Key points are not available for this paper at this time.
Anatomical structures that are displaced periodically during respiration are repeated as ghosts in magnetic resonance (MR) images. These ghosts can be suppressed in many ways: the averaging of multiple sets of data, respiratory gating, deliberate positioning of ghosts, and respiratory ordering of phase encoding. Each method has a unique mechanism, which is described in detail. A theoretical investigation has been conducted into the effects that the methods have on the point spread function of a moving point. Data acquired in Fourier imaging are actually in the spatial frequency domain, so that respiratory motion can be regarded as a function of spatial frequency. The four methods above modify this functional dependence in different ways, allowing a unified comparison. Motion artifact suppression imposes additional constraints on image acquisition, which can prolong the imaging time. A technique has been developed that keeps the imaging time short by using the configuration of the subject to regulate the timing of image acquisition.
Wood et al. (Sat,) studied this question.