Assessing cognitive mental workload via EEG signals and an ensemble deep learning classifier based on denoising autoencoders

• Published in: Computers in biology and medicine Vol. 109; pp. 159 - 170
• Main Authors: Yang, Shuo, Yin, Zhong, Wang, Yagang, Zhang, Wei, Wang, Yongxiong, Zhang, Jianhua
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.06.2019; Elsevier Limited
• Subjects: Accuracy; Algorithms; Artificial intelligence; Classification; Classifiers; Cognitive ability; Deep learning; EEG; Electroencephalogram; Electroencephalography; Human performance; Human-machine system; Internal Medicine; Mapping; Mental workload; Neural networks; Noise reduction; Nuclear power plants; Operator performance; Other; Physiology; Preservation; Stacked denoising autoencoder; Task complexity; Workload; Workloads; Deep learning; Electroencephalogram; Mental workload; Human-machine system; Stacked denoising autoencoder
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2019.04.034

To estimate the reliability and cognitive states of operator performance in a human-machine collaborative environment, we propose a novel human mental workload (MW) recognizer based on deep learning principles and utilizing the features of the electroencephalogram (EEG). To determine personalized properties in high dimensional EEG indicators, we introduce a feature mapping layer in stacked denoising autoencoder (SDAE) that is capable of preserving the local information in EEG dynamics. The ensemble classifier is then built via the subject-specific integrated deep learning committee, and adapts to the cognitive properties of a specific human operator and alleviates inter-subject feature variations. We validate our algorithms and the ensemble SDAE classifier with local information preservation (denoted by EL-SDAE) on an EEG database collected during the execution of complex human-machine tasks. The classification performance indicates that the EL-SDAE outperforms several classical MW estimators when its optimal network architecture has been identified. •Temporal and frequential features of EEG are filtered by deep denoising autoencoders.•Assessing cognitive mental workload via an ensemble autoencoders preserving local information.•The architecture of the deep ensemble learning network has been determined.