Characterizing Sensory Re-weighting for Human Postural Control

Abstract

In order to survive in the wide range of sensory contexts that comprise our physical world, the nervous system employs adaptive mechanisms that optimize functional behaviors within a given sensory environment. Human bipedal stance control requires that the nervous system obtain relevant information about the environment and the body's relationship with it from multiple sensory systems. How does the nervous system accomplish this when the sensory environment compromises the information available from a given sensory system? In previous theoretical and empirical work, we have provided evidence of nonlinearities that are consistent with an hypothesis of sensory re-weighting: The nervous system adapts to changing sensory contexts by decreasing its dependence, or weighting, on the compromised system and increases its weighting of other inputs. This thesis presents empirical findings that further support the sensory re-weighting hypothesis and further efforts towards characterizing sensory re-weighting by providing empirical results that provide important constraints on any proposed sensory re-weighting scheme. First, postural responses to complex visual motion consisting of the sum of 10 different sinusoidal components, were measured at two different amplitudes. Changes in the gain of body sway to visual motion were consistent with the nonlinearities previously interpreted as evidence for sensory re-weighting. Further, the observed changes in gain did not vary significantly as a function of stimulus frequency. Second, we found evidence indicating a temporal asymmetry in the sensory re-weighting process dependent upon the direction of the change in stimulus motion amplitude: the change in postural response is faster to a rapid increase versus decrease in stimulus amplitude. This temporal asymmetry was interpreted functionally: an increase in visual environmental motion may threaten balance, requiring a rapid down-weighting of vision if a strong dependence upon visual information would increase postural response beyond the stability boundaries of stance. Conversely, if stance is already stable in the face of large visual motion amplitude, a decrease in motion amplitude does not threaten balance and adapting rapidly to the new sensory conditions is not critical to avoid falling.