Automated Detection of Mind Wandering: A Mobile Application

Abstract

There is growing interest in mindfulness-based training of attention. A particular challenge for novices is learning to sustain focused attention while ensuring that the mind does not wander. This paper presents the development of a tool for the automated detection of episodes of mind wandering (MW), on the basis of biosignals, while normal healthy participants engaged in brief mindfulness-based training (BMT) of attention. BMT required five 20-minute training sessions on consecutive days and entailed practice of breath-focused attention, a typical exercise in mindfulness-based techniques of stressreduction. Heart rate, respiratory rate, electrodermal and electromyographic activity were measured, and participants pressed a button to indicate the subjective detection of MW during training. The data showed that BMT did not influence our measures of stress but BMT was effective in reducing the frequency of subjectively detected MW events. The algorithm for offline detection of MW achieved an accuracy of 85%. Based on this algorithm, a mobile application was developed for automated MW detection in real-time. The application requires the use of easily placeable sensors, provides a new approach to the real-time MW detection, and could be developed further for use in MW-related investigations and interventions (such as mindfulness-based training of focused attention). Abstract:There is growing interest in mindfulness-based training of attention. A particular challenge for novices is learning to sustain focused attention while ensuring that the mind does not wander. This paper presents the development of a tool for the automated detection of episodes of mind wandering (MW), on the basis of biosignals, while normal healthy participants engaged in brief mindfulness-based training (BMT) of attention.BMT required five 20-minute training sessions on consecutive days and entailed practice of breath-focused attention, a typical exercise in mindfulness-based techniques of stress-reduction. Heart rate, respiratory rate, electrodermal and electromyographic activity were measured, and participants pressed a button to indicate the subjective detection of MW during training. The data showed that BMT did not influence our measures of stress but BMT was effective in reducing the frequency of subjectively detected MW events. The algorithm for offline detection of MW achieved an accuracy of 85%. Based on this algorithm, a mobile application was developed for automated MW detection in real-time. The application requires the use of easily placeable sensors, provides a new approach to the real-time MW detection, and could be developed further for use in MW-related investigations and interventions (such as mindfulness-based training of focused attention).

Mind Wandering in a Multimodal Reading Setting: Behavior Analysis & Automatic Detection Using Eye-Tracking and an EDA Sensor

“…Cheetham and colleagues [ 22 ] investigated the feasibility of using only EDA features for the automatic detection of mind wandering in the context of meditation. Using peaks observed from a low-pass filtered EDA signal as a feature, authors reported an area under the curve (AUC) of 0.81.…”

“…Complex and rather expensive techniques for mind wandering observation include fMRI [ 8 , 9 , 10 ] and EEG [ 2 , 11 , 12 , 13 ]. Simpler and more scalable detection can be accomplished using eye-trackers [ 4 , 14 , 15 , 16 , 17 , 18 , 19 ] and the measurement of electrodermal activity [ 20 , 21 , 22 ].…”

Bored, distracted, and forgetful: The impact of mind wandering and boredom on memory encoding

“…However, this marker may not be entirely reliable, as two studies show no differences in heart rate between mind wandering and being on-task (Ottaviani et al, , 2015. A few studies also pointed out that mind wandering could be associated with a lowered level of skin conductance response (SCR; Blanchard et al, 2014;Cheetham et al, 2016;Smallwood, Davies, et al, 2004;Smallwood et al, 2007). Concerning boredom, a similar pattern of increased HR (London et al, 1972;Merrifield & Danckert, 2014;Pattyn et al, 2008) and decreased SCR (Giakoumis et al, 2010;London et al, 1972;Merrifield & Danckert, 2014) has been reported.…”

“…Mind wandering and boredom also seem to be associated with a similar pattern of physiological modifications. Indeed, engaging in task-unrelated thoughts appears to be consistently tied with a short-term elevation in mean heart rate (HR; Cheetham et al, 2016;Pham & Wang, 2015;Smallwood, Davies, et al, 2004;Smallwood et al, 2007;. This elevation is considered to be a marker for emotional arousal supposedly generated by the personal salience of internal thoughts (Klinger, 1999).…”

Overview of Biosignal Analysis Methods for the Assessment of Stress

“…Through the appropriate categorization and detection of the activity of specific muscle groups, such as cheekbones, forehead, mouth or cheeks, a conclusion can be drawn about the psychological state of the person, as well as his level of stress or satisfaction [25,26]. However, it is not the safest method to draw conclusions and it is proposed to use it as an additional marker [27].…”