Stress-indicators and exploratory gaze for the analysis of hazard perception in patients with visual field loss

Abstract

The perception of traffic hazards is crucial for timely driving reactions. In this study we examine if the perception of hazardous objects can be determined by means of combined eye-tracking recordings and stress-related indicators in drivers suffering from binocular visual field loss. An increased rate of traffic accidents has been reported for this group of patients, which may lead to driving prohibition. The sole recording of gaze movements may not be sufficient for reliable identification of traffic object perception. Even if a traffic hazard is fixated by the driver, inattention or the presence of higher visual deficits may lead to an inappropriate driving response.We examined how stress-related indicators, i.e., skin conductance and heart rate, can complement and improve the eye-tracking based analysis of object perception. In a driving simulator study, we assessed the driving performance and hazard perception ability of subjects with binocular visual field loss (eight with homonymous visual field defects and seven with advanced glaucoma) and 15 healthy-sighted control subjects. Seven out of 15 subjects with visual field defects failed the driving assessment, eight passed the test. One healthy-sighted subject failed the driving test.Drivers with visual field defects showed a significant increase in skin conductance in 93% of situations with adequate reaction to the approaching hazard (69% for heart rate changes. 92% and 81% for healthy drivers, respectively). Furthermore, we identified five out of 120 total situations where the drivers looked at the traffic hazard but failed to react (look-but-not-react). Four of these situations were associated with a change in skin conductance, which can be interpreted as a perceived hazard, but an inadequate behavioral driving response. In one situation, no change in skin conductance was observed (look-but-not-perceived). In eight situations, although drivers did not fixate the traffic hazard, adequate driving reaction was observed. Four of these situations were also associated with changes in skin conductance (perceived-without-look). However, 43% and 44% of all significant increases in skin conductance and heart rate, respectively, occurred without the presence of a traffic hazard.In summary, our findings suggest that the sole analysis of the driver’s eye movements does not reliably reveal whether a traffic hazard was perceived. Additional recordings of stress-related parameters, such as heart rate, and especially skin conductance, can confirm hazard perception and be used as a complementary method for the evaluation of exploratory eye movements and driving behavior.