• Integration of neuropsychological tests, driving simulation, and eye movement analysis to assess driving ability. • Saccade amplitude is a promising marker distinguishing mild/severe brain injury patients from healthy controls. • Findings support the development of evidence-based fitness-to-drive assessments for clinical and rehabilitation applications. Stroke survivors often wish to resume driving, but objective and reliable indicators for assessing fitness to drive are lacking. We aimed to establish effective measures for evaluating safe driving ability in brain-injured patients by integrating driving simulator (DS) performance and eye movement analysis. Participants included brain-injured patients, classified into mild and severe groups using Trail Making Test-B scores and the presence of visual field defects, neglect, or aphasia, alongside healthy controls. Neuropsychological assessments (Trail Making Test-A/B, Kohs Block Design Test, Mini-Mental State Examination) were conducted. Driving performance was evaluated using the Honda Safety Navi DS, focusing on the standard deviation (SD) of steering angle on straight roads and the SD of velocity on curved roads. Eye movements were recorded with Tobii Pro Glasses 2 during hazard detection and dangerous situation scenes, with particular attention to saccade amplitude. Group differences were analyzed using Kruskal–Wallis and Mann–Whitney U tests. The SD of steering angle on straight roads and the SD of velocity on curved roads were significantly higher in mild or severe brain-injured groups compared with healthy controls. However, saccade amplitude was significantly lower in both mild and severe brain-injured groups than in healthy controls during hazard detection scenes, indicating impaired visual exploration. To measure driving ability in both mild and severe brain-injured patients, saccade amplitude provided a promising objective indicator for evaluating driving ability instead of DS alone. These findings support the development of evidence-based fitness to drive assessments for clinical and rehabilitation applications.
Inoue et al. (Thu,) studied this question.