Multisensory integration and sensory reweighting


1  Accurate and efficient sensorimotor behavior depends on precise localization of the body in space, which may be estimated using multiple sensory modalities (i.e., vision and proprioception)

2  The most direct sensory information concerning the body arises from somatosensory (cutaneous touch and proprioception)  along with signals from the vestibular system which provide direct information about the body's positioning with respect to the environment.

The somatosensory and vestibular signals are combined, and integrated with inputs from with visual and auditory sesnors providing information about the relationship of our bodies with the external environment. 

3  Sensory information is always multimodal, some information being useful and other redundant with respect to task performance. The movement brain filters and weights information in order create the best representation of the present situation. 

Reweighting depends on unimodal acuity

Static visual acuity has been shown to develop  in young children whereas proprioceptive acuity continues to improve throughout childhood and adolescence. 

Reasearch by King et al suggests that multisensory-motor integration in children is a flexible process influenced by the functioning of unimodal inputs. In the context of their study, improvements in proprioceptive localization in 7-13 year-old children resulted in an increased contribution of proprioception to the multisensory estimate; and, critically, this finding was independent of age. This result extends the findings of previous research examining multisensory-motor integration in postural control tasks. In children younger than three years, vision has been considered the dominant modality and the contribution of other sensory inputs (i.e., proprioception) increased with age until approximately 7-10 years. Older children exhibited adult-like multisensory integration, whereas the young children (i.e., 4- to 6-years) were thought to be in a transition period for multisensory integration.

"Evidence from a variety of tasks suggests that the adult CNS utilizes probabilistic mechanisms to reduce the uncertainty inherent in multisensory-motor processing . In a probabilistic framework, available sensory information is differentially utilized, or flexibly ‘re-weighted’, to reduce the uncertainty associated with a multisensory estimate. (Wolpert 2007).  One such example includes the weighting of sensory inputs based on the inverse of the variability of the unimodal estimates. Results from the current study suggest that multisensory-motor re-weighting is dependent on unimodal sensory functioning, a finding that is consistent with a probabilistic framework. (King 2010).


Stretch reflex

Safavynia SA, Ting LH. Long-latency muscle activity reflects continuous, delayed sensorimotor feedback of task-level and not joint-level error. J Neurophysiol. 2013 Sep;110(6):1278-90. doi: 10.1152/jn.00609.2012. Epub 2013 Jun 26. PubMed PMID: 23803325; PubMed Central PMCID: PMC3763153.

Turpin NA, Levin MF, Feldman AG. Implicit learning and generalization of stretch response modulation in humans. J Neurophysiol. 2016 Jun 1;115(6):3186-94. doi: 10.1152/jn.01143.2015. Epub 2016 Apr 6. PubMed PMID: 27052586; PubMed Central PMCID: PMC4946602.

Dimitriou M, Franklin DW, Wolpert DM. Task-dependent coordination of rapid bimanual motor responses. J Neurophysiol. 2012 Feb;107(3):890-901. doi: 10.1152/jn.00787.2011. Epub 2011 Nov 9. PubMed PMID: 22072514; PubMed Central PMCID: PMC3289469.

Kurtzer, I. L. (2014). Long-latency reflexes account for limb biomechanics through several supraspinal pathways. Frontiers in Integrative Neuroscience, 8, 99.


Bair, W.-N., Kiemel, T., Jeka, J. J., & Clark, J. E. (2007). DEVELOPMENT OF MULTISENSORY REWEIGHTING FOR POSTURE CONTROL IN CHILDREN. Experimental Brain Research. Experimentelle Hirnforschung. Experimentation Cerebrale, 183(4), 435–446.

Multisensory processes are fundamental in scaffolding perception, cognition, learning and behaviour. How and when stimuli from different sensory modalities are integrated rather than treated as separate entities is poorly understood. We review how the relative reliance on stimulus characteristics versus learned associations dynamically shapes multisensory processes. We illustrate the dynamism in multisensory function across two timescales: one long-term that operates across the lifespan and one short-term that operates during the learning of new multisensory relations. In addition, we highlight the importance of task contingencies. We conclude that these highly dynamic multisensory processes, based on the relative weighting of stimulus characteristics and learned associations, provide both stability and flexibility to brain functions over a wide range of temporal scales.

Bair, W.-N., Kiemel, T., Jeka, J. J., & Clark, J. E. (2007). DEVELOPMENT OF MULTISENSORY REWEIGHTING FOR POSTURE CONTROL IN CHILDREN. Experimental Brain Research. Experimentelle Hirnforschung. Experimentation Cerebrale, 183(4), 435–446.

