Measures of multisensory integration based on dependent probability summation: from spike counts to reaction times

A single neuron is categorized as “multisensory” if there is a statistically significant difference between the response evoked by a cross-modal stimulus combination and that evoked by the most effective of its components individually. The most widely applied quantitative index expresses multisensory enhancement (or inhibition) as a proportion of the strongest unisensory response. However, it has no theoretical foundation in terms of the possible operations a neuron may perform in combining unisensory inputs to yield the multisensory response. In particular, being responsive to multiple sensory modalities does not guarantee that a neuron has actually engaged in integrating its multiple sensory inputs rather than simply responding to the most salient stimulus. Here, a new index is proposed based on a dependent probability summation mechanism. It compares the mean observed cross-modal response of a neuron with the largest cross-modal mean achievable by stochastically coupling its unisensory responses. Because this new measure is, in general, more restrictive than the traditional one, many neurons previously categorized as “multisensory” may possibly lose that property. The approach is illustrated by data from a single cat superior colliculus neuron. Computation of the new index is straightforward and it is as amenable to statistical testing as the traditional one. Moreover, it is completely analogous to a widely-used procedure in behavioral studies of multisensory integration (“race model inequality”). 1 Defining and measuring multisensory integration Single neurons in the deep layers of the mammalian superior colliculus (SC) have been shown to integrate afferent visual, auditory, and somatosensory cues and to generate efferent motor commands to structures innervating the musculature of, e.g., the eyes and hands [1, 2]. From a wealth of neurophysiological studies, some general “rules” for multisensory integration (MI) have emerged. Cross-modal cues (e.g., visual-auditory) that are in close spatial and temporal register enhance the response of multisensory neurons implementing the “spatial and temporal rules”, whereas those that are spatially or temporally disparate often elicit response depression or fail to be integrated [3](see also [4]). A third principle refers to the observation that proportionately greater effects of crossmodal cues are obtained when those individual cues are weakly effective. Thus, the magnitude of multisensory integration is inversely related to the efficacy of the stimuli being integrated ( “inverse effectiveness rule”) [5]. Moreover, those rules are closely mirrored by findings of behavioral studies measuring (saccadic or manual) reaction time [6]. Whereas empirical results are, at least in general, unequivocal regarding these rules, the issue of exactly how to measure “multisensory integration” School VI Medicine and Health Sciences, European Medical School Oldenburg-Groningen, and Cluster of Excellence, Hearing4all. Web: www.staff.uni-oldenburg.de/hans.colonius/ Center for Neurosensory Systems and Science, Oldenburg University