Algebraic reconstruction method for off-resonance and eddy-current correction in functional and diffusion weighted magnetic resonance imaging

Stanford researchers have designed a method to perform an off-resonance corrected MRI reconstruction by modeling the off-resonance terms as part of the image encoding process. By acquiring two sets of images with inverted readout polarity, the underlying geometric distortion map can be derived.

This method allows for a joint reconstruction of acquisitions with opposing readout polarity and thereby improves the conditioning of the reconstruction (inversion) process. This method also allows for an iteration between image domain methods and k-space based methods.

Existing methods correct for geometric distortions in the image domain after the reconstruction from the MR-signal is completed. The new method accounts for the distortions during the reconstruction step (k-space based).