Functional Connectivity Growth Charts

• Activity Number: 565
• Type: Topic Contributed
• Date/Time: Wednesday, August 4, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Philip T. Reiss*+ and Lei Huang and Eva Petkova and Michael P. Milham and F. Xavier Castellanos
• Companies: New York University and New York University and New York University and New York University and New York University
• Keywords: quantile regression ; penalized splines ; functional magnetic resonance imaging ; resting state ; smoothing parameter selection

Abstract

Recent studies have documented age-related changes in functional connectivity, and there is some evidence linking psychiatric disorders to abnormal developmental trajectories. This work has motivated proposals to establish growth charts for functional connectivity measures. Whereas growth chart methods ordinarily presuppose a known measure whose change is being charted, a key challenge in our context is to select appropriate pairs of regions of interest (ROIs) for which connectivity growth charts can usefully screen for disorders. This talk presents recent work in that direction, using data from the 1000 Functional Connectomes database. We develop a penalized spline method to estimate quantiles of resting-state functional connectivity between ROIs as a function of age. Our method features an improved approach to smoothness selection for nonparametric quantile regression.