Algebraic reconstruction method for off-resonance and eddy-current correction in functional and diffusion weighted magnetic resonance imaging
Abstract:
Improved correction of MRI images is provided making use of three main principles. First, a hybrid representation of the image reconstruction problem is used where the raw MRI data is Fourier transformed in x to provide a x-ky-kz representation. Second, blip-up blip-down acquisition is performed to provide data having equal and opposite regular k-space encoding. Third, joint least squares reconstruction is performed using a singular value decomposition of a joint blip-up, blip-down encoding matrix. The encoding matrix includes at least the effects of regular k-space encoding and known phase variations due to system non-ideality. Preferably, suitable regularization of the encoding matrix based on its singular values is used to automatically remove poorly conditioned parts of the data.
Information query
Patent Agency Ranking
0/0