Nonstationarity in the ‘resting brain's’ modular architecture

摘要: Task-free functional magnetic resonance imaging (TF-fMRI) has great potential for advancing the understanding and treatment of neurologic illness. However, as with all measures neural activity, variability is a hallmark intrinsic connectivity networks (ICNs) identified by TF-fMRI. This hampered efforts to define robust metric suitable biomarker We hypothesized that some this rather than representing noise in measurement process, related fundamental feature within ICNs, which their non-stationary nature. To test hypothesis, we used large (n=892) population-based sample older subjects construct well characterized atlas 68 regions, were categorized based on independent component analysis network origin, anatomical locations, meta-analysis. These regions then dynamic graphical representations brain sliding time window each subject. allowed us demonstrate nature brain’s modular organization assign region ‘‘meta-modular’’ group. Using grouping, compared dwell strong sub-network configurations default mode (DMN) between 28 Alzheimer’s dementia 56 cognitively normal elderly matched 1:2 age, gender, education. found differences others have previously observed disease can be explained DMN configurations, steady state magnitude. specific may also underlie TF-fMRI findings been described mild cognitive impairment who are at risk dementia.