Template ICA: Estimating Resting-State Networks from fMRI in Individual Subjects Using Empirical Population Priors

• Activity Number: 14 - Translational Methods for the Assessment of Brain Function
• Type: Invited
• Date/Time: Sunday, July 29, 2018 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Imaging
• Author(s): Amanda Mejia* and Yikai Wang and Brian Caffo and Ying Guo
• Companies: Indiana University and Emory University and Johns Hopkins University and Emory University
• Keywords: fMRI; big data; empirical Bayes; connectivity; expectation maximization; independent component analysis

Abstract

Independent component analysis (ICA) is commonly applied to fMRI data to identify resting-state networks (RSNs), regions of the brain that spontaneously coactivate. Due to high noise levels in fMRI, group RSNs are typically estimated by combining data from many subjects in a group ICA (GICA). Subject-level RSNs are then estimated by relating GICA results to subject-level fMRI data. Recently, model-based methods that estimate subject-level and group RSNs simultaneously have been shown to result in more reliable subject-level RSNs. However, this approach is computationally demanding and inappropriate for small group or single-subject studies. To address these issues, we propose a model-based approach to estimate RSNs in a single subject using empirical population priors based on large fMRI datasets. We develop an EM algorithm to obtain posterior means of subject-level RSNs, along with a faster approximate EM algorithm. We conduct a simulation study to examine the performance of both algorithms. Using fMRI data from the Human Connectome Project, we find that the proposed methods result in subject-level RSN estimates that are much more reliable than those of competing methods.