Concussion Biomarkers: Novel Imaging and Advanced Statistics - Video

Despite major investments in scientific and clinical research, no significant progress in concussion pharmacotherapy has been achieved to date. Concussion heterogeneity has been identified as a major obstacle in advancing concussion care. Concussions are extensively heterogeneous in terms of patient characteristics, neuropathology, and symptom profile. Developing improved stratification strategies for future successful clinical trials requires parsing concussion heterogeneity in a way that accounts for the complex relationships between symptoms and brain structure. In this talk, I will outline a recent study that addresses this challenge. We applied a pattern learning approach on novel diffusion MRI data and symptom data from a large cohort of previously-concussed children. We found cross-demographic multi-tract multi-symptom relationships that captured more information than traditional approaches and predicted meaningful clinical outcomes in unseen data. These results both recapitulated well-known findings from the concussion literature and revealed new insights about white matter structure/symptom relationships.