Abstract: In this talk, I’ll present some of the challenges inherent in the application of neuroimaging for personalized medicine, and in doing so, motivate Neuroblox, a new computational ecosystem we’re developing for that purpose. To provide background, I’ll discuss lessons learned from three complementary directions pursued by our team. The first applies methods adapted from statistical physics towards multi-scale modeling of neuroscience data. This research, using neuron-to-clinical scale imaging, aims to develop data-driven models of brain function that include insulin resistance and the role of neuron-glial control circuits governing metabolism, with direct applications for human brain aging and Alzheimer’s disease. The second builds upon the problem of quantifying allostatic regulation of neural and physiological control circuits associated with psychiatric disorders. The third is our work on achieving the fMRI signal/noise required for single-subject-level generative modeling. Finally, we conclude with some directions for the future, including scaling circuit discovery and community-wide curation of libraries of integrated models, experimentally validated across different species, populations, and applications.
Noyce Conference Room
US Mountain Time
Our campus is closed to the public for this event.
Lilianne Mujica-ParodiDirector of Laboratory for Computational Neurodiagnostics, SUNY Stony Brook University