Neural and cognitive correlates of performance in dynamic multi-modal settings

Abstract

The endeavour to understand human cognition has largely relied upon investigation of task-related brain activity. However, resting-state brain activity can also offer insights into individual information processing and performance capabilities. Previous research has identified electroencephalographic resting-state characteristics (most prominently: the individual alpha frequency; IAF) that predict cognitive function. However, it has largely overlooked a second component of electrophysiological signals: aperiodic 1/f activity. The current study examined how both oscillatory and aperiodic resting-state EEG measures, alongside traditional cognitive tests, can predict performance in a dynamic and complex, semi-naturalistic cognitive task. Participants’ resting-state EEG was recorded prior to engaging in a Target Motion Analysis (TMA) task in a simulated submarine control room environment (CRUSE), which required participants to integrate dynamically changing information over time. We demonstrated that the relationship between IAF and cognitive performance extends from simple cognitive tasks (e.g., digit span) to complex, dynamic measures of information processing. Further, our results showed that individual 1/f parameters (slope and intercept) differentially predicted performance across practice and testing sessions, whereby flatter slopes were associated with improved performance during learning, while higher intercepts were linked to better performance during testing. In addition to the EEG predictors, we demonstrate a link between cognitive skills most closely related to the TMA task (i.e., spatial imagery) and subsequent performance. Overall, the current study highlights (1) how resting-state metrics – both oscillatory and aperiodic - have the potential to index higher-order cognitive capacity, while (2) emphasising the importance of examining these electrophysiological components within more dynamic settings and over time.

Oscillatory and aperiodic neural activity jointly predict language learning

“…The effect of differing levels of 1/ f slope on, for instance, beta power and behavioural performance likely reflect more nuanced inter-individual differences in information processing capacities (Dziego et al, 2022; Immink, Cross et al, 2021; Thuwal, Banerjee, & Roy, 2021), which may explain behavioural gains that are otherwise related to the manifestation of oscillatory activity (e.g., Kepinska et al, 2017). For example, here we observed that a decrease in beta power predicted better behavioural performance for flexible rules, while the inverse was seen for fixed word order rules.…”

Prior context and individual alpha frequency (IAF) influence predictive processing during language comprehension

“…Additionally, whilst IAF influenced N400 amplitudes, we did not detect any effects relating to the 1/f aperiodic slope. Although we investigated 1/f as a between-subjects (trait-like) measure, in line with prior research (e.g., Demuru & Fraschini, 2020;Dziego et al, 2023;McSweeney et al, 2021;, the 1/f slope also shows state-like changes and has been linked with norepinephrine activity and phasic changes in neural gain (Pertermann et al, 2019;Rosenblum, 2023). Consequently, future research may benefit from examining 1/f on a trial-by-trial basis to determine whether changes relating to prediction may be more transient.…”

“…Despite periodic oscillatory components such as IAF traditionally being linked to cognitive processing, the importance of considering the aperiodic noise-like activity (particularly the 1/f slope of the distribution) that underlies the EEG spectra has been recently emphasised in the literature (Cross et al, 2022;Demuru & Fraschini, 2020;Dziego et al, 2023;Immink et al, 2021;Merkin et al, 2023;Ouyang et al, 2020;. Although prior literature recommends controlling for this activity to directly probe periodic (i.e., oscillatory) patterns (Donoghue et al, 2020), the 1/f component has been shown to predict information processing speeds (Ouyang et al, 2020), and to mediate the relationship between age and cognitive functioning (Pathania et al, 2022;, suggesting that it may be relevant for cognition.…”

Electrophysiological indices of individual differences in adult language learning

“…Individual alpha frequency (IAF) is associated with memory (Lebedev, 1994), language (Bornkessel et al, 2004;Bornkessel-Schlesewsky et al, 2015), attention (Angelakis et al, 2004, Klimesch et al, 1993, general cognitive ability (g-factor intelligence; Grandy et al, 2013a, Zakharov et al, 2020, and has recently been shown to modulate sleep-related memory consolidation (Chatburn et al, 2021;Cross et al, 2020b). Specifically, a growing body of work shows that IAF is a more sensitive measure of individual cognitive performance than behavioural measures alone in studying language comprehension (Bornkessel et al, 2004;Bornkessel-Schlesewsky et al, 2015;Kurthen et al, 2020), as well as in naturalistic paradigms of cognition (Dziego et al, 2022) The functional significance of IAF as a robust marker of information processing makes it a potentially informative metric for examining individual differences in L2 learning, as it is advantaged by its stable, trait-like characteristics (Posthuma et al, 2001) and test-retest reliability in samples of healthy individuals (Grandy et al, 2013b).…”