Thalamocortical excitability modulation guides human perception under uncertainty

Abstract

Knowledge about the relevance of environmental features can guide stimulus processing. However, it remains unclear how processing is adjusted when feature relevance is uncertain. We hypothesized that (a) heightened uncertainty would shift cortical networks from a rhythmic, selective processing-oriented state toward an asynchronous (“excited”) state that boosts sensitivity to all stimulus features, and that (b) the thalamus provides a subcortical nexus for such uncertainty-related shifts. Here, we had young adults attend to varying numbers of task-relevant features during EEG and fMRI acquisition to test these hypotheses. Behavioral modeling and electrophysiological signatures revealed that greater uncertainty lowered the rate of evidence accumulation for individual stimulus features, shifted the cortex from a rhythmic to an asynchronous/excited regime, and heightened neuromodulatory arousal. Crucially, this unified constellation of within-person effects was dominantly reflected in the uncertainty-driven upregulation of thalamic activity. We argue that neuromodulatory processes involving the thalamus play a central role in how the brain modulates neural excitability in the face of momentary uncertainty.

The EEG spectral fingerprints of meditation and mind wandering differ between experienced meditators and novices

“…Interestingly, some of the EEG spectral modulations associated to drowsiness (steeper 1/f slope and reduced IAF) were also identified when comparing meditation and rest conditions in experienced meditation practitioners (see Figure 3G). In this way, both a relative decrease in IAF and a steeper 1/f slope have been previously associated to reduced arousal levels (Cantero et al, 2002;Kosciessa et al, 2021;Lendner et al, 2020;Mierau et al, 2017). Since a relative decrease in arousal can be expected during both drowsiness and a state of relaxation during meditation practice (Jagannathan et al, 2018;Matko et al, 2021;Petitmengin et al, 2017), we speculate that steeper 1/f slope and reduced IAF during meditation in experienced practitioners are reflective of reduced arousal due to relaxation.…”

“…Recent evidence has shown that the commonly termed '1/f trend' varies within subjects and is functionally relevant (Donoghue, Haller, et al, 2020;Kosciessa et al, 2021;Lendner et al, 2020;Voytek et al, 2015;Waschke et al, 2021). For example, the slope of the 1/f trend is thought to be reflective of excitation-inhibition balance in the brain (steeper slope is associated to greater inhibition) (R. Gao et al, 2017), thereby having important implications for cognition and consciousness (Kosciessa et al, 2021;Lendner et al, 2020;Ouyang et al, 2020). Crucially, it has been recently shown that changes in the 1/f trend of the EEG spectrum can conflate the estimation oscillatory activity (Donoghue, Dominguez, et al, 2020;Donoghue, Haller, et al, 2020).…”

“…Neural oscillations are embedded within non-oscillatory / scale free neural activity that has a 1/f-like distribution (decreased power in higher frequencies) (Donoghue, Haller, et al, 2020). Recent evidence has shown that the commonly termed '1/f trend' varies within subjects and is functionally relevant (Donoghue, Haller, et al, 2020;Kosciessa et al, 2021;Lendner et al, 2020;Voytek et al, 2015;Waschke et al, 2021). For example, the slope of the 1/f trend is thought to be reflective of excitation-inhibition balance in the brain (steeper slope is associated to greater inhibition) (R. Gao et al, 2017), thereby having important implications for cognition and consciousness (Kosciessa et al, 2021;Lendner et al, 2020;Ouyang et al, 2020).…”

Bilingualism: A Neurocognitive Exercise in Managing Uncertainty

“…Task performance may become more variable as people try new strategies to learn more about the context and seek out further information that could be used to make inferences (Hsu et al, 2005;Yu & Dayan, 2005; see also, Kosciessa et al, 2021).…”

“…This network interacts with broader networks involved in cognitive control, including the frontal-parietal network and the cingulo-opercular Downloaded from http://direct.mit.edu/nol/article-pdf/doi/10.1162/nol_a_00044/1927613/nol_a_00044.pdf by guest on 13 July 2021 network (including anterior cingulate cortex, supplementary motor area, and insula; T. Wu et al, 2020). Recent evidence suggests that the thalamaus may play a central role in cortical shifts that occur during decision making under uncertainty (Kosciessa et al, 2021). To give one example, when comparing situations with unexpected uncertainties, where there is risk that is unknown beforehand (e.g., a deck of cards where probabilities are unknown; also called ambiguous choices), to those with expected uncertainties, where risk is known beforehand (e.g., a familiar deck of cards where probabilities are known; also called risky choices), there is differential activation of frontal (orbitofrontal cortex) vs. striatal (basal ganglia, caudate) areas.…”