NeuroQLab DTI - Probabilistic Parameter Adaption for Efficient Fiber Tracking

NeuroQLab DTI is a software tool which we developed for efficient and robust fiber tractography and fiber quantification. It consists of deterministic and probabilistic fiber tracking algorithms and offers multiple options for seeding and filtering fiber tracts. This paper shows the fiber tracking results for three different patient data sets, published within the context of the 2015 DTI tractography challenge. All patients suffered from a tumor in the close vicinity of the precentral gyrus. The task of the challenge was to reconstruct the pyramidal tracts (left and right hemisphere) for all three patients. We propose to use a probabilistic approach which adapts its control parameters locally. This allows for a reconstruction of fiber bundles whose geometric properties and whose corresponding underlying diffusion processes vary throughout passing brain regions. For defining seed regions, we utilized a 3D picking mechanism so that seed ROIs can interactively be placed on the brain surface or within the brain. Our results show that the pyramidal tracts can reliably and efficiently be tracked for the given data sets. Fig. 1. The pyramidal tract is an accumulation of upper motor neuron fibers that start in the precentral gyrus and terminate either in the brainstem (corticobulbar tract) or in the spinal cord (corticospinal tract). The image data have been provided within the context of the MICCAI tractography challenge 2015 (patient 1).