Frontiers in Systems Neuroscience Systems Neuroscience

Abstract

magnetic resonance imaging (fMRI) in humans, and electrophysiology in anesthetized cats – with results in each case demonstrating that SOA exhibits reproducible spatiotemporal patterns that can be linked to underlying neural circuits. Using fMRI, Biswal et al. (1995) demonstrated that spontaneous activity within a functional sensorimotor network showed strong covariation even when that network was completely at rest, a phenomenon they dubbed “functional connectivity” based largely on previous electrophysiological work (Gerstein and Perkel, 1969; Gochin et al., 1991; Friston, 1994). In the same year, Arieli et al. (1995) used electrophysiological and optical techniques to show that patterns of intrinsic electrical activity in the visual cortex of anesthetized cats is coordinated at spatial scales up to several millimeters. Over the next decade, fMRI studies in humans and electrophysiological studies in animals elaborated the degree of organization of SOA signals. Functional connectivity computed from fMRI collected during rest revealed multiple distinct “networks” of covarying (i.e., functionally connected) areas (for a review, see Fox and Raichle, 2007). Early studies focused on cortical networks (Lowe et al., 1998; Greicius et al., 2003; Fox et al., 2005), with more recent ones also demonstrating subcortical networks (Di Martino et al., 2008; Zhang et al., 2008; O’Reilly et al., 2010). Relatively few imaging studies have been conducted in animal models to date, but the basic pattern of resting state networks appears similar in monkeys IntroductIon There is abundant activity in the brain in the absence of explicit sensory input or behavioral output, a phenomenon that is commonly observed in both electrophysiological and brain imaging experiments. In fact, most of the brain’s enormous energy expenditure appears to arise from spontaneously driven, intrinsic processes, rather than from the interaction with the sensory environment. Sensory stimulation increases local energy consumption above this background of high metabolic activity by only a few percent (Clarke and Sokoloff, 1999; Shulman et al., 2004; Raichle and Mintun, 2006). Yet despite its prominence, the neural processes associated with this spontaneous ongoing activity (SOA) have not been examined in detail until recently, and their significance for normal brain function is poorly understood. Moment-by-moment fluctuations in neural activity that cannot be ascribed to a stimulus or task event are typically treated as random “noise,” and are thus averaged out over multiple experimental trials. Analyzing spontaneous neural activity poses certain experimental challenges, as there are no clearly defined task events to serve as points about which to average. A common approach has therefore been to investigate the temporal covariation between pairs of signals simultaneously measured at different positions in the brain. Approximately 15 years ago, this approach was applied in two different branches of experimental neuroscience – functional Distinct superficial and deep laminar domains of activity in the visual cortex during rest and stimulation