Stefan Weick1,
2, Flavio Carinci, 13, Cord Meyer1, 2,
Felix A. Breuer4, Frederick Mantel2, Philip Kleine2,
Peter M. Jakob1, 4
1Department
of Experimental Physics 5, University of Wuerzburg, Wuerzburg, Bavaria,
Germany; 2Department of Radiation Oncology, University Hospital
Wuerzburg, Wuerzburg, Bavaria, Germany; 3MRB Research Center,
Wuerzburg, Bavaria, Germany; 4Research Center Magnetic Resonance
Bavaria e. V. (MRB), Wuerzburg, Bavaria, Germany
In this work, it is shown that image registration in combination with averaging allows for the reconstruction of 3D data sets of the human lung with improved SNR. Retrospective respiratory self-gating was first used to reconstruct 3D data sets of multiple respiratory phases. These data sets were then registered onto one target data set and finally averaged. The resulting data set shows improved SNR due to averaging. Additionally, the data set shows preserved motion artifact reduction although data from different respiratory phases were used within the reconstruction process.