Abstract
Multisensory integration is the perceptual enhancement deriving from the integration of stimuli from different sensory channels, due to a neural coactivation mechanism. Multisensory integration has been suggested to be different from unisensory integration ( i.e. the statistical facilitation in a behavioural response when two stimuli of the same sensory modality are presented) and to be subserved by the activity of Superior Colliculus (SC). In order to verify these hypotheses, in the present study, a group of patients with subcortical lesions at the SC and a control group of healthy subjects were tested in a speeded detection task (Experiment 1) and a localization task (Experiment 2). Subjects were presented with modality-specific stimuli (visual or auditory), cross-modal stimulus pairs (audio-visual) and within-modal stimulus pairs (visual-visual). In Experiment 1, control subjects showed a multisensory enhancement effect with a violation of the race model inequality only for audio-visual stimuli, whereas visual-visual stimuli induced a statistical facilitation effect. On the other hand, SC patients did not show any significant violation of the race inequality, demonstrating only a statistical facilitation effect both for audio-visual and visual-visual stimuli. In Experiment 2, control subjects exhibited significantly enhanced localization accuracy for audio-visual stimuli, while in SC patients no differences were found in the localization performances. Overall these results suggest that multisensory and unisensory integration are two distinct phenomena and that the former, due to a neural coactivation mechanism, requires SC activity to occur, whereas the latter is due to a statistical facilitation and is independent from SC activity.
