Quantifying the maladaptive neurophysiological correlates leading to lapses of attention during the SART: towards real-time mental state monitoring of mind-wandering

Mind-wandering (MW) constitutes one of the most ubiquitous mental activity humans engage in but it comes at a significant cost. Internal distractions are believed to be a leading cause of performance errors and unhappiness. A brain computer interface (BCI) able to predict the disengagement of attention, e.g. lapses of attention resulting from mindwandering episodes, harbors numerous useful applications. In this study, the SART was applied to quantify EEG correlates of lapses of attention to assess the viability of a BCI able to monitor attentional states in real-time. Both spatio-temporal classification and filter bank common spatial patterns were applied on a single-trial basis with accuracy reaching 92%. This work represents a potential step towards enhanced human-machine systems and BCI-based treatments of perseverative disorders such as depression.