Driver drowsiness detection using the in-ear EEG.

Area of Science:

• Biomedical Engineering
• Neuroscience
• Transportation Safety

Background:

• Driver drowsiness is a major cause of road accidents, necessitating effective monitoring technologies.
• Existing methods using electroencephalogram (EEG) and peripheral signals have limitations in convenience or accuracy.
• In-ear EEG offers a less obtrusive alternative, but its effectiveness for drowsiness detection requires validation.

Purpose of the Study:

• To evaluate the performance of in-ear EEG for driver alertness-drowsiness classification.
• To compare in-ear EEG effectiveness against on-scalp EEG and peripheral physiological signals.
• To assess the viability of in-ear EEG as a practical drowsiness monitoring solution.

Main Methods:

• Collected in-ear EEG data alongside on-scalp EEG, electrocardiogram (ECG), photoplethysmogram (PPG), and galvanic skin response (GSR) from drivers.
• Performed classification analysis to distinguish between alertness and drowsiness states.
• Compared classification accuracy across different signal modalities and combinations.

Main Results:

• In-ear EEG achieved high classification accuracy, comparable to individual on-scalp EEG channels.
• Peripheral signals (ECG, PPG, GSR) showed competitive performance only when used in pairwise combinations.
• In-ear EEG demonstrated effectiveness comparable to established methods, despite its unobtrusive nature.

Conclusions:

• In-ear EEG is a viable technology for driver drowsiness monitoring.
• It offers a practical and accurate alternative to single-channel on-scalp EEG and individual peripheral signals.
• This technology holds potential for improving road safety through advanced drowsiness detection.