Study And Application Of Alertness Monitoring Methods Based On EEG

Alertness refers to the ability of people to maintain their focus of attention and to remain alert to perform a task.It may affect our normal life,and even bring great harm to our lives and property safety if alertness is in a relatively low state when people are working normally.At present,electroencephalogram(EEG)is one of the most widely used physiological indices in the study of alertness detection.Some characteristics of EEG signals vary with the degree of human's alertness.Therefore,alertness level can be monitored by assessing the characteristic changes of EEG signals.In this paper,we mainly study the characteristics of EEG signals in three applications of alertness detection.Three different ways to induce fatigue were used: stimulated driving task,sleep deprivation and use of a visual brain-computer interface(BCI),which provided useful information for the study of EEG-based alertness monitoring in other applications.The main contents of this paper include the following three aspects:1?The fatigue was induced by a simulated driving task.Firstly,the simulated driving dataset from Taiwan National Chiao Tung University(NCTU)was used to develop the EEG-based alertness-monitoring algorithm.Then software for simulated driving was designed in a Unity 3D environment for performing the simulated driving experiment.The EEG data and behavioral data were collected simultaneously and analyzed by the same algorithm.The results verify the effectiveness and efficiency of the simulated driving platform and the developed alertness-monitoring algorithm.2?The fatigue was induced by sleep deprivation,which lasted 36 hours.EEG signals in eyes-open and eyes-closed resting states and during a psychomotor vigilance task(PVT)were collected every 4 hours.Subjects were also asked to fill out 4 questionnaire scales in each recording at the beginning of the experiment.The reaction time in the PVT task indicated that subject's alertness decreased over time during sleep deprivation.Power spectral density(PSD)analysis was introduced to analyze fatigue-related changes of EEG.The relationships between EEG's PSD,reaction time,and data of scales were established.3?The fatigue was induced by mental workload with a certain amount of time and task to study alertness monitoring during the use of a visual BCI paradigm.The BCI paradigm is a spelling system based on steady-state visual evoked potential(SSVEP),which was usedto study whether subjects had increased fatigue and the fatigue influenced the performance of the BCI system.This study aimed to provide suggestions and guidelines on EEG-based alertness detection in BCI research.