The Relationship between Real-Time EEG Engagement, Distraction and Workload Estimates and Simulator-Based Driving Performance

Identifying potentially impaired drivers is often dependent upon using cognitive testing from a controlled environment (clinic, laboratory) to predict behavior in a dynamic and unpredictable real world driving environment. The goal of this study was to determine the feasibility, and validity, of using a wireless EEG system to ultimately differentiate between impaired and unimpaired drivers. We utilized the B-Alert X10 portable wireless EEG/ECG system and applied previously validated EEG algorithms estimating engagement, workload, and distraction within a sample of normal control (n = 10) and HIV seropositive individuals (n = 14). Participants underwent a 30-minute fully interactive driving simulation. Overall, the HIV+ group evidenced significantly higher distraction during the simulation. When grouped according to poor and good performers on the simulation (regardless of HIV serostatus), those performing worse on the simulation had higher distraction, with a trend for lower workload, levels. We then examined EEG profiles immediately preceding a crash. Prior to a crash, participants evidenced a significant increase in distraction ~ 10-14 seconds leading up to the crash; the greatest increase was seen in the HIV+ group. These preliminary data support the potential utility of using EEG data in patient populations to identify individuals who might be at risk for impaired drivin