Group independent component analysis reveals consistent resting-state networks across multiple sessions

Abstract

Group independent component analysis (gICA) was performed on resting-state data from 14 healthy subjects scanned on 5 fMRI scan sessions across 16 days. The data were reduced and aggregated in 3 steps using Principal Components Analysis (PCA, within scan, within session and across session) and subjected to gICA procedures. The amount of reduction was estimated by an improved method that utilizes a first-order autoregressive fitting technique to the PCA spectrum. Analyses were performed using all sessions in order to maximize sensitivity and alleviate the problem of component identification across session. Across-session consistency was examined by three methods, all using back-reconstruction of the single session or single subject/session maps from the grand (5 session) maps. The gICA analysis produced 55 spatially independent maps. Obvious artifactual maps were eliminated and the remainder were grouped based upon physiological recognizability. Biologically relevant component maps were found, including sensory, motor and a 'default-mode' map. All analysis methods showed that components were remarkably consistent across session. Critically, the components with the most obvious physiological relevance were the most consistent. The consistency of these maps suggests that, at least over a period of several weeks, these networks would be useful to follow longitudinal treatment-related manipulations.

Evaluating the effective connectivity of resting state networks using conditional Granger causality

“…Our networks have been also reported in a number of previous resting state studies (Beckmann et al 2005;Chen et al 2008;Damoiseaux et al 2006;De Luca et al 2006;Mantini et al 2007;van den Heuvel et al 2008), although there is no complete consensus on the number and topology of the RSNs.…”

“…ICA is a method capable of separating independent spatio-temporal patterns of synchronized neural activity from fMRI data (Bartels and Zeki 2005), without prior knowledge about their activity waveforms or locations (McKeown et al 1998). Numerous authors demonstrated that voxels belonging to a given ICA network have higher BOLD temporal correlations among themselves compared with voxels belonging to other patterns, making ICA particularly suitable for functional connectivity studies (Beckmann et al 2005;Chen et al 2008;Damoiseaux et al 2006;De Luca et al 2006;Greicius et al 2004;Jafri et al 2008;Mantini et al 2007;Stevens et al 2009). Using ICA, we obtained reproducible RSN spatial maps across subjects, along with their associated time-courses, which could be used for effective connectivity analysis.…”

“…Interestingly, such patterns, namely resting state networks (RSNs), recapitulate the functional architecture of somato-motor, visual, auditory, attention, language, and memory networks that are commonly modulated during active behavioral tasks (Bartels and Zeki 2005;Corbetta and Shulman 2002;. Despite there is not a consensus yet, a number of consistent RSNs, as detected by ICA, could be jointly reported in the same study, i.e., the default-mode network (DMN) Raichle et al 2001), and other task-positive networks corresponding to dorsal attention, visual, auditory, and sensorimotor processing (Beckmann et al 2005;Chen et al 2008;Damoiseaux et al 2006;De Luca et al 2006;Mantini et al 2007;van de Ven et al 2008). Interactions among these RSNs are still largely unexplored .…”

Detecting resting-state brain activity by spontaneous cerebral blood volume fluctuations using whole brain vascular space occupancy imaging

“…The human brain has been shown to be composed of multiple coherent networks that support sensory, motor and cognitive functions (Buzsáki and Draguhn, 2004;De Luca et al, 2006;Smith et al, 2009). These brain networks appear to be consistent across time within and between individuals (Chen et al, 2008;Damoiseaux et al, 2006), and constrained to anatomically connected regions (Greicius et al, 2009;Honey et al, 2009). Interestingly, the strength of functional connectivity in these networks at "rest" is able to predict relevant task-induced activation and behavioral performance (Hampson et al, 2006;Zou et al, in press).…”

Identification of Resting State Networks Involved in Executive Function

“…These networks have been shown to be consistent across time (Chen et al, 2008) and across subjects (Damoiseaux et al, 2006). While investigations into these ''intrinsic connectivity networks'' (Seeley et al, 2007) have harnessed other imaging modalities such as EEG and PET, fMRI has been the tool of choice for the majority of research in this area (Smith et al, 2009).…”