Selection of input features across subjects for classifying crewmember workload using artificial neural networks

The issue of crewmember workload is important in complex system operation because operator overload leads to decreased mission effectiveness. Psychophysiological research on mental workload uses measures such as electroencephalogram (EEG), cardiac, eye-blink, and respiration measures to identify mental workload levels. This paper reports a research effort whose primary objective was to determine if one parsimonious set of salient psychophysiological features can be identified to accurately classify mental workload levels across multiple test subjects performing a multiple task battery. To accomplish this objective, a stepwise multivariate discriminant analysis heuristic and artificial neural network feature selection with a signal-to-noise ratio (SNR) are used. In general, EEG power in the 31-40-Hz frequency range and ocular input features appeared highly salient. The second objective was to assess the feasibility of a single model to classify mental workload across different subjects. A classification accuracy of 87% was obtained for seven independent validation subjects using neural network models trained with data from other subjects. This result provides initial evidence for the potential use of generalized classification models in multitask workload assessment.

[1]  Carolyne R Swain,et al.  Electrophysiological, behavioral, and subjective indexes of workload when performing multiple tasks: manipulations of task difficulty and training. , 1999, International journal of psychophysiology : official journal of the International Organization of Psychophysiology.

[2]  K. Ramachandran,et al.  Mathematical Statistics with Applications. , 1992 .

[3]  Glenn F. Wilson,et al.  Psychophysiological responses to changes in workload during simulated air traffic control , 1996, Biological Psychology.

[4]  G F Wilson,et al.  Air-to-ground training missions: a psychophysiological workload analysis. , 1993, Ergonomics.

[5]  Michael E. Smith,et al.  Monitoring Working Memory Load during Computer-Based Tasks with EEG Pattern Recognition Methods , 1998, Hum. Factors.

[6]  Hankins Tc,et al.  A comparison of heart rate, eye activity, EEG and subjective measures of pilot mental workload during flight. , 1998, Aviation, space, and environmental medicine.

[7]  Glenn F. Wilson,et al.  Workload Related Changes In Eye, Cardiac, Respiratory and Brain Activity During Simulated Air Traffic Control. , 1995 .

[8]  M. Sterman,et al.  Multiband topographic EEG analysis of a simulated visuomotor aviation task. , 1994, International journal of psychophysiology : official journal of the International Organization of Psychophysiology.

[9]  M. B. Sterman,et al.  Quantitative EEG patterns of differential in-flight workload , 1993 .

[10]  H. Jasper,et al.  Electrical Activity of the Brain , 1941 .

[11]  Glenn F. Wilson,et al.  Operator Functional State Classification Using Neural Networks with Combined Physiological and Performance Features , 1999 .

[12]  Kenneth W. Bauer,et al.  Feature screening using signal-to-noise ratios , 2000, Neurocomputing.

[13]  G. Meek.  Mathematical statistics with applications , 1973 .

[14]  H. Jasper,et al.  The ten-twenty electrode system of the International Federation. The International Federation of Clinical Neurophysiology. , 1999, Electroencephalography and clinical neurophysiology. Supplement.

[15]  Richard S.H. Mah,et al.  Pattern recognition using artificial neural networks , 1992 .

[16]  Glenn F. Wilson,et al.  Performance Enhancement with Real-Time Physiologically Controlled Adaptive Aiding , 2000 .

[17]  M. Sterman,et al.  Concepts and applications of EEG analysis in aviation performance evaluation , 1995, Biological Psychology.

[18]  M. B. Sterman,et al.  Topographic EEG Correlates of Good and Poor Performance in a Signal Recognition Task , 1993 .

[19]  M. Goldstein,et al.  Multivariate Analysis: Methods and Applications , 1984 .

[20]  A Gevins,et al.  Test–retest reliability of cognitive EEG , 2000, Clinical Neurophysiology.

[21]  P G Jorna,et al.  Heart rate and workload variations in actual and simulated flight. , 1993, Ergonomics.

[22]  Christopher M. Bishop,et al.  Neural networks for pattern recognition , 1995 .

[23]  William J. Ray.  EEG and chaos: Description of underlying dynamics and its relation to dissociative states , 1994 .

[24]  A. Gevins,et al.  Detecting transient cognitive impairment with EEG pattern recognition methods. , 1999, Aviation, space, and environmental medicine.

[25]  N Galley,et al.  The evaluation of the electrooculogram as a psychophysiological measuring instrument in the driver study of driver behaviour. , 1993, Ergonomics.

[26]  J. Doyle,et al.  Electroencephalogram correlates of higher cortical functions. , 1979, Science.

[27]  A. H. Roscoe,et al.  Heart rate as a psychophysiological measure for in-flight workload assessment. , 1993, Ergonomics.

[28]  G. Wilson,et al.  Cognitive task classification based upon topographic EEG data , 1995, Biological Psychology.

[29]  C. Wientjes.  Respiration in psychophysiology: methods and applications , 1992, Biological Psychology.

[30]  C D Wickens,et al.  Assessment of pilot performance and mental workload in rotary wing aircraft. , 1993, Ergonomics.

[31]  S. Hart,et al.  Development of NASA-TLX (Task Load Index): Results of Empirical and Theoretical Research , 1988 .

[32]  John A. Caldwell,et al.  Effects of Pilot Workload on EEG Activity Recorded during the Performance of In-Flight Maneuvers in a UH-1 Helicopter. , 1997 .

[33]  G F Wilson,et al.  The use of cardiac and eye blink measures to determine flight segment in F4 crews. , 1991, Aviation, space, and environmental medicine.