Abstract:Sleep loss causes decrements in cognitive performance, which increases risks to those in safety-sensitive fields, including medicine and aviation. Mathematical models can be formulated to predict performance decrement in response to sleep loss, with the goal of identifying when an individual may be at highest risk for an accident. This work produces an Ensemble Mixed Effects Model that combines a traditional Linear Mixed Effects (LME) model with a semi-parametric, nonlinear model called Mixed Effects Random Forest (MERF). Using this model, we predict performance on the Psychomotor Vigilance Task (PVT), a test of sustained attention, using biologically motivated features extracted from a dataset containing demographic, sleep, and cognitive test data from 44 healthy participants studied during inpatient sleep loss laboratory experiments. Our Ensemble Mixed Effects Model accurately predicts an individual's trend in PVT performance, and fits the data better than prior published models. The ensemble successfully combines MERF's high rate of peak identification with LME's conservative predictions. We investigate two questions relevant to this model's potential use in operational settings: the tradeoff between additional model features versus ease of collecting these features in real-world settings, and how recent a cognitive task must have been administered to produce strong predictions. This work addresses limitations of previous approaches by developing a predictive model that accounts for interindividual differences and utilizes a nonlinear, semi-parametric method called MERF. We methodologically address the modeling decisions required for this prediction problem, including the choice of cross-validation method. This work is novel in its use of data from a highly-controlled inpatient study protocol that uncouples the influence of the sleep-wake cycle from the endogenous circadian rhythm on the cognitive task being modeled. This uncoupling provides a clearer picture of the model's real-world predictive ability for situations in which people work at different circadian times (e.g., night- or shift-work).

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

Computational Modeling of the Effects of Sleep Deprivation on the Vigilance Decrement

Accurate detection of acute sleep deprivation using a metabolomic biomarker—A machine learning approach

An ensemble mixed effects model of sleep loss and performance

Driving Drowsiness Detection With Eeg Using A Modified Hierarchical Extreme Learning Machine Algorithm With Particle Swarm Optimization: A Pilot Study

Prediction of Vigilant Attention and Cognitive Performance Using Self-Reported Alertness, Circadian Phase, Hours since Awakening, and Accumulated Sleep Loss

Estimating vigilance from the pre‐work shift sleep using an under‐mattress sleep sensor

Application of Machine Learning to Sleep Stage Classification

Tracking Vigilance Fluctuations in Real-Time: A Sliding-Window HRV-based Machine-Learning Approach

Sleepiness and safety: Where biology needs technology

Detection of microsleep states from the EEG: a comparison of feature reduction methods

Novel Tools for Driving Fatigue Prediction: (1) Dry Eeg Sensor and (2) Eye Tracker.

Generalized Deep Learning EEG Models for Cross-Participant and Cross-Task Detection of the Vigilance Decrement in Sustained Attention Tasks

A Hybrid Vigilance Monitoring Study for Mental Fatigue and Its Neural Activities

Machine Learning models for the Identification of Cognitive Tasks using Autonomic Reactions from Heart Rate Variability and Electrodermal Activity

Enhancing construction safety: predicting worker sleep deprivation using machine learning algorithms

Multi-sensor driver monitoring for drowsiness prediction

Estimate Vigilance in Driving Simulation Based on Detection of Light Drowsiness.

Exploring Neuro-Physiological Correlates of Drivers' Mental Fatigue Caused by Sleep Deprivation Using Simultaneous EEG, ECG, and fNIRS Data