Investigating the Performance of Sensor-Driven Biometrics for the Assessment of Cognitive Workload

Cognitive workload changes in the assessment of human performance have long been monitored through the use of subjective self-reports, physiological, or neuro-physiological measures. Wearable sensors allow us to monitor direct physiological changes associated with cognitive workload in real time. This thesis focuses on utilizing multiple physiological and neurological measures: functional near-infrared spectroscopy (fNIRS), eye-tracking, electrodermal activity, heart rate, and respiratory rate; in order to assess cognitive workload changes during different simulation-based tasks designed within the context of Federal Law Enforcement Training. The primary objective is to use multi-modal sensors and machine learning (ML) approach as an objective assessment method and to investigate which set of biometrics serves as the best predictor of cognitive workload. To achieve this objective, several biometrics, including prefrontal cortex (PFC) oxygenation changes, heart rate, and respiratory rate, were collected in both high and low workload conditions in simulated use of force (UOF) scenarios. These biometrics were then used to train logistic regression, feature mapping, support vector machine (SVM) and linear discriminant analysis (LDA) classifiers. F1-score and accuracy measures were used to evaluate the performance of these classifiers. Oxygenation changes in the left anterior medial PFC region and the combined biomarkers including oxygenation changes in the PFC and heart rate revealed the highest performing feature sets across all the classifiers. Future work should include the use of non-linear classification models and study of additional biomarkers to enhance training outcomes and performance.