Citation

Radüntz T, Fürstenau N, Mühlhausen T, Meffert B. Front. Physiol. 2020; 11: e300.

Affiliation

Signal Processing and Pattern Recognition, Department of Computer Science, Humboldt-Universität zu Berlin, Berlin, Germany.

Copyright

(Copyright © 2020, Frontiers Research Foundation)

DOI

10.3389/fphys.2020.00300

PMID

32372970

PMCID

PMC7186426

Abstract

In our digitized society, advanced information and communication technology and highly interactive work environments impose high demands on cognitive capacity. Optimal workload conditions are important for assuring employee's health and safety of other persons. This is particularly relevant in safety-critical occupations, such as air traffic control. For measuring mental workload using the EEG, we have developed the method of Dual Frequency Head Maps (DFHM). The method was tested and validated already under laboratory conditions. However, validation of the method regarding reliability and reproducibility of results under realistic settings and real world scenarios was still required. In our study, we examined 21 air traffic controllers during arrival management tasks. Mental workload variations were achieved by simulation scenarios with different number of aircraft and the occurrence of a priority-flight request as an exceptional event. The workload was assessed using the EEG-based DFHM-workload index and instantaneous self-assessment questionnaire. The DFHM-workload index gave stable results with highly significant correlations between scenarios with similar traffic-load conditions (r between 0.671 and 0.809, p ≤ 0.001). For subjects reporting that they experienced workload variation between the different scenarios, the DFHM-workload index yielded significant differences between traffic-load levels and priority-flight request conditions. For subjects who did not report to experience workload variations between the scenarios, the DFHM-workload index did not yield any significant differences for any of the factors. We currently conclude that the DFHM-workload index reveals potential for applications outside the laboratory and yields stable results without retraining of the classifiers neither regarding new subjects nor new tasks.

Copyright © 2020 Radüntz, Fürstenau, Mühlhausen and Meffert.

Language: en

Keywords

air traffic controllers; biomedical signal processing; electroencephalography; mental workload; pattern recognition; psychophysiology; state monitoring