Estimation of Human Workload from the Auditory Steady-State Response Recorded via a Wearable Electroencephalography System during Walking

Abstract

Workload in the human brain can be a useful marker of internal brain state. However, due to technical limitations, previous workload studies have been unable to record brain activity via conventional electroencephalography (EEG) and magnetoencephalography (MEG) devices in mobile participants. In this study, we used a wearable EEG system to estimate workload while participants walked in a naturalistic environment. Specifically, we used the auditory steady-state response (ASSR) which is an oscillatory brain activity evoked by repetitive auditory stimuli, as an estimation index of workload. Participants performed three types of N-back tasks, which were expected to command different workloads, while walking at a constant speed. We used a binaural 500 Hz pure tone with amplitude modulation at 40 Hz to evoke the ASSR. We found that the phase-locking index (PLI) of ASSR activity was significantly correlated with the degree of task difficulty, even for EEG data from few electrodes. Thus, ASSR appears to be an effective indicator of workload during walking in an ecologically valid environment.

Selective Attention Measurement of Experienced Simultaneous Interpreters Using EEG Phase-Locked Response

“…On the other hand, the B group showed almost no significant differences between tasks. Previous studies reported that ITCs decrease with increasing difficulty during tasks of different difficulty while listening to 40-Hz pulses (Griskova-Bulanova et al, 2011;Yokota and Naruse, 2015;Yokota et al, 2017). As SI experience grows, interpreters focus on the target stimuli during multitasking (Strobach et al, 2015;Henrard and Daele, 2017;Van der Linden et al, 2018).…”

“…However, body movements and measurement sounds cause much noise. In recent years, the development of active electrodes and portable EEG has made it possible to measure brain function in situations closer to realistic environments, such as while walking (Yokota and Naruse, 2015 ; Yokota et al, 2017 ).…”

Visual load does not decrease the auditory steady‐state response to 40‐Hz amplitude‐modulated tones

“…In a recent study (Yokota et al, 2017), the load effect on ITC varied with time on task: The ITC was higher during low load than during high load only in the first half of the experiment. To determine whether the load effect varied over time, we conducted exploratory ANOVAs of load (2 levels: low vs. high) by block (4 levels), as shown in the Supporting Information.…”

“…Only a few previous studies have tested the effects of crossmodal attention on 40‐Hz ASSRs by manipulating the difficulty or load of the visual task. These visual tasks had several different load levels: detecting an easy versus difficult target (Parks, Hilimire, & Corballis, 2011), detecting versus discriminating a change in target brightness (de Jong, Toffanin, & Harbers, 2010), reading versus performing a visual search task (Griskova‐Bulanova, Ruksenas, Dapsys, Maciulis, & Arnfred, 2011), playing an easy versus difficult Tetris game (Roth et al., 2013), flying an airplane in a flight simulator during self‐reported low versus high mental workload (Tsuruhara, Arake, Ogawa, Aiba, & Tomitsuka, 2015), and performing n ‐back tasks with different n to vary the difficulty (Yokota & Naruse, 2015; Yokota, Tanaka, Miyamoto, & Naruse, 2017).…”

“…Four studies showed a statistically significant decrease in ASSRs as a result of visual load. Of these studies, three showed smaller ASSR amplitude for high load than for low load (Roth et al., 2013; Tsuruhara et al., 2015; Yokota & Naruse, 2015), and two showed lower phase locking for high than for low load (Yokota et al., 2017, Yokota & Naruse, 2015). However, in other studies, either the effect was unclear, or there was no apparent statistically significant effect.…”