Identification of mental fatigue in language comprehension tasks based on EEG and deep learning

Chunhua Ye,Zhong Yin,Chenxi Wu,Xiayidai Abulaiti,Yixing Zhang,Zhenqi Sun,Jianhua Zhang

DOI: https://doi.org/10.48550/arXiv.2104.08337

2021-04-14

Abstract:Mental fatigue increases the risk of operator error in language comprehension tasks. In order to prevent operator performance degradation, we used EEG signals to assess the mental fatigue of operators in human-computer systems. This study presents an experimental design for fatigue detection in language comprehension tasks. We obtained EEG signals from a 14-channel wireless EEG detector in 15 healthy participants. Each participant was given a cognitive test of a language comprehension task, in the form of multiple choice questions, in which pronoun references were selected between nominal and surrogate sentences. In this paper, the 2400 EEG fragments collected are divided into three data sets according to different utilization rates, namely 1200s data set with 50% utilization rate, 1500s data set with 62.5% utilization rate, and 1800s data set with 75% utilization rate. In the aspect of feature extraction, different EEG features were extracted, including time domain features, frequency domain features and entropy features, and the effects of different features and feature combinations on classification accuracy were explored. In terms of classification, we introduced the Convolutional Neural Network (CNN) method as the preferred method, It was compared with Least Squares Support Vector Machines(LSSVM),Support Vector Machines(SVM),Logistic Regression (LR), Random Forest(RF), Naive Bayes (NB), K-Nearest Neighbor (KNN) and Decision Tree(DT).According to the results, the classification accuracy of convolutional neural network (CNN) is higher than that of other classification methods. The classification results show that the classification accuracy of 1200S dataset is higher than the other two datasets. The combination of Frequency and entropy feature and CNN has the highest classification accuracy, which is 85.34%.

Artificial Intelligence,Machine Learning,Signal Processing

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to identify the mental fatigue state of operators through electroencephalogram (EEG) signals in language understanding tasks, so as to prevent the decline of operators' performance. Specifically, the research designed an experiment to detect the mental fatigue of operators in the human - computer interaction system. The research collected EEG signals from 15 healthy participants and used deep - learning methods (especially convolutional neural networks, CNN) to classify these signals, thereby identifying the fatigue state. The paper explored the influence of different feature combinations on classification accuracy and compared the effects of multiple classification methods. Eventually, it was found that the combination of frequency - domain and entropy features with CNN had the best classification effect, and its classification accuracy rate reached 85.34%.