Abstract #1069

Storey P, Sodomsky M, Chavez s, Graham S

Spike noise is a common source of image artefact on MRI systems and is particularly problematic for diffusion tensor imaging (DTI). It is a problem that is difficult to trouble-shoot and often results in loss of data. Currently, there is no magnitude-based spike noise correction algorithm that can be directly applied to DTI. This paper presents a methodology for spike noise detection and correction based on a magnitude-derived pseudo-k-space. The methodology makes use of spike-free images to identify and correct the spike signature. It has been shown to work successfully on a wide range of spike noise occurrences.