Large-cohort neuroscience database projects, such as the Cardiovascular Health Study and the Baltimore Longitudinal Study of Aging, have as a critical component neuroradiology variables, such as infarct-like lesions on magnetic-resonance examination of the brain. Statistical analysis of the images and functional variables is central to this lesion-deficit research. I will describe a scalable relational-database architecture for this brain-image database, will present preliminary results of analysis of these data, and will propose future work for analyzing the spatial and signal components, in conjunction with clinical variables.

Back to Workshop Schedule