Indices for Probabilistic Tractography through regions of fibre crossing
Williams G, Hosey T, Carpenter T, Harding S
University of Cambridge
In recent years many algorithms for probabilistic fibre tracking have been proposed. Conventionally, partial volume problems due to crossing fibres have been considered a source of noise in the data. Where information about fibre heterogeneity is available, the path of least deflection has been chosen and tracks are assumed not to bifurcate. Using a two fibre model we present an algorithm for calculating the probability of connection using tractography and simultaneously estimating the proportion of fibres within a seed voxel that connect to target voxels.