Improved probabilistic streamlines tractography by 2 nd order integration over fibre orientation distributions

Abstract

The probabilistic streamlines [1] approach is one of the most promising methods to perform diffusion-weighted tractography. It takes into consideration the inherent noisy nature of DW data, extends naturally to allow fibre tracking through regions containing crossing fibres, and can be implemented to be relatively fast. However, current implementations of this algorithm work by stepping along an orientation derived from the fibre orientation distribution (FOD) at the current point only, and can hence be viewed as first order integration methods. Such methods are known to ‘overshoot’, and this effect has indeed previously been reported for the deterministic streamlines algorithm [2]. Higher order (Runge-Kutta) integration methods have been proposed to improve the quality of tracking results obtained using tensor-based deterministic streamlines [3], but unfortunately these are not suitable for probabilistic multi-fibre fibre- tracking applications. In this study, we propose a novel higher order probabilistic streamlines method, based on 2nd order integration over fibre orientation distributions (iFOD2).