Open AccessReal Time Sleep Onset Detection from Single Channel EEG Signal Using Block Sample Entropy
Author(s) -
Talha Zobaed,
Saadaldeen Rashid Ahmed,
Abu Saleh Musa Miah,
Salma Masuda Binta,
Mohammed Rashid Ahmed,
Mamunur Rashid
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/928/3/032021
Subject(s) - computer science , novelty , sample entropy , real time computing , classifier (uml) , microcontroller , pattern recognition (psychology) , artificial intelligence , electroencephalography , speech recognition , computer hardware , psychology , psychiatry , philosophy , theology
In recent years, driver’s temporary state has been one in each of the foremost causes of road accidents and would possibly lead to severe physical damaging, mortality and necessary and noticeable economic losses. Maximum road accidents possible to avoided, if possible, to properly monitored driver’s drowsiness and a system are given warnings. In this work, a simple and inexpensive method has been offered to detect driver’s drowsiness or sleep onset detection with single channel EEG signal analysis. The key novelty of this work is to identify the sleep onset detection from a publicly available graph signal dataset by exploitation only one feature, simply implementable filter in any microcontroller device or smartphone and a threshold based mostly classification. Since, threshold-based classification techniques don’t need to train the classifier, hence, new subject adaptation is comparatively easier and real time implementation is more feasible. This novel approach can be easily implemented in smartphone to design and expand a drowsiness detection and alarming system for vehicle’s driver. On a variety of subjects, the experimental results show 95.68% accuracy.