Applying deep learning to single-trial EEG data provides evidence for complementary theories on action control

Abstract

Efficient action control is indispensable for goal-directed behaviour. Different theories have stressed the importance of either attention or response selection sub-processes for action control. Yet, it is unclear to what extent these processes can be identified in the dynamics of neurophysiological (EEG) processes at the single-trial level and be used to predict the presence of conflicts in a given moment. Applying deep learning, which was blind to cognitive theory, on single-trial EEG data allowed to predict the presence of conflict in~95% of subjects 33% above chance level. Neurophysiological features related to attentional and motor response selection processes in the occipital cortex and the superior frontal gyrus contributed most to prediction accuracy. Importantly, deep learning was able to identify predictive neurophysiological processes in single-trial neural dynamics. Hence, mathematical (artificial intelligence) approaches may be used to foster the validation and development of links between cognitive theory and neurophysiology of human behavior.

Temporally-precise disruption of prefrontal cortex informed by the timing of beta bursts impairs human action-stopping

“…This result is highly consistent with the idea that the stop process is initiated via the right prefrontal cortex. It is also a relatively rare demonstration of a within-participant, single-trial, scalp EEG correlation with the executive control of human actions (for other examples, see: Cohen and van Gaal 2014 ; Vahid et al, 2020 ). A separate cohort participated in experiments 2 (MRI and EEG) and 3 (TMS).…”

Event-Related Variability is Modulated by Task and Development

“…Our approach has methodological limitations that could be overcome by future developments. For instance, the extraction of ERP templates in our case depended on a manual inspection, but more sophisticated algorithms for the temporal clustering of single-trial ERP topographies(Vahid et al, 2020) could be used and combined with our variability analyses schemes. The occurrence of large within-trial variations also need to be put in correspondence with microstate transitions or other approaches to describe ongoing dynamics of intrinsic functional connectivity(Lombardo et al, 2020) which also reveal power-law distributed statistics of fluctuations.…”

Evaluating Deep Learning EEG-Based Anxiety Classification in Adolescents with Autism for Breathing Entrainment BCI

“…Despite our binary FBCSP-SVM classifiers reaching a satisfactory overall classification accuracy of around 82% across all 4 condition pairs, there are several trade-offs pertaining to the use of SVM when compared to deep learning. Although SVMs require less optimizing parameters, do not suffer from the problem of local minima, and are less computationally demanding than neural networks, they are constrained to a small number of features [77], even when these features are extracted by algorithms [78]. In addition, SVMs cannot consider a robust set of EEG timepoints, rendering them unable to examine the EEG time domain, which is a critical dimension for analyses [78].…”

“…Although SVMs require less optimizing parameters, do not suffer from the problem of local minima, and are less computationally demanding than neural networks, they are constrained to a small number of features [77], even when these features are extracted by algorithms [78]. In addition, SVMs cannot consider a robust set of EEG timepoints, rendering them unable to examine the EEG time domain, which is a critical dimension for analyses [78]. Contrastingly, LSTMs are well able to handle temporal information, given their ability to choose to remember or discard information depending on contextual information.…”