Premium
Real‐time adaptive sequential design for optimal acquisition of arterial spin labeling MRI data[Note *. * M: male. F: female. yr: year. ...]
Author(s) -
Xie Jingyi,
Clare Stuart,
Gallichan Daniel,
Gunn Roger N.,
Jezzard Peter
Publication year - 2010
Publication title -
magnetic resonance in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.696
H-Index - 225
eISSN - 1522-2594
pISSN - 0740-3194
DOI - 10.1002/mrm.22398
Subject(s) - computer science , adaptive sampling , sampling (signal processing) , fisher information , schedule , algorithm , scheduling (production processes) , a priori and a posteriori , arterial spin labeling , mathematical optimization , inversion (geology) , magnetic resonance imaging , mathematics , machine learning , statistics , computer vision , monte carlo method , medicine , paleontology , philosophy , filter (signal processing) , epistemology , radiology , operating system , structural basin , biology
An optimal sampling schedule strategy based on the Fisher information matrix and the D‐optimality criterion has previously been proposed as a formal framework for optimizing inversion time scheduling for multi‐inversion‐time arterial spin labeling experiments. Optimal sampling schedule possesses the primary advantage of improving parameter estimation precision but requires a priori estimation of plausible parameter distributions that may not be available in all situations. An adaptive sequential design approach addresses this issue by incorporating the optimal sampling schedule strategy into an adaptive process that iteratively updates the parameter estimates and adjusts the optimal sampling schedule accordingly as data are acquired. In this study, the adaptive sequential design method was experimentally implemented with a real‐time feedback scheme on a clinical MRI scanner and was tested in six normal volunteers. Adapted schedules were found to accommodate the intrinsically prolonged arterial transit times in the occipital lobe of the brain. Simulation of applying the adaptive sequential design approach on subjects with pathologically reduced perfusion was also implemented. Simulation results show that the adaptive sequential design approach is capable of incorporating pathologic parameter information into an optimal arterial spin labeling scheduling design within a clinically useful experimental time. Magn Reson Med, 2010. © 2010 Wiley‐Liss, Inc.