Machine Learning Models for the Classification of Sleep Deprivation Induced Performance Impairment During a Psychomotor Vigilance Task Using Indices of Eye and Face Tracking

Abstract

High risk professions, such as pilots, police officers, and TSA agents, require sustained vigilance over long periods of time and/or under conditions of little sleep. This can lead to performance impairment in occupational tasks. Predicting impaired states before performance decrement manifests is critical to prevent costly and damaging mistakes. We hypothesize that machine learning models developed to analyze indices of eye and face tracking technologies can accurately predict impaired states. To test this we trained 12 types of machine learning algorithms using five methods of feature selection with indices of eye and face tracking to predict the performance of individual subjects during a psychomotor vigilance task completed at 2-h intervals during a 25-h sleep deprivation protocol. Our results show that (1) indices of eye and face tracking are sensitive to physiological and behavioral changes concomitant with impairment; (2) methods of feature selection heavily influence classification performance of machine learning algorithms; and (3) machine learning models using indices of eye and face tracking can correctly predict whether an individual's performance is "normal" or "impaired" with an accuracy up to 81.6%. These methods can be used to develop machine learning based systems intended to prevent operational mishaps due to sleep deprivation by predicting operator impairment, using indices of eye and face tracking.

A methodology to detect pilot perception of warning information by eye movement data and deep residual shrinkage networks

“…When the features of the Part I, Part II and Part III were selected, the accuracy of the model to identify the warning information reached 89%. In previous studies, the research methods for judging human perception based on physiological features include statistical analysis and machine learning, and the accuracy ranges from 50% to 82% [41][42][43][44]. In contrast, the method proposed in this paper had a certain degree of improvement in accuracy.…”

Towards Classifying Cognitive Performance by Sensing Electrodermal Activity in Children With Specific Learning Disorders

“…Statistical analysis [14], [17]- [19], [27] and machine learning [8], [13], [18], [21], [28]- [31] were found as the most-used analysis techniques for the processed signals (to infer the psychological event from the physiological signal). Authors established relationships between EDA/HRV changes and cognitive states such as cognitive load [13], [14], [21], [23], [24], [32], [33], attention [8], [28], memory [19], autistic aspects [17], engagement [20], anxiety states and stress [27], and different cognitive tasks [30] under impairment conditions [31]. There are works recognizing cognition from EEG signals.…”

“…The most common way to identify cognitive states is to measure performance indices from tasks [19], [21], [27], [38], standardized tests [8], [14], [17], [30], [31], self-reports [13], [28], and expert observations [18]. However, these identification strategies are considered subjective and cannot be measured continuously [39].…”