Is the primary visual cortex multisensory?: Comment on “Crossmodal influences on visual perception” by Prof. Ladan Shams

Abstract

In our everyday life we are confronted with a constant flow of information from several modalities (visual, auditory, proprioceptive, etc.) and multisensory integration (MSI) allows us to link the unimodal sensory analysis with the polysensory context in which they appear (crossmodal fusion). As extensively described by Shams and Kim in this issue [1], MSI results in perceptual improvements in various tasks, from simple detections to complex discriminations, memory or learning tasks. The fact that such large fields are involved in multisensory interactions implies the participation of an equally large variety of cortical networks. During the last decade, we have observed a reappraisal of the neuronal networks supporting MSI, emphasizing the original theoretical concept of a broad allocation of the cortical regions susceptible to host multisensory interactions [2]. In their review, Shams and Kim bring into focus the intriguing observation that multisensory interactions can occur at very short delays in regions historically defined as unimodal, so excluding the origin of MSI in a feedback influence from high order associative areas. Non-human primates constitute a good model to study the neuronal mechanisms of MSI as they show the same perceptual multisensory benefits as reported in humans [3]. Thus to corroborate human brain imaging studies, anatomical and electrophysiological approaches have been developed to assess if crossmodal influences can occur at early stages of sensory processing such as the primary visual cortex (V1). A recent set of anatomical studies revealed a network of heteromodal connections that link directly cortical regions processing different modalities [4]. A common feature of these heteromodal connections is a coupling between specific sensory representations across modalities (visual field or body part representations) suggesting a functional role restricted to specialized behaviors. Among these heteromodal connections, the presence of direct anatomical links between the primary visual and auditory areas suggests that this pathway can support the visuo-auditory interactions described in human and reviewed by Shams and Kim. I would like to stress that recording of single unit activity in the primary visual cortex does not reveal pure auditory responses. Instead, neurons recorded in V1 showed a modulation of their visual response as expressed as a significant reduction in visual response latencies specifically in suboptimal conditions [5] when approaching perceptual threshold. It should be noted that, in the primary auditory cortex of the monkey, other authors have reported the absence of non-specific