Benjamin Zahneisen1, Thimo Grotz1, Maxim Zaitsev1, Juergen Hennig1
The aim of this work is to estimate dynamic changes of precession frequencies (off-resonance map) from dynamic changes of the image phase of a reconstructed time series. The dynamic changes are measured relative to a reference time point. Feeding back the dynamic off-resonance map to the reconstruction results in a ~64% reduction of respiration fluctuations in the signal time series compared to ~40% for the DORK-method. The approach provides a spatial map of the off-resonances and therefore allows one to restrict the calculation of the field map to a ROI and exclude unwanted regions that also influence the objects phase.