Reweighting to multisensory inputs adaptively contributes to stable and flexible upright stance control. However, few studies have examined how early a child develops multisensory reweighting ability, or how this ability develops through childhood. The purpose of the study was to characterize a developmental landscape of multisensory reweighting for upright postural control in children 4 to 10 years of age. Children were presented with simultaneous small-amplitude somatosensory and visual environmental movement at 0.28 and 0.2 Hz, respectively, within five conditions that independently varied the amplitude of the stimuli. The primary measure was body sway amplitude relative to each stimulus: touch gain and vision gain. We found that children can reweight to multisensory inputs from 4 years on. Specifically, intra-modal reweighting was exhibited by children as young as 4 years of age; however, inter-modal reweighting was only observed in the older children. The amount of reweighting increased with age indicating development of a better adaptive ability. Our results rigorously demonstrate the development of simultaneous reweighting to two sensory inputs for postural control in children. The present results provide further evidence that the development of multisensory reweighting contributes to more stable and flexible control of upright stance which ultimately serves as the foundation for functional behaviors such as locomotion and reaching.

Dusing SC. Postural variability and sensorimotor development in infancy. Dev Med Child Neurol. 2016 Mar;58 Suppl 4:17-21. doi: 10.1111/dmcn.13045. Review. PubMed PMID: 27027603

Infants develop skills through a coupling between their sensory and motor systems. Newborn infants must interpret sensory information and use it to modify movements and organize the postural control system based on the task demands. This paper starts with a brief review of evidence on the use of sensory information in the first months of life, and describes the importance of movement variability and postural control in infancy. This introduction is followed by a review of the evidence for the interactions between the sensory, motor, and postural control systems in typically development infants. The paper highlights the ability of young infants to use sensory information to modify motor behaviors and learn from their experiences. Last, the paper highlights evidence of atypical use of sensory, motor, and postural control in the first months of life in infants who were born preterm, with neonatal brain injury or later diagnosed with cerebral palsy (CP).

King, B. R., Pangelinan, M. M., Kagerer, F. A., & Clark, J. E. (2010). Improvements in proprioceptive functioning influence multisensory-motor integration in 7- to 13-year-old children. Neuroscience Letters, 483(1), 36–40.

Accurate and efficient sensorimotor behavior depends on precise localization of the body in space, which may be estimated using multiple sensory modalities (i.e., vision and proprioception). Although age-related differences in multisensory-motor integration across childhood have been previously reported, the extent to which age-related changes in unimodal functioning affect multisensory-motor integration is unclear. The purpose of the current study was to address this knowledge gap. Thirty-seven 7- to 13-year-old children moved their dominant hand in a target localization task to visual, proprioceptive, and concurrent visual and proprioceptive stimuli. During a subsequent experimental phase, we introduced a perturbation that placed the concurrent visual and proprioceptive stimuli in conflicting locations (incongruent condition) to determine the relative contributions of vision and proprioception to the multisensory estimate of target position. Results revealed age-related differences in the localization of incongruent stimuli in which the visual estimate of target position contributed more to the multisensory estimate in the younger children whereas the proprioceptive estimate was up-weighted in the older children. Moreover, above and beyond the effects of age, differences in proprioceptive functioning systematically influenced the relative contributions of vision and proprioception to the multisensory estimate during the incongruent trials. Specifically, improvements in proprioceptive functioning resulted in an up-weighting of proprioception, suggesting that the central nervous system of school-aged children utilizes information about unimodal functioning to integrate redundant sensorimotor inputs.

​Wolpert DM. Probabilistic models in human sensorimotor control. Hum Mov Sci. 2007;26:511–524.

Sensory and motor uncertainty form a fundamental constraint on human sensorimotor control. Bayesian decision theory (BDT) has emerged as a unifying framework to understand how the central nervous system performs optimal estimation and control in the face of such uncertainty. BDT has two components: Bayesian statistics and decision theory. Here we review Bayesian statistics and show how it applies to estimating the state of the world and our own body. Recent results suggest that when learning novel tasks we are able to learn the statistical properties of both the world and our own sensory apparatus so as to perform estimation using Bayesian statistics. We review studies which suggest that humans can combine multiple sources of information to form maximum likelihood estimates, can incorporate prior beliefs about possible states of the world so as to generate maximum a posteriori estimates and can use Kalman filter-based processes to estimate time-varying states. Finally, we review Bayesian decision theory in motor control and how the central nervous system processes errors to determine loss functions and optimal actions. We review results that suggest we plan movements based on statistics of our actions that result from signal-dependent noise on our motor outputs. Taken together these studies provide a statistical framework for how the motor system performs in the presence of uncertainty.