Bridging the gap – Spontaneous fluctuations shape stimulus-evoked spectral power

Abstract

Spontaneous fluctuations of neural activity have been shown to influence trial-by-trial variation in perceptual, cognitive, and behavioural outcomes. This implies that these fluctuations affect stimulus-related neural processes, and hence should affect stimulus-evoked neural activity. However, the mechanisms by which spontaneous neural activity shapes stimulus-evoked neural activity have rarely been examined. Employing a large-scale magnetoencephalographic dataset, as well as an electroencephalographic replication dataset, we observed that for high-frequency power, high pre-stimulus activity leads to greater evoked desynchronization (negative interaction); in contrast, for low-frequency power, high pre-stimulus activity induces greater event-related synchronization (positive interaction). We show that both positive and negative interactions are manifest primarily in cortical oscillations, rather than scale-free activity, and can also be observed in the time domain. In summary, we demonstrate positive and negative spontaneous-evoked interaction in multiple electrophysiological processes; these mechanisms "bridge the gap" between spontaneous and evoked activity and provide novel insights into how spontaneous activity influences behaviour and cognition.

Ongoing Brain Activity and Its Role in Cognition: Dual versus Baseline Models

“…In addition, fMRI and EEG/MEG data demonstrate nonadditive rather than merely additive modulation of task-related activity by prestimulus and/or resting state activity (Hesselmann and others 2008; He 2013; Huang and others 2017; Sadaghiani and others 2010 2015; Wainio-Theberge and others 2020). Finally, recent studies demonstrate that alterations (i.e., decreases or increases) in rest/prestimulus task interaction may lead to major cognitive abnormalities as in schizophrenia (Northoff and others 2021; Northoff and Gomez-Pilar 2021) and anxiety disorders (Lucherini and others 2021).…”

“…A second component of ongoing activity can be found in prestimulus activity, that is, the neural activity immediately preceding the onset of a specific stimulus or task (i.e., the 100–1000 ms preceding stimulus onset) (Bai and others 2015; He 2013; Huang and others 2017; Wainio-Theberge and others 2021; Wolff and others 2019a, 2019b). Finally, a third component of ongoing activity can be found during task-related activity where it is present in what is also coined “background activity”: it may represent that component of task-related activity that, operating in the background, reflects the carryover and continuation of the ongoing activity during task states (Barnes and others 2009; Cole and others 2012, 2013, 2016; Di and Biswal 2019; Di and others 2013a, 2013b; Di and others 2020; Fair and others 2007; Ito and others 2020; Northoff and others 2010a, 2010b; Smith and others 2009) (see Fig.…”

“…Analogously, the spontaneous activity’s shapes task-evoked activity through its degree of scale-freeness (as measured with PLE), which is related to distinct aspects of self-consciousness (Huang and others 2016; Wolff and others 2019a, 2019b; Kolvoort and others 2020; Scalabrini and others 2017), animate social interaction through active touch (Scalabrini and others 2019), multisensory perception (Ferri and others 2015, 2017), and even complex learning chemistry formulae (Bongers and others 2020). Moreover, two recent studies using fMRI and MEG demonstrated that the spontaneous activity’s dynamics (e.g., its power law exponent, delta/alpha power, and oscillatory/fractal components) correlate with the degree of nonadditivity during pre/poststimulus interaction (Aru and others 2019; Huang and others 2017; Wainio-Theberge and others 2020; Wolff and others 2020).…”

“…Even though the dual model seems to be the dominant background assumption in current studies on cognition, it shows some inconsistencies with the empirical data. Various imaging studies using fMRI and EEG/MEG demonstrate that, on the purely neuronal level, spontaneous activity is present in all three: resting state (i.e., the absence of any overt task stimulus), prestimulus activity, and task-related activity (Bolt and others 2017; Cheng and others 2018; Cole and others 2012, 2014, 2016; He 2011; Huang and others 2017; Ito and others 2020; Mennes and others 2010, 2011; Wainio-Theberge and others 2020; Wolff and others 2019a, 2019b, 2020).…”

“…The baseline model is based on the following empirical data: nonadditive rather than additive interaction of rest and task states, suggesting that rest serves as internal reference for external task states (He 2013; Huang and others 2017; Wainio-Theberge and others 2020); occurrence of activation in DMN and deactivation in non-DMN regions, which is dependent upon the general task-context (Chen and others 2015, 2017; Simony and others 2016; Vatansever and others 2015, 2017); recent findings on the brain’s global activity showing how, during the rest-to-task transition, it modulates the frequency of the regional/network pattern as measured by dynamic CAP of DMN and non-DMN networks (Liu and others 2018; Zhang and others 2020); association of internally oriented cognition like self and mind wandering with activation in non-DMN (i.e., seemingly task-positive networks like salience and control-executive networks) (Christoff and others 2009b; Christoff and others 2016; Qin and others 2020; Scalabrini and others 2017, 2019; Smallwood and Schooler 2015; Vatansever and others 2020); and, at the same time, externally oriented cognition that also induces activation in DMN as a seemingly task-negative network (Chen and others 2015, 2017; Simony and others 2016; Vatansever and others 2015, 2017; Zhang and others 2020).…”