Why do Auditory Warnings during Steering Allow for Faster Visual Target Recognition?

Introduction Aim: Auditory cues are often used to capture and direct attention , away from an ongoing task to a critical situation. In the context of driving, previous research have shown that looming sounds, which convey time-to-contact information in their rising intensity profiles, promote faster braking times to potential front collisions 1. The current experiment investigates the role of auditory warnings in facilitating the identification of visual objects in the visual periphery during steering. This approximates the use of auditory warnings for cueing possible candidates for side-collisions. We expected faster response times for visual targets cued by a looming sound compared to a constant sound. Electroencephalography (EEG) was recorded to determine whether faster response times were due to either earlier or stronger neural responses to the visual target. We hypothesize: 1) earlier event-related-potentials (ERPs) for cued compared to not-cued visualtargets, and 2) larger amplitudes for visual targets that were cued by looming versus constant sounds. Methods: While performing a primary steering task, participants (N=20) had to identify visual stimuli (i.e., Gabors patches ) presented in the periphery and to discriminate them for their tilt-orientation. In 50 of the trials, visual stimuli were preceded by a 400Hz sound with either constant or looming profiles. The looming sounds’ intensity profile increased exponentially over time, while the intensity profile for the constant sound did not change across the 500ms. Results: Our results show that participants responded faster to cued targets than to trials without a warning [t(19)=-9.054, p< .000]. The maximum peaks in the ERPs to visual targets were earlier for those that were cued with an auditory warning (black and red curve) compared to those without warning cues (blue curve). Next, looming warnings resulted in faster visual discrimination than constant sounds [F(1,19) = 6.934, p = .016]. The maximum peaks of ERPs to cued visual targets were larger for those that were cued by looming sounds compared to those that were cued by constant sounds. Looming sounds result in a larger ERP at 350ms than constant sounds. Discussion Conclusion: The maximal ERP peak that we report is likely to be the P3 component that is related to visual object recognition performance. This response occurs later in the absence of a warning cue. Looming auditory warnings might have induced a larger P3 component to visual targets, relative to constant warnings, by being more effective attentional cues. Our EEG data corresponds to behavioral benefit of looming auditory cues observed in faster reaction times. Interestingly, warning signals can prepare the brain to respond earlier to visual events, even with a predictability of only 50. These findings can directly be applied to the design of auditory warnings where fast but also accurate reaction times are preferable