Item

Abstract

Despite recent advances in tractography, the gap remains wide between the descriptions of white-matter pathways in the literature and the methods to reconstruct and study them from dMRI images. Here, we tackle this challenge by proposing a language to define white matter tracts, namely WMQLT, and a tool to automatically reconstruct pathways from their WMQLT queries. Our method is performant, flexible enough to allow defining tracts using multiple modalities, and allows to extend ROI-based reconstruction methods. Leveraging our language, we define 19 major brain tracts, alongside their subdivisions, and reconstruct them in a large population. We show that the shape of the reconstructed pathways, as well as their connectivity and lateralizations are in accordance with the current neuroanatomical literature. Finally, we showcase our technique in two scenarios: computing the functional subdivisions of a tract, and assessing the role of handedness and gender in the lateralization of language-related tracts.