Using spatial vibrotactile cues to direct visual attention in driving scenes

Despite 2 centuries of research, the question of whether attending to a sensory modality speeds the perception of stimuli in that modality has yet to be resolved. The authors highlight weaknesses inherent in this previous research and report the results of 4 experiments in which a novel methodology was used to investigate the effects on temporal order judgments (TOJs) of attending to a particular sensory modality or spatial location. Participants were presented with pairs of visual and tactile stimuli from the left and/or right at varying stimulus onset asynchronies and were required to make unspeeded TOJs regarding which stimulus appeared first. The results provide the strongest evidence to date for the existence of multisensory prior entry and support previous claims for attentional biases toward the visual modality and toward the right side of space. These findings have important implications for studies in many areas of human and animal cognition.

Forty years ago a project to explore late brain plasticity was initiated that was to lead into a broad area of sensory substitution studies. The questions at that time were: Can a person who has never seen learn to see as an adult? Is the brain sufficiently plastic to develop an entirely new sensory system? The short answer to both questions is yes, first clearly demonstrated in 1969 ((Bach-y-Rita et al., 1969)). To reach that conclusion, it was first necessary to find a way to get visual information to the brain. That took many years and is still the most challenging aspect of the research and the development of practical sensory substitution and augmentation systems. The sensor array is not a problem: a TV camera for blind persons; an accelerometer for persons with vestibular loss; a microphone for deaf persons. These are common and fully developed devices. The problem is the brain-machine interface (BMI). In this short report, only two substitution systems are discussed, vision and vestibular substitution.

Tactile-visual links in spatial attention were examined by presenting spatially nonpredictive tactile cues to the left or right hand, shortly prior to visual targets in the left or right hemifield. To examine the spatial coordinates of any crossmodal links, different postures were examined. The hands were either uncrossed, or crossed so that the left hand lay in the right visual field and vice versa. Visual judgments were better on the side where the stimulated hand lay, though this effect was somewhat smaller with longer intervals between cue and target, and with crossed hands. Event-related brain potentials (ERPs) showed a similar pattern. Larger amplitude occipital N1 components were obtained for visual events on the same side as the preceding tactile cue, at ipsilateral electrode sites. Negativities in the Nd2 interval at midline and lateral central sites, and in the Nd1 interval at electrode Pz, were also enhanced for the cued side. As in the psychophysical results, ERP cueing effects during the crossed posture were determined by the side of space in which the stimulated hand lay, not by the anatomical side of the initial hemispheric projection for the tactile cue. These results demonstrate that crossmodal links in spatial attention can influence sensory brain responses as early as the N1, and that these links operate in a spatial frame-of-reference that can remap between the modalities across changes in posture.