Computational Neuroimaging of Human Brainstem at 9.4 T
BMBF Project
Upper panels are 3T measurements; left EVC and right SC:
a-b) Sinusoidal fits measure contrast;
c-d) CNR on surface models;
e-f) phase on surface models showing CNR averaged over the ROIs. Lower panel
g-h) show 9.4T measurements in SC, including average CNR
UHF MRI offers a tremendous, likely supra-linear, boost in SNR and CNR over conventional field strengths. Yet this merit comes at substantial expense both in cost of the scanner, and in the analysis necessary to take advantage of the clearer data. On the one hand, UHF MRI is useful in cortex because it allows us to push the boundaries of spatial and temporal resolution. On the other hand, UHF is particularly attractive for subcortical brain regions because CNR is an order-of-magnitude lower than cortex even in strongly responding nuclei like SC. In general, brainstem function is an extremely important but vastly under-studied research topic because of its physical inaccessibility in animal models, and its low CNR for fMRI in human studies. Here we propose a panoply of measurements, methods, and modeling that should open up brainstem imaging to more general use at UHF. This argument is strengthened by the increasing availability of 7T scanners at many research sites.
PI and Coordinator (USA):
David Ress, Baylor College of Medicine
PIs and Coordinator (Germany):
Klaus Scheffler, MPI for Biological Cybernetics & University of Tuebingen
Gisela Hagberg, MPI for Biological Cybernetics & University of Tuebingen
Marc Himmelbach, University of Tuebingen