Evaluation of Super-Resolution Methods for Magnetic Resonance Images

Abstract Super-resolution is the process of reconstructing a high-resolution magnetic resonance image (MRI) sequence from one or many low-resolution MRI sequences. In MRI, low-resolution images are usually interpolated to reduce the voxel size and enhance the perceptible resolution. Conventional interpolation methods are not able to recover the high-frequency information lost during the acquisition process. In this manuscript a novel super-resolution technique is proposed for the recovery of such information using coplanar high-resolution images. The proposed methodology benefits from the fact that in typical clinical settings both high- and low-resolution images of different types are taken of the same subject. These available high-resolution images can be used to enhance the resolution of other coplanar lower resolution images. Experiments on MRI data are described to show the efficacy of the proposed approach. An evaluation with conventional interpolation techniques is presented to illustrate the superior performance of the proposed methodology relative to preceding techniques.