Predicting the Degree of Distracted Driving Based on fNIRS Functional Connectivity: A Pilot Study

Area of Science:

• Neuroscience
• Transportation Safety
• Biomedical Engineering

Background:

• Distracted driving is a leading cause of traffic accidents.
• Predicting driver attentional state can prevent distractions and improve safety.

Purpose of the Study:

• To develop a brain activity-based model for predicting the degree of driver distraction.
• To understand the neural basis of distracted driving.

Main Methods:

• Used functional near-infrared spectroscopy (fNIRS) to measure brain activity (oxyhemoglobin concentrations) in drivers.
• Constructed participant-specific regression models using functional connectivity and brake reaction time (BRT).
• Analyzed and clustered accurate prediction models using hierarchical clustering.

Main Results:

• Developed a predictive model with a mean absolute error of 5.58 x 10^2 ms for BRT across 12 participants.
• Identified common functional connectivity patterns involving dorsal attention network (DAN), sensory-motor network (SMN), and ventral attention network (VAN).
• Confirmed multiple types of prediction models with varying network connectivity patterns.

Conclusions:

• Driver distraction degree can be predicted using brain activity during real driving.
• DAN, SMN, and VAN connectivity are crucial for predicting distraction in complex driving tasks.
• Findings contribute to developing safer driving systems and understanding the neural basis of distraction.