Motion Compensation with Floating Navigator and GRAPPA Operators

Wei Lin¹, Feng Huang¹, Yu Li¹, Charles Saylor¹, Arne Reykowski¹

A method for motion correction in multi-coil imaging applications, involving both data collection and reconstruction, is presented. Floating navigator (FNAV) method, which acquires a readout line along k_y≠0 line, is expanded to detect translation/rotation and non-consistent motion, using a correlation measure. The flexibility of GRAPPA operator is exploited by (a) extrapolating readout lines to fill in missing pie-slice of k-space caused by rotational motion, and (b) regenerating full k-space from a reduced dataset, therefore allowing subsequent correction or the replacement of rejected non-consistent views. In vivo turbo spin-echo brain imaging experiments demonstrate the correction of severe motion artifacts.