Theta oscillations shift towards optimal frequency for cognitive control

Abstract

Cognitive control is supported by theta band (4-7Hz) neural oscillations coordinating neural populations for task implementation. Task performance has been shown to depend on theta amplitude but a second critical aspect of theta oscillations, its peak frequency, has mostly been overlooked. Using modelling, behavioral and electrophysiological recordings, we show that theta oscillations adapt to task demands by shifting towards the optimal frequency..

Theta-Phase Connectivity between Medial Prefrontal and Posterior Areas Underlies Novel Instructions Implementation

“…In the context of instructions implementation, we hypothesized that this optimized format consists of strengthened connections between sensory and motor areas representing task-relevant information, coordinated by medial prefrontal structures through long-range phase synchronization. This operationalization is grounded in a body of computational work addressing the role of mPFC theta oscillations in flexibly binding task-relevant areas for upcoming task demands ( Verguts, 2017 ; Verbeke and Verguts, 2019 ; Verbeke et al, 2020 ; Senoussi et al, 2022 ).…”

“…Stemming from this literature, mPFC has been attributed a central role in determining and instantiating control policies in several computational models of proactive and reactive cognitive control ( Dosenbach et al, 2008 ; Botvinick and Cohen, 2014 ; Shenhav et al, 2017 ; Verguts, 2017 ; Holroyd and Verguts, 2021 ). In particular, in the work of Verguts and colleagues, mPFC theta oscillations signal the need for adjustments to reach the current goal ( Verguts, 2017 ; Verbeke and Verguts, 2019 ; Verbeke et al, 2020 ; Senoussi et al, 2022 ), and achieve them by synchronizing the activity of task-relevant pairs of sensory and action units, thereby allowing them to communicate more efficiently ( Fries, 2005 , 2015 ). Therefore, these models identify in the mPFC the structure responsible to coordinate and operationalize the flexible binding of lower-level modules to meet task demands.…”

“…According to the computational model proposed by Verguts (2017) and updated by his colleagues ( Verbeke and Verguts, 2019 ; Verbeke et al, 2020 ; Senoussi et al, 2022 ), theta oscillations from the mPFC control unit serve the purpose of synchronizing posterior task-relevant processing units. Therefore, we were primarily interested in investigating whether task demands (i.e., preparing for Implementation vs Maintenance) affect connectivity between frontal and posterior areas.…”

“…Theories and evidence emerged regarding slow oscillations as prime mechanism for establishing top-down neural communication and for efficiently setting up functional networks optimized to meet task demands (Bonnefond et al, 2017;McLelland & VanRullen, 2016;Riddle, Scimeca, et al, 2020;Riddle, Vogelsang, et al, 2020;Sauseng & Klimesch, 2008;Voloh & Womelsdorf, 2016). In line with this framework, recent computational models have attributed to theta oscillations a crucial role in flexibly binding posterior task-relevant areas (Senoussi et al, 2022;Verbeke et al, 2020;Verguts, 2017). Burst of theta-locked activity originating in the medial prefrontal cortex (mPFC) are directed towards the appropriate processing units (i.e., areas encoding task-relevant information), inducing the synchronization of their firing patterns and thus enabling efficient inter-areal communication (Fries, 2005(Fries, , 2015.…”

Neural Representations of Task Context and Temporal Order During Action Sequence Execution

“…One other potential reason why context is much more difficult to decode is that it might be the case that during relatively long sequences, as used in this coffee–tea making task, the brain first sets up the sequence of actions to execute based on contextual information and then mostly represent progression through that specific sequence in different brain areas. It would be very interesting in future studies to investigate whether other aspects of neural activity reflect task context, such as fast neural dynamics or connectivity across frontal brain regions (Formica, González‐García, Senoussi, & Brass, 2021; Senoussi et al., 2020; Smith et al., 2019), and what specific aspects of an extended task sequence could require a more evident or widespread representation of task context across brain areas.…”