Non-local diffusion weighted image super-resolution using collaborative joint information

Due to the clinical durable scanning time, spatial resolution of diffusion-weighted magnetic resonance imaging (DWI) is highly limited. It has been evidenced that supper-resolution of DWI holds out the potential in both improving investigation of smaller white matter structure and reducing partial volume effect. In this paper, a new non-local DWI super-resolution method is proposed to increase image resolution of DWI. Based on non-local strategy, we take the advantage of the joint information by a new searching framework in the DWI volumes. Besides this, weighting scheme is also adapted for more accurate comparison as well. The proposed method was tested over in vivo DWI datasets quantitatively and qualitatively. Comparison with currently used method demonstrates the competitive results of our proposed approach in terms of improvements on diffusion weighted image reconstruction and estimation accuracy for diffusion tensor imaging (DTI)