ABSTRACT
Using a wearable, electronic, eye-frame-mounted ocular monitoring system that works in real time, we validated several oculometric measures as an index of driver inattention and drowsiness in several sleep disorder conditions. Thirty-one participants (nine controls, seven obstructive sleep hypopnea/apnoea (OSH/OSA), eight narcolepsy and seven attention deficit disorder with or without hyperactivity (ADD/ADHD)) completed a 36-hour sleep deprivation cycle. Every six hours, a drive simulator session was synchronised with oculometric and electroencephalography (EEG) measures, followed by objective (test of variables of attention (TOVA), maintenance of wakefulness test (MWT)) and subjective (Stanford sleepiness scale (SSS)) measures of alertness and performance. A sleep-deprivation effect was confirmed throughout all six sessions. Narcoleptic participants performed significantly worse in the drive simulator, followed by the ADD/ADHD group, while the control and OSH/OSA groups had similar performance. Accident events in the drive simulator had significant regression with percentage of eyelid closure over the pupil over time (PERCLOS), eye blink duration (EBD) and eye blink rate (EBR) at sessions four and five, accounting for a significant percentage of the variations in crash accidents and road-edge excursion. Oculometric variables were sensitive to sleep deprivation with predictive circadian effects on the performance/vigilance relationship. These results contributed to the early design of a composite oculometric fatigue index (COFI) and safety response algorithm, which can be integrated into different real-world operator domains.
