Decoding Study-Independent Mind-Wandering from EEG using Convolutional Neural Networks

Abstract

ObjectiveMind-wandering is a mental phenomenon where the internal thought process disengages from the external environment periodically. In the current study, we trained EEG classifiers using convolutional neural networks (CNN) to track mind-wandering across studies. ApproachWe transformed the input from raw EEG to band-frequency information (power), singletrial ERP (stERP) patterns, and connectivity matrices between channels (based on inter-site phase clustering, ISPC). We trained CNN models for each input type from each EEG channel as the input model for the meta-learner. To verify the generalizability, we used leave-N-participant-out crossvalidations (N=6) and tested the meta-learner on the data from an independent study for across-study predictions. Main resultsThe current results show limited generalizability across participants and tasks. Nevertheless, our meta-learned trained with the stERPs performed the best among the state-of-theart neural networks. The mapping of each input model to the output of the meta-learner indicates the importance of each EEG channel. SignificanceOur study makes the first attempt to train study-independent mind-wandering classifiers. The results indicate that this remains challenging. The current proposed stacking neural network design allows an easy inspection of channel importance and feature maps.