Motion correction of magnetic resonance images using phase difference of two orthogonal acquisitions
AssigneeMayo Foundation for Medical Education and Research
InventorsEdward Brian Welch, Armando Manduca
Life of this patent
Estimating the term…
Abstract
Two magnetic resonance frequency domain data sets are acquired with the phase and frequency encoding gradient axes swapped. A k-space phase difference data set is computed from the two acquisitions and corrections for interview in-plane, rigid body translational motion are calculated from the phase difference information and a large set of simultaneous linear equations. The two acquired data sets are corrected with phase changes to compensate for the subject motion and then combined to form a single image.
Claims13
A method for correcting for in-plane, rigid body translational subject motion during data acquisition with a magnetic resonance imaging (MRI) system, the steps comprising:
- a) acquiring a first k-space data set with the MRI system from which an image of the subject may be reconstructed using a pulse sequence having a first imaging gradient direction arrangement in which a first imaging gradient is directed along a first axis and a second imaging gradient is directed along a second axis that is substantially orthogonal to the first axis;
- b) acquiring a second k-space data set with the MRI system from which an image of the subject may be reconstructed using a pulse sequence having a second imaging gradient direction arrangement in which the gradient directions are swapped such that the first imaging gradient is directed along the second axis and the second imaging gradient is directed along the first axis;
- c) calculating a k-space phase difference data set from phase information in the first and second k-space data sets;
- d) calculating corrections for subject motion using the k-space phase difference data set by solving a set of simultaneous linear equations that describe phase changes caused by subject motion;
- and
- e) applying the phase corrections produced in step d) to the first k-space data set.
Dependencies are inferred from the claim text.