Re-Thinking Image Artifacts: From Errors to A Valuable Source of Information

Motivation: Image artifacts are usually considered as unwanted signal contributions in reconstructed images. We want to demonstrate that artifacts can also be understood as an encoding mechanism which can provide additional useful information. Goal(s): We want to demonstrate the idea behind artifacts as an encoding scheme in different MRI scenarios. Approach: Anatomical data with object motion is acquired as well as Arterial Spin Labeling data using an EPI-based PROPELLER readout without spectral fat saturation. Results: It was demonstrated that motion can appear as an additional slice encoding mechanism and that water/fat separated images can be obtained from artifacts in PROPELLER data. Impact: This study might introduce a novel understanding of artifacts in MRI images. Instead of aiming for artifact prevention, we propose to think about potential additional (diagnostic) information about the object of interest which might be reconstructed from those image artifacts.