An Approach to Classification of Mental Fatigue based on Electroencephalography (EEG) Signals

• Mylena N. M. R. Ferreira UFPA / Instituto Senai de Inovação
• Ana C. Q. Siravenha Instituto Senai de Inovação
• Schubert R. Carvalho Instituto Tecnológico Vale
• Bruno D. Gomes UFPA
• Ronaldo F. Zampolo UFPA
• Agostinho S. Castro UFPA
• Adriana R. G. Castro UFPA

DOI: https://doi.org/10.5753/kdmile.2018.27387

Abstract

The complexity of the analysis of mental fatigue in healthy people is evidenced by the absence of abrupt disturbances in the electroencephalography signal and by the uniqueness and variability of the cognitive profile of each individual. Identifying this type of mental state requires the analysis of factors that characterize it, such as the behavior of frequency bands and brain regions. This work proposes to classify mental fatigue based on the analysis of frequency bands and reasons for these bands in two machine learning models: Multiple-layer Perceptron Neural Network and chained self-associative Neural Networks. Three frequencies and four ratios were calculated from the electroencephalographic data in terms of spectral energy density: α, β, θ, and the θ / α, (α + θ) / β, β / α and (α + θ) ratios / (α + β). We also propose a strategy for channel selection based on Wilcoxon's statistical significance between samples of normal and fatigued data. In addition, the normalization of the vector of characteristics is used in order to reduce the variability of the data and improve the characterization of the states. Tests show that the use of normalization effectively increases the accuracy of the classification, regardless of the model used. The selection of channels reduced the number of sensors from 30 to 11 and slightly impacted the accuracy of the models. The maximum accuracy of 99, 97% was achieved when using standardized data with channel selection, trained with self-associative Neural Networks.

Keywords: Electroencephalography, Artificial neural networks, Fatigue, Selection of characteristics

References

Bourlard, H. and Kamp, Y. Auto-association by multilayer perceptrons and singular value decomposition. Biological cybernetics 59 (4-5): 291–294, 1988.

Brookhuis, K. A. and De Waard, D. The use of psychophysiology to assess driver status. Ergonomics 36 (9): 1099–1110, 1993.

Cajochen, C., Brunner, D. P., Krauchi, K., Graw, P., and Wirz-Justice, A. Power density in theta/alpha frequencies of the waking eeg progressively increases during sustained wakefulness. Sleep 18 (10): 890–894, 1995.

Chalder, T., Berelowitz, G., Pawlikowska, T., Watts, L., Wessely, S., Wright, D., and Wallace, E. Development of a fatigue scale. Journal of psychosomatic research 37 (2): 147–153, 1993.

Cheron, G., Petit, G., Cheron, J., Leroy, A., Cebolla, A., Cevallos, C., Petieau, M., Hoellinger, T., Zarka, D., Clarinval, A.-M., et al. Brain oscillations in sport: toward eeg biomarkers of performance. Frontiers in psychology vol. 7, pp. 246, 2016.

Dimitrakopoulos, G. N., Kakkos, I., Dai, Z., Wang, H., Sgarbas, K., Thakor, N., Bezerianos, A., and SUN, Y. Functional connectivity analysis of mental fatigue reveals different network topological alterations between driving and vigilance tasks. IEEE Transactions on Neural Systems and Rehabilitation Engineering 4320 (c): 1–10, 2018.

Eoh, H. J., Chung, M. K., and Kim, S.-H. Electroencephalographic study of drowsiness in simulated driving with sleep deprivation. International Journal of Industrial Ergonomics 35 (4): 307–320, 2005.

Helton, W. S. and Russell, P. N. Working memory load and the vigilance decrement. Experimental Brain Research 212 (3): 429–437, 2011.

Hinton, G. E. and Zemel, R. S. Autoencoders, minimum description length and helmholtz free energy. In Advances in neural information processing systems. pp. 3–10, 1994.

Jap, B. T., Lal, S., Fischer, P., and Bekiaris, E. Using eeg spectral components to assess algorithms for detecting fatigue. Expert Systems with Applications 36 (2): 2352–2359, 2009.

Kohavi, R. A study of cross-validation and bootstrap for accuracy estimation and model selection. In Proceedings of the 14th International Joint Conference on Artificial Intelligence - Volume 2. IJCAI’95. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA, pp. 1137–1143, 1995.

Li, J., Lim, J., Chen, Y., Wong, K., Thakor, N., Bezerianos, A., and Sun, Y. Mid-Task Break Improves Global Integration of Functional Connectivity in Lower Alpha Band. Frontiers in Human Neuroscience 10 (June): 1–12, 2016.

Liu, Y. T., Wu, S. L., Chou, K. P., Lin, Y. Y., Lu, J., Zhang, G., Lin, W. C., and Lin, C. T. Driving fatigue prediction with pre-event electroencephalography (EEG) via a recurrent fuzzy neural network. 2016 IEEE International Conference on Fuzzy Systems, FUZZ-IEEE 2016, 2016.

Min, J., Wang, P., and Hu, J. Driver fatigue detection through multiple entropy fusion analysis in an EEG-based system. PLOS ONE 12 (12): e0188756, dec, 2017.

Pyun, H. and Kim, J. A study on the effect of emotion-evoking advertisement with eeg analysis. In Proceedings of 2000 Joint Conference of KIIE and KORMS, KIIE and KORMS, Seoul. Vol. 413416, 2000.

Schier, M. A. Changes in eeg alpha power during simulated driving: a demonstration. International Journal of Psychophysiology 37 (2): 155–162, 2000.

Scott G. Paris, A. H. P. Classroom applications of research on self-regulated learning. Educational Psychologist 36 (2): 89–101, 2001.

Siravenha, A. C. and Carvalho, S. R. Plant classification from leaf textures. In Digital Image Computing: Techniques and Applications (DICTA), 2016 International Conference on. IEEE, pp. 1–8, 2016.

Wascher, E., Rasch, B., Sunger, J., Hoffmann, S., Schneider, D., Rinkenauer, G., Heuer, H., and Gutberlet, I. Frontal theta activity reflects distinct aspects of mental fatigue. Biological Psychology 96 (1): 57–65, 2014.