System and method for reconstruction of magnetic resonance images acquired with partial Fourier acquisition

1. A method for medical image registration in k-space domain, the method comprising:

receiving, using a controller, a first partial k-space dataset for an object and a second partial k-space dataset for the object;

selecting, using the controller, the first partial k-space dataset as a reference;

selecting, using the controller, feature for estimating a transformation matrix for transforming k-space data;

estimating, using the controller, a transformation matrix based on the feature of entire or part of the first partial k-space dataset and the feature of the second partial k-space dataset corresponding to the entire or part of the first partial k-space dataset;

correcting, using the controller, the second partial k-space dataset based on the transformation matrix;

obtaining, using the controller, the corrected second partial k-space dataset; and

storing the corrected second partial k-space dataset in an image data storage unit for magnetic resonance reconstruction.

2. A method for Magnetic Resonance Imaging (MRI) partial Fourier reconstruction, the method comprising:

acquiring, using a controller, a first partial k-space dataset for a target area using an MRI scanner;

creating, using the controller, a second partial k-space dataset for the target area based on Hermitian symmetry of the first partial k-space dataset;

calculating, using the controller, a transformation matrix based on the first partial k-space dataset and the second partial k-space dataset;

correcting, using the controller, the second partial k-space dataset using the transformation matrix;

obtaining, using the controller, a full k-space dataset for the target area based on at least one of the corrected second partial k-space dataset, the first partial k-space dataset, and a third k-space dataset;

constructing, using the controller, an image for the target area based on the full k-space dataset; and

storing the corrected second partial k-space dataset in an image data storage unit for magnetic resonance reconstruction.

3. A magnetic resonance imaging (MRI) system comprising:

a magnetic field generating unit configured to apply a plurality of RF pulses with a variable flip angle to a target area in an object;

a receiver configured to receive MR signals from the target area;

a processing unit;

a system memory; and

machine readable instructions stored in the system memory that, when executed by the processing unit, cause the processing unit to:

acquire a first partial k-space dataset for the target area based on the MR signals;

create a second partial k-space dataset for the target area based on Hermitian symmetry of the first partial k-space dataset;

calculate a transformation matrix based on the first partial k-space dataset and the second partial k-space dataset;

correct the second partial k-space dataset using the transformation matrix;

obtain a full k-space dataset for the target area based on at least one of the corrected second partial k-space dataset, the first partial k-space dataset, and a third k-space dataset;

reconstruct an image for the target area based on the full k-space dataset; and

output the image through an output device.

4. The method of claim 1 , wherein receiving the first partial k-space dataset and the second partial k-space dataset comprises:

acquiring the first partial k-space dataset and the second partial k-space dataset from at least one modality; or

determining the first partial k-space dataset and the second partial k-space dataset from corresponding images of electronic storage on image domain; or

determining the first partial k-space dataset and the second partial k-space dataset from a mathematic model.

5. The method of claim 1 , wherein the feature includes at least one of a phase of k-space data that object feature is mainly weighted by image magnitude of a magnetic resonance image associated with the k-space domain, or a magnitude of k-space data that object feature is mainly weighted by image phase of the magnetic resonance image associated with the k-space domain.

6. The method of claim 5 , wherein the feature comprises at least one of a phase of k-space data that are acquired with T 1 -weighted or its variation sequence, or a magnitude of k-space data that are acquired with T 2 -weighted, T 2 *-weighted, or their variation sequence.

7. The method of claim 1 , wherein the transformation matrix is configured to transform the first partial k-space dataset or the second partial k-space dataset by at least one of translation, rotation, scaling and shearing of the first partial k-space dataset, or the second partial k-space dataset.

8. The method of claim 4 , wherein the at least one modality is at least one of optical image, CT, MRI, ultrasound, and PET.

9. The method of claim 2 , wherein the first partial k-space dataset is a partial Fourier k-space dataset; and

wherein the partial Fourier k-space dataset is acquired in at least one of phase-encoding and frequency-encoding direction.

10. The method of claim 2 , wherein calculating the transformation matrix comprises:

estimating the transformation matrix based on a part of the first partial k-space dataset and a corresponding part of the second partial k-space dataset.

11. The method of claim 2 , wherein calculating the transformation matrix comprises:

acquiring a full Fourier k-space dataset for the target area using an MR scanner with a resolution that is lower than a resolution used for acquiring the first partial k-space dataset; and

calculating the transformation matrix based on the full Fourier k-space dataset and the second partial k-space dataset.

12. The method of claim 11 , wherein the transformation matrix is configured to correct translation, rotation, scaling and shearing of an entire or part of k-space dataset.

13. The method of claim 2 , wherein the third k-space dataset comprises a k-space dataset which is not covered by the first k-space dataset and the second k-space dataset.

14. The method of claim 13 , wherein the third k-space dataset is filled with zeros, or the third k-space dataset extends a boundary of the first k-space dataset or the second k-space dataset.

15. The method of claim 2 , wherein the full k-space dataset is filtered using at least one of hard thresholding, Hamming filtering, Hanning filtering, Blackman filtering, Lanczos filtering, Gaussian filtering, and Wiener filtering.

16. The method of claim 2 , wherein the first partial k-space dataset is acquired using at least one of rectilinear, echo planar, a propeller, a blade, radial, Cartesian, non-Cartesian, Zig-Zag, stochastic, rosette, TWIRL, WHIRL and spiral trajectories.

17. The method of claim 2 , wherein the first partial k-space dataset is acquired using at least one of a sequential sampling order, a centric sampling order, an interleave sampling order, a reverse sampling order, a random sampling order, or a hybrid sampling order.

18. The method of claim 2 , wherein acquiring the first partial k-space dataset for the target area using the MRI scanner comprises controlling the MRI scanner to acquire the first partial k-space dataset with a fraction of less than 0.8.

19. The method of claim 2 , wherein acquiring the first partial k-space dataset for the target area using the MRI scanner comprises controlling the MRI scanner to acquire the first partial k-space dataset with a fraction of less than 0.6.

20. The method of claim 2 , wherein acquiring the first partial k-space dataset for the target area using the MRI scanner comprises controlling the MRI scanner to acquire the first partial k-space dataset with a fraction of less than 0.4.