Dmitry Kurzhunov1, Robert Borowiak1,2, Marco Reisert1, Philipp Wagner1, Axel Krafft1,2, and Michael Bock1
1University Medical Center Freiburg, Dept. of Radiology - Medical Physics, Freiburg, Germany, 2German Cancer Research Center (DKFZ), German Cancer Consortium (DKTK), Heidelberg, Germany
This work presents a comparison analysis of different
reconstruction techniques for quantification of 3D maps of the cerebral
metabolic rate of oxygen consumption (CMRO2) in human brain. Several
17O-MR 3D data sets of a healthy volunteer’s brain were acquired at a
clinical 3 Tesla MR system with inhalation of 70%-enriched 17O2
gas. Iterative image reconstruction procedures, e.g. where different
co-registered 1H MR image data sets of high spatial resolution act
as edge-preserving constraints, are compared and used to improve the image
quality and the precision of CMRO2 mapping. Anisotropic Diffusion as
non-Homogeneous Constraint (ADHOC) is shown to be superior.
