Benjamin A. Rowland

In this viewpoint article, Florencia Iacaruso, Benjamin A. Rowland and Barry E. Stein examine the principles and open questions surrounding single-neuron dynamics in multisensory integration. Drawing on decades of work, they emphasize that individual neurons, particularly in structures such as the superior colliculus, exhibit rapid, nonlinear interactions when combining inputs from different sensory modalities. These dynamics are shaped by experience, circuit architecture, and the precise timing and balance of converging inputs, rather than reflecting fixed or purely feedforward processes. The authors highlight that single-neuron responses provide a critical bridge between cellular mechanisms and behaviour, revealing how multisensory signals are integrated in real time to support detection, localization, and decision-making. Recent methodological advances, including high-density electrophysiology, calcium imaging, and computational modelling, are enabling more detailed characterization of these dynamics within distributed and context-dependent networks. Key open questions focus on the origin of multisensory effects observed at the single-neuron level, including whether they arise from local circuit computations or reflect inputs inherited from broader network interactions. The variability of neuronal responses across contexts, such as changes in attention, behavioural state, and task demands, further complicates the identification of general principles. Looking forward, the authors argue that integrating large-scale functional data with anatomical connectivity and artificial intelligence approaches will be essential for linking single-neuron dynamics to system-level function. They conclude that understanding multisensory processing at the level of individual neurons remains central to explaining how the brain constructs coherent perception from diverse sensory inputs.

The 1993 book The Merging of the Senses has proven to be a profoundly impactful text that has shaped research programs studying the interaction between the senses for the last three decades. The book combines skillful and approachable narration with engaging illustrations and was received with rave reviews on publication as one of the first comprehensive approaches to the subject. It captures the impressive breadth of domains in which multisensory integration impacts the daily life of all animals and promotes a systematic approach to understanding its underlying operation by interrogating the nervous system at multiple levels, from the peripheral organ, through convergence, integration, and decision-making, to effected behavior. Thirty years later, the multiple generations of scientists that have been inspired by the text have built an amazing structure on this foundation, through advancing refinements in theory and experimental technique, investigation of new domains and species, an understanding of the origins, maturation, and plasticity of the process, the translation of biological principles to artificial systems, and discovering new applications of multisensory research in clinical and rehabilitative domains.

Midbrain multisensory neurons undergo a significant postnatal transition in how they process cross-modal (e.g. visual–auditory) signals. In early stages, signals derived from common events are processed competitively; however, at later stages they are processed cooperatively such that their salience is enhanced. This transition reflects adaptation to cross-modal configurations that are consistently experienced and become informative about which correspond to common events. Tested here was the assumption that overt behaviors follow a similar maturation. Cats were reared in omnidirectional sound thereby compromising the experience needed for this developmental process. Animals were then repeatedly exposed to different configurations of visual and auditory stimuli (e.g. spatiotemporally congruent or spatially disparate) that varied on each side of space and their behavior was assessed using a detection/localization task. Animals showed enhanced performance to stimuli consistent with the experience provided: congruent stimuli elicited enhanced behaviors where spatially congruent cross-modal experience was provided, and spatially disparate stimuli elicited enhanced behaviors where spatially disparate cross-modal experience was provided. Cross-modal configurations not consistent with experience did not enhance responses. The presumptive benefit of such flexibility in the multisensory developmental process is to sensitize neural circuits (and the behaviors they control) to the features of the environment in which they will function. These experiments reveal that these processes have a high degree of flexibility, such that two (conflicting) multisensory principles can be implemented by cross-modal experience on opposite sides of space even within the same animal.

Hemianopia is a common consequence of unilateral damage to visual cortex that manifests as a profound blindness in contralesional space. A noninvasive cross-modal (visual–auditory) exposure paradigm has been developed in an animal model to ameliorate this disorder. Repeated stimulation of a visual–auditory stimulus restores overt responses to visual stimuli in the blinded hemifield. It is believed to accomplish this by enhancing the visual sensitivity of circuits remaining after a lesion of visual cortex; in particular, circuits involving the multisensory neurons of the superior colliculus. Neurons in this midbrain structure are known to integrate spatiotemporally congruent visual and auditory signals to amplify their responses, which, in turn, enhances behavioral performance. Here we evaluated the relationship between the rehabilitation of hemianopia and this process of multisensory integration. Induction of hemianopia also eliminated multisensory enhancement in the blinded hemifield. Both vision and multisensory enhancement rapidly recovered with the rehabilitative cross-modal exposures. However, although both reached pre-lesion levels at similar rates, they did so with different spatial patterns. The results suggest that the capability for multisensory integration and enhancement is not a pre-requisite for visual recovery in hemianopia, and that the underlying mechanisms for recovery may be more complex than currently appreciated.

