Lucilio Cordero-Grande1, Anthony Price1, Jana Hutter1, Emer Hughes1, and Joseph V. Hajnal1
A 2D CG-SENSE framework is proposed aiming at an integrated treatment of the main error sources in SMS-EPI reconstruction. Our pipeline jointly estimates the sensitivity profiles, Nyquist ghosting parameters, and image to be unfolded. In addition, an artifact-SNR tradeoff is established at a pixel level. Assessment by a phantom experiment has shown that all the main functionalities of the method do help diminish reconstruction artifacts. Stable results have been obtained when applying the framework in a large cohort of motion corrupted fMRI and DWI neonatal studies.