R2* mapping in the presence of macroscopic B0 field variations

R-2(Z) mapping has important applications in MRI, including functional imaging, tracking of super-paramagnetic particles, and measurement of tissue iron levels. However, R-2(Z) measurements can be confounded by several effects, particularly the presence of fat and macroscopic B-0 field variations. Fat introduces additional modulations in the signal. Macroscopic field variations introduce additional dephasing that results in accelerated signal decay. These effects produce systematic errors in the resulting R-2(Z) maps and make the estimated R-2(Z) values dependent on the acquisition parameters. In this study, we develop a complex-reconstruction, confounder-corrected R-2(Z) mapping technique, which addresses the presence of fat and macroscopic field variations for both 2D and 3D acquisitions. This technique extends previous chemical shift-encoded methods for R-2(Z), fat and water mapping by measuring and correcting for the effect of macroscopic field variations in the acquired signal. The proposed method is tested on several 2D and 3D phantom and in vivo liver, cardiac, and brain datasets. Magn Reson Med, 2012. (c) 2011 Wiley Periodicals, Inc.