Classification of Bruxism Based on Time- frequency and Nonlinear Features of single channel EEG

Abstract

Abstract Background In the classification of Bruxism patients based on electroencephalogram (EEG), feature extraction is essential. The method of using multi-channel EEG and fusing electrocardiogram (ECG) and Electromyography (EMG) signal features has been proved to have good performance in Bruxism classification, but the classification performance based on single channel EEG signal is still unsatisfactory. Methods Extract time-domain, frequency-domain and nonlinear features based on single channel EEG signal to improve the classification performance of Bruxism. Five common bipolar EEG recordings from 2 bruxism patients and 4 healthy controls during REM sleep were analyzed. The time domain (mean, standard deviation, root mean squared value), frequency domain (absolute, relative and ratios power spectral density (PSD)), and non-linear features include (sample entropy) of different EEG frequency bands were extracted from five channels of participant. Fine tree algorithm was trained and tested for classifying sleep bruxism and healthy controls using five-fold cross-validation. Results Our results suggest that the C4P4 EEG channel was found to be most useful for classification of sleep bruxism and yielded 95.59% sensitivity, 98.44% specificity, 97.84% accuracy, and 94.20% positive predictive value (PPV). Conclusions Our research has proved that the proposed features are effective, further demonstrating the feasibility of using only one EEG channel to classify sleep Bruxism, and providing experimental basis for developing a portable automatic sleep Bruxism detection system.