Nonlinear motion artifact reduction in event-triggered gradient-echo FMRI

Event-triggered FMRI was designed to yield good spatial and time resolution on conventional hardware. However, due to longer overall measurement time this method is even more prone to motion artifacts than regular FLASH-based methods. We demonstrate a simple and efficient method to detect and reduce linear and non-linear motion artifacts by manipulating k-space MR data. Results show that (a) motion artifacts may be reduced from typically 10% to noise level (2–3%); (b) small, well-localized hemodynamic changes may be differentiated from motion artifacts; and (c) are visualized with high contrast using an appropriate model function. This way, hemodynamic changes detected by event-triggered FMRI can be quantitated reliably in the lower percent range.