired in an undersampled manner, wherein the first and second dataset in k-space are steady-state free precession (SSFP) repetition acquisitions of one phase of multiple phases in a cine angiography;

the processor generating a sparsified third dataset prior to image reconstruction by creating a difference set in k-space by subtracting the first from the second dataset in k-space and by applying a Karhunen-Loeve Transform (KLT) over a temporal dimension of an inverse Fourier Transform of the difference set; and

the processor iteratively minimizing an objective function under a constraint related to the sparsified third dataset to reconstruct the image of the object, wherein one or more artifacts due to the acquisition in the undersampled manner are suppressed.

2. The method of claim 1, wherein the minimizing of the objective function under the constraint is expressed as

arg min _f{∥

φ

(f)∥

₁ subject to ∥

Af−

y∥

₂≦

ε

}, wherein

f represents an iterative image;

φ

(f) represents the Karhunen-Loeve Transform of f;

y represents the sparsified third dataset;

A represents a Fourier operator that maps the iterative image into k-space; and

ε

represents a parameter that accounts for a deviation between a measurement and ideal data.

3. The method of claim 1, wherein the object is a patient.

4. The method of claim 1, wherein the reconstructed image is an angiogram.

5. The method of claim 4, wherein the angiogram is a 4D angiogram.

6. The method of claim 1, wherein the method is used in diagnosing a vascular disease.

7. The method of claim 1, wherein the suppression of the one or more artifacts introduced by the undersampled acquisition, which is expressed as a relative mean squared error between an energy-normalized reconstructed 4D volume using the steps of claim 1 and a fully sampled ground truth, compared to a relative mean squared error between an energy-normalized reconstructed 4D volume that does not apply KLT on the undersampled acquisition and a fully sampled ground truth, is at least 35% lower.

8. A system to reconstruct an image of an object from image data, comprising:

a memory enabled to store data;

a processor enabled to execute instructions to perform the steps;

receiving non-contrast-enhanced magnetic resonance image data of the object containing a first and a second dataset in k-space obtained with a medical imaging device which are acquired in an undersampled manner, wherein the first and second dataset in k-space are steady-state free precession (SSFP) repetition acquisitions of one phase of multiple phases in a cine angiography;

generating a sparsified third dataset prior to image reconstruction by creating a difference set in k-space by subtracting the first from the second dataset in k-space and by applying a Karhunen-Loeve Transform (KLT) over a temporal dimension of an inverse Fourier Transform of the difference set; and

iteratively minimizing an objective function under a constraint related to the sparsified third dataset to reconstruct the image of the object, wherein one or more artifacts due to the acquisition in the undersampled manner are suppressed.

9. The system of claim 8, wherein the minimizing of the objective function under the constraint is expressed as

arg min _f{∥

φ

(f)∥

₁ subject to ∥

Af−

y∥

₂≦

ε

}, wherein

f represents an iterative image;

φ

(f) represents a the Karhunen-Loeve Transform of f;

y represents the sparsified third dataset;

A represents a Fourier operator that maps the iterative image into k-space; and

ε

represents a parameter that accounts for a deviation between a measurement and ideal data.

10. The system of claim 8, wherein the object is a patient.

11. The system of claim 8, wherein the reconstructed image is an angiogram.

12. The system of claim 11, wherein the angiogram is a 4D angiogram.

13. The system of claim 8, wherein the system is used in diagnosing a vascular disease.

14. The system of claim 8, wherein the suppression of the one or more artifacts introduced by the undersampled acquisition, which is expressed as a relative mean squared error between an energy-normalized reconstructed 4D volume using the steps of claim 8 and a fully sampled ground truth, compared to a relative mean squared error between an energy-normalized reconstructed 4D volume that does not apply KLT on the undersampled acquisition and a fully sampled ground truth, is at least 35% lower.