Dr. Chris Rorden (University of South Carolina Center for the Study of Aphasia Recovery (C-STAR))

My team has been exploring how brain imaging can predict recovery from stroke and guide therapy. Many of the novel open-source tools we created to solve these problems have broader applications to aid other neuroimaging research.

First, I will describe NiiVue, a domain-specific visualization tool that can work on any device and read the dominant formats of voxels, meshes, streamlines, and connectomes in our field. Providing a universal visualization tool allows users to employ the best processing pipeline for the task at hand and enables cloud analyses. Second, I describe how feature selection can aid machine learning for modest-sized clinical neuroimaging datasets.

Third, I describe the challenges and opportunities we faced when openly sharing large neuroimaging datasets of acute and chronic stroke. Finally, I will discuss new tools that can improve the spatial normalization of clinical neuroimaging data.

This event is also available via Zoom.