Abstract #0416

Thijs Dhollander^{1,2},
Wim Van Hecke^{1,3}, Frederik Maes^{1,2}, Stefan Sunaert^{1,3},
Paul Suetens^{1,2}

^{1}Medical
Imaging Research Center (MIRC), K.U.Leuven, Leuven, Belgium; ^{2}Department
of Electrical Engineering (ESAT), K.U.Leuven, Leuven, Belgium; ^{3}Department
of Radiology, University Hospitals of the K.U.Leuven, Leuven, Belgium

The number of atlas construction methods for diffusion weighted imaging is at least equal to the number of possible models which can be reconstructed from the original data in q-space. Therefore, we aim at constructing an atlas before the reconstruction of any of these models: a hybrid diffusion imaging atlas containing signal functions for multiple shells of q-space. In this work, we show how this can be achieved by creating such an atlas (containing signal functions for b = 700, 1000, 2800) and reconstructing some possible models of diffusion (diffusion tensor and fiber orientation distribution function) from the result.