Nonlinear Estimation of Cylindrically Symmetric Magnetic Susceptibility Anisotropy in Image Space Without a Rigid DTI Prior

Cynthia Wisnieff¹, Pascal Spincemaille², Yi Wang¹

In this work we explore methods of estimating magnetic susceptibility anisotropy, MSA, of the human brain in vivo in image space using only a shape constraint on the susceptibility tensor without imposing prior information on the fiber orientation. Estimation of cylindrically symmetric tensor anisotropy is explored in a carbon fiber phantom and human in vivo data. It was found that anisotropy estimation was similar with some sensitivity to noise in the nonlinear method compared to previously explored k-space MSA estimation methods.