A data-driven method for analysis of associated time courses found from group Independent Component Analysis
Schmithorst V, Holland S
Children's Hospital Medical Center
A data-driven method for analyzing the associated time courses obtained from group Independent Component Analysis (ICA) of functional MRI (fMRI) data is proposed. The method can be used for between-groups analyses or correlations with covariates of interest in a General Linear Model (GLM) framework. Changes in the shape of the hemodynamic response function (HRF) yield different information than changes in the activation intensities. The effectiveness of the method, with greater sensitivity than a hypothesis-driven approach, is demonstrated via the investigation of age-related and performance-related effects in a cohort of normal children ages 5-18 performing a word-picture matching task.