Sleep loss and change detection in simulated driving

A Deep Learning Approach to Classify Sitting and Sleep History from Raw Accelerometry Data during Simulated Driving

Prolonged sitting and inadequate sleep can impact driving performance. Therefore, objective knowledge of a driver's recent sitting and sleep history could help reduce safety risks. This study aimed to apply deep learning to raw accelerometry data collected during a simulated driving task to classify recent sitting and sleep history. Participants (n = 84, Mean ± SD age = 23.5 ± 4.8, 49% Female) completed a seven-day laboratory study. Raw accelerometry data were collected from a thigh-worn accelerometer during a 20-min simulated drive (8:10 h and 17:30 h each day). Two convolutional neural networks (CNNs; ResNet-18 and DixonNet) were trained to classify accelerometry data into four classes (sitting or breaking up sitting and 9-h or 5-h sleep). Accuracy was determined using five-fold cross-validation. ResNet-18 produced higher accuracy scores: 88.6 ± 1.3% for activity (compared to 77.2 ± 2.6% from DixonNet) and 88.6 ± 1.1% for sleep history (compared to 75.2 ± 2.6% from DixonNet). Class activation mapping revealed distinct patterns of movement and postural changes between classes. Findings demonstrate the suitability of CNNs in classifying sitting and sleep history using thigh-worn accelerometer data collected during a simulated drive. This approach has implications for the identification of drivers at risk of fatigue-related impairment.

Change blindness in simulated driving in individuals with homonymous visual field loss

Individuals with homonymous visual field loss (HVFL) fail to perceive visual information that falls within the blind portions of their visual field. This places additional burden on memory to represent information in their blind visual field, which may make visual changes in the scene more difficult to detect. Failing to detect changes could have serious implications in the context of driving. A change blindness driving simulator experiment was conducted with individuals with HVFL (n = 17) and in those with normal vision (NV; n = 16) where changes (pedestrians appearing) were triggered based on the driver’s gaze location. Gaze was used to ensure that the location of the change was visible before and after the change occurred. There were wide individual differences in both vision groups, ranging from no change blindness to more than 33% of events. Those with HVFL had more change blindness than those with NV (16.7% vs. 6.3%, p < 0.001) and more change blindness to pedestrians appearing in their blind than seeing hemifield (34.6% vs. 10.4%, p < 0.001). Further, there was more change blindness for events appearing in the seeing hemifield for those with HVFL than normal vision (p = 0.023). These results suggest that individuals with HVFL may be more susceptible to failures of awareness, such as change blindness, than individuals with normal vision. Increased risk for failures of awareness may result in motor vehicle crashes where the driver fails to notice the other road user (looked-but-failed-to-see incidents).