EXTended Rosette ACquisition Technique (EXTRACT) : a dynamic R2* mapping method using extended rosette trajectory

S. Lee, D. Noll, J. Fessler EECS, University of Michigan, Ann Arbor, MI, United States, Biomedical engineering, University of Michigan, Ann Arbor, MI, United States Introduction Dynamic R2* mapping has been used in several MRI applications, such as functional imaging [1-4], and tissue identification by bolus tracking [5]. In functional imaging, R2* mapping has several advantages. It provides better functional contrast than BOLD imaging. The comparison between different subjects and scans is easier, because it is not dependent on the scan parameters such as TR and TE, or on baseline drift. [3] In addition, the separation of inflow effect from BOLD signal is possible [4]. However, conventional dynamic R2* mapping [1-4] relies on multiple gradient echo (2 ~ 8 echoes) imaging methods, suffering from either significantly increased data acquisition time, or/and motion artifact and physiological noise. In this work, we propose the EXTended Rosette ACquisition Technique (EXTRACT), which can provide a short acquisition time of ~50ms, and dynamic estimates of T2*-weighted images, R2* maps and field inhomogeneity maps simultaneously. To demonstrate the accuracy of the method, we performed a phantom study and compared EXTRACT to a mono-exponential fit of 6-echo spin warp images. We also present a functional study result from BOLD imaging and the proposed method. Method Formulation of segmented image reconstruction scheme: The MR signal S(t) with monoexponential R2* decay can be modeled as a series of time-segmented data sets Si(t) (i=1...L)