Multimodal biomarkers based on brain dynamics, omics and clinical data for neuropathologies: Examples of stroke, ALS and bipolar disorder

Abstract:
Recent progress in neuroimaging techniques give access to the brain activity in vivo at the whole-brain level, which holds many promises for the study of cognition and neuropathologies. As an example, an active direction in clinical research is the extraction of markers for the prognosis of patient evolution using fMRI. To go beyond a phenomenological analysis, models have also been developed to go beyond a pure (black-box) machine-learning approach applied on the recorded signals, aiming to develop interpretable markers or signatures of the brain dynamics. In this presentation I will focus on analyzing the propagation dynamics within brain subnetworks, which is also used to uncover information processing during cognitive tasks. Then, I will discuss data-fusion strategies to incorporate transcriptomics and clinical data to whole-brain dynamics data, in order to go towards multimodal biomarkers and uncover pathological mechanisms.