Maokun Lin

Mental fatigue is a common phenomenon in our daily lives. Long-term fatigue can lead to a decline in a person’s operational functions and seriously affect work efficiency. In this paper, a method that recognizes the degree of mental fatigue based on relative band power and fuzzy entropy of Electroencephalogram (EEG) is proposed. The N-back experiment was used to induce mental fatigue in subjects, and the corresponding EEG signals were recorded during the experiment. A preprocessing method based on complementary ensemble empirical modal decomposition (CEEMD) and independent component analysis (ICA) was designed to remove noise from the raw EEG signal. The relative band power feature, which has been used extensively in fatigue recognition studies, was extracted from the EEG signals. Meanwhile, fuzzy entropy, a feature commonly used in attention recognition, was also extracted for fatigue recognition, based on previous findings that an increase in fatigue is accompanied by a decrease in attention. The two features were fed into an extreme gradient boosting (XGBoost) classifier to distinguish three different degrees of fatigue, which resulted in an average accuracy of 92.39% based on data from eight subjects. The promising results indicate the effectiveness of the proposed method in mental fatigue degree identification.

Epilepsy is a common neurological disorder with sudden and recurrent seizures. Early prediction of seizures and effective intervention can significantly reduce the harm suffered by patients. In this paper, a method based on nonlinear features of EEG signal and gradient boosting decision tree (GBDT) is proposed for early prediction of epilepsy seizures. First, the EEG signals were divided into two categories: those that had seizures onset over a period of time (represented by InT) and those that did not. Second, the noise in the EEG was removed using complementary ensemble empirical mode decomposition (CEEMD) and wavelet threshold denoising. Third, the nonlinear features of the two categories of EEG were extracted, including approximate entropy, sample entropy, permutation entropy, spectral entropy and wavelet entropy. Fourth, a GBDT classifier with random forest as the initial result was designed to distinguish the two categories of EEG. Fifth, a two-step “k of n” method was used to reduce the number of false alarms. The proposed method was evaluated on 13 patients’ EEG data from the CHB-MIT Scalp EEG Database. Based on ten-fold cross validation, the average accuracy was 91.76% when the InT was taken at 30 min, and 38 out of 39 seizures were successfully predicted. When the InT was taken for 40 min, the average accuracy was 92.50% and all 42 seizures selected were successfully predicted. The results indicate the effectiveness of the proposed method for predicting epilepsy seizures.