Super Resolution Reconstruction from Differently Oriented Diffusion Tensor Datasets

Keywords

acquisition added affine angular appreciated averaging axial badly benefit benefits biomedical brain chosen clinical clinically close comfort composed conditioned consequently constant constraints coronal corrupted datasets decreased decreases described differently diffusion direct directly dirk enable encoded error evaluate even exhibit feasible fiber filters finer fractional free function generator gradient grid ground head hence imposes improved in vivo increasing inherently integrate intensity investigation keeping larger leading leads least likelihood limited loss making maps matrix measures media median metrics minimal minimizing modality model modeled molecules motion neglect newton noise noiseless noisy noninvasive nonlinear numerical optimal oriented orthogonal outputs partial parts patient pipeline preserving press problem processed projection promising proposed quality quantify quantitative reconstructed reconstruction regularization regularized relation represents required resolution respect sampling served sets simulate simulated simulation simulations slice slices solution solutions solved space spatial sqrt squared squares structures suffers super tensor thereby thinner trade transform transformation transformations transforms trust truth typically unidirectional unprecedented volume water