Hemianopia (unilateral blindness), a common consequence of stroke and trauma to visual cortex, is a debilitating disorder for which there are few treatments. Research in an animal model has suggested that visual-auditory stimulation therapy, which exploits the multisensory architecture of the brain, may be effective in restoring visual sensitivity in hemianopia. It was tested in two male human patients who were hemianopic for at least 8 months following a stroke. The patients were repeatedly exposed to congruent visual-auditory stimuli within their blinded hemifield during 2 h sessions over several weeks. The results were dramatic. Both recovered the ability to detect and describe visual stimuli throughout their formerly blind field within a few weeks. They could also localize these stimuli, identify some of their features, and perceive multiple visual stimuli simultaneously in both fields. These results indicate that the multisensory therapy is a rapid and effective method for restoring visual function in hemianopia.SIGNIFICANCE STATEMENTHemianopia (blindness on one side of space) is widely considered to be a permanent disorder. Here, we show that a simple multisensory training paradigm can ameliorate this disorder in human patients.

Concordant visual–auditory stimuli enhance the responses of individual superior colliculus (SC) neurons. This neuronal capacity for “multisensory integration” is not innate: it is acquired only after substantial cross-modal (e.g. auditory–visual) experience. Masking transient auditory cues by raising animals in omnidirectional sound (“noise-rearing”) precludes their ability to obtain this experience and the ability of the SC to construct a normal multisensory (auditory–visual) transform. SC responses to combinations of concordant visual–auditory stimuli are depressed, rather than enhanced. The present experiments examined the behavioral consequence of this rearing condition in a simple detection/localization task. In the first experiment, the auditory component of the concordant cross-modal pair was novel, and only the visual stimulus was a target. In the second experiment, both component stimuli were targets. Noise-reared animals failed to show multisensory performance benefits in either experiment. These results reveal a close parallel between behavior and single neuron physiology in the multisensory deficits that are induced when noise disrupts early visual–auditory experience.

The superior colliculus (SC) is richly endowed with neurons that integrate cues from different senses to enhance their physiological responses and the overt behaviors they mediate. However, in the absence of experience with cross‐modal combinations (e.g., visual–auditory), they fail to develop this characteristic multisensory capability: Their multisensory responses are no greater than their most effective unisensory responses. Presumably, this impairment in neural development would be reflected as corresponding impairments in SC‐mediated behavioral capabilities such as detection and localization performance. Here, we tested that assumption directly in cats raised to adulthood in darkness. They, along with a normally reared cohort, were trained to approach brief visual or auditory stimuli. The animals were then tested with these stimuli individually and in combination under ambient light conditions consistent with their rearing conditions and home environment as well as under the opposite lighting condition. As expected, normally reared animals detected and localized the cross‐modal combinations significantly better than their individual component stimuli. However, dark‐reared animals showed significant defects in multisensory detection and localization performance. The results indicate that a physiological impairment in single multisensory SC neurons is predictive of an impairment in overt multisensory behaviors.

Although the ability to integrate information across the senses is compromised in some individuals for unknown reasons, similar defects have been observed when animals are reared without multisensory experience. The experience-dependent development of multisensory integration has been studied most extensively using the visual-auditory neuron of the cat superior colliculus (SC) as a neural model. In the normally-developed adult, SC neurons react to concordant visual-auditory stimuli by integrating their inputs in real-time to produce non-linearly amplified multisensory responses. However, when prevented from gathering visual-auditory experience, their multisensory responses are no more robust than their responses to the individual component stimuli. The mechanisms operating in this defective state are poorly understood. Here we examined the responses of SC neurons in “naïve” (i.e., dark-reared) and “neurotypic” (i.e., normally-reared) animals on a millisecond-by-millisecond basis to determine whether multisensory experience changes the operation by which unisensory signals are converted into multisensory outputs (the “multisensory transform”), or whether it changes the dynamics of the unisensory inputs to that transform (e.g., their synchronization and/or alignment). The results reveal that the major impact of experience was on the multisensory transform itself. Whereas neurotypic multisensory responses exhibited non-linear amplification near their onset followed by linear amplification thereafter, the naive responses showed no integration in the initial phase of the response and a computation consistent with competition in its later phases. The results suggest that multisensory experience creates an entirely new computation by which convergent unisensory inputs are used cooperatively to enhance the physiological salience of cross-modal events and thereby facilitate normal perception and behavior.

Hemianopia can be rehabilitated by an auditory-visual “training” procedure, which restores visual responsiveness in midbrain neurons indirectly compromised by the cortical lesion and reinstates vision in contralesional space. Presumably, these rehabilitative changes are induced via mechanisms of multisensory integration/plasticity. If so, the paradigm should fail if the stimulus configurations violate the spatiotemporal principles that govern these midbrain processes. To test this possibility, hemianopic cats were provided spatially or temporally noncongruent auditory-visual training. Rehabilitation failed in all cases even after approximately twice the number of training trials normally required for recovery, and even after animals learned to approach the location of the undetected visual stimulus. When training was repeated with these stimuli in spatiotemporal concordance, hemianopia was resolved. The results identify the conditions needed to engage changes in remaining neural circuits required to support vision in the absence of visual cortex, and have implications for rehabilitative strategies in human patients.