The broad goal of the proposed research is to further understand the role of intrinsic brain networks in cognition. Intrinsic networks (INs) are collections of disparate brain regions that are consistently identified in task-free functional magnetic resonance imaging (fMRI), and are presumed to underlie sensory, motor, and cognitive functions. Further understanding requires detailed knowledge about the dynamics of the networks, both in task-free """"""""resting state"""""""" as well as during cognitive function. What are the interactions between Ins and the traditional """"""""task-active"""""""" regions? Is there an integration of information along the networks that may subserve task responses? The current proposal focuses on using fMRI to elucidate dynamic properties of a particular IN known as the default-mode network (DMN).
In Specific Aim 1, we model directions of task-evoked information flow among the DMN and task activated regions. The chronometry of initial responses to a brief, attentionally demanding task will be obtained using onset latency analysis, and the evolution of temporal dependences in the (long) poststimulus interval will be quantified using time-varying Granger causality.
Specific Aims 2 and 3 pertain to methodological issues in studying INs using fMRI.
In Aim 2, we quantify the stationarity of coupling strength and phase differences among regions of the DMN in task-free resting state. Currently, the majority of studies define INs using algorithms that assume that the coupling of IN regions remains constant over time. We employ time-varying functional connectivity and ICA over a long resting-state scan to examine the degree of change overtime, and relate changes in connectivity strength to physiological variables that may signify changes in arousal or awareness level.
In Aim 3, we correct for regional differences in hemodynamic latency due to vascular reactivity, so that the relative timing of BOLD signals reflects underlying neural interactions more closely. Our approach is to measure vascular response latencies across the brain using a breath holding task, and develop methods for correcting for these non-neural latency differences. Investigating the temporal behavior of the DMN will provide valuable insight into how the network underlies cognitive dysfunction, as well as function. While the DMN is receiving widespread attention for its purported role in episodic memory and task performance, it is also proving effective as a biomarker for disorders such as Alzheimer's Disease, major depression, ADHD, and schizophrenia. The current proposal provides a framework for studying the temporal behavior of the DMN (and other INs), which will serve as a basis for further investigation of network responses and connectivity in patient populations.
