Rapid adaptation of the vestibulo-ocular reflex and induced self-motion perception

Simultaneous adaptation of retinal and extra-retinal motion signals

A number of models of motion perception include estimates of eye velocity to help compensate for the incidental retinal motion produced by smooth pursuit. The 'classical' model uses extra-retinal motor command signals to obtain the estimate. More recent 'reference-signal' models use retinal motion information to enhance the extra-retinal signal. The consequence of simultaneously adapting to pursuit and retinal motion is thought to favour the reference-signal model, largely because the perception of motion during pursuit ('perceived stability') changes despite the absence of a standard motion aftereffect. The current experiments investigated whether the classical model could also account for these findings. Experiment 1 replicated the changes to perceived stability and then showed how simultaneous motion adaptation changes perceived retinal speed (a velocity aftereffect). Contrary to claims made by proponents of the reference-signal model, adapting simultaneously to pursuit and retinal motion therefore alters the retinal motion inputs to the stability computation. Experiment 2 tested the idea that simultaneous motion adaptation sets up a competitive interaction between two types of velocity aftereffect, one retinal and one extra-retinal. The results showed that pursuit adaptation by itself drove perceived stability in one direction and that adding adapting retinal motion drove perceived stability in the other. Moreover, perceived stability changed in conditions that contained no mismatch between adapting pursuit and adapting retinal motion, contrary to the reference-signal account. Experiment 3 investigated whether the effects of simultaneous motion adaptation were directionally tuned. Surprisingly no tuning was found, but this was true for both perceived stability and retinal velocity aftereffect. The three experiments suggest that simultaneous motion adaptation alters perceived stability based on separable changes to retinal and extra-retinal inputs. Possible mechanisms underlying the extra-retinal velocity aftereffect are discussed.

Optic flow dominates visual scene polarity in causing adaptive modification of locomotor trajectory

Locomotion and posture are influenced and controlled by vestibular, visual and somatosensory information. Optic flow and scene polarity are two characteristics of a visual scene that have been identified as being critical in how they affect perceived body orientation and self motion. The goal of this study was to determine the role of optic flow and visual scene polarity on adaptive modification in locomotor trajectory. An object is said to have visual polarity, or to be "visually polarized", when it contains an identifiable principal axis with one end distinct from the other. Two computer-generated virtual reality scenes were shown to subjects during 20 min of treadmill walking. One scene was a highly polarized scene, while the other was composed of objects displayed in a non-polarized fashion. Both virtual scenes depicted constant rate self motion equivalent to walking counterclockwise around the perimeter of a room. Subjects performed Stepping Tests blindfolded before and after scene exposure to assess adaptive changes in locomotor trajectory. Subjects showed a significant difference in heading direction, between pre- and post-adaptation Stepping Tests, when exposed to either scene during treadmill walking. However, there was no significant difference in the subjects' heading direction between the two visual scene polarity conditions. Therefore, it was inferred from these data that optic flow has a greater role than visual polarity in influencing adaptive locomotor function.

Exposure to a rotating virtual environment during treadmill locomotion causes adaptation in heading direction

The objective of this study was to investigate the adaptive effects of variation in the direction of optic flow, experienced during linear treadmill walking, on modifying locomotor trajectory. Subjects (n=30) walked on a motorized linear treadmill at 4.0 km h(-1) for 24 min while viewing the interior of a 3D virtual scene projected on to a screen 1.5 m in front of them. The virtual scene depicted constant self-motion equivalent to either (1) walking around the perimeter of a room to one's left (Rotating Room group) or (2) walking down the center of a hallway (Infinite Corridor group). The scene was static for the first 4 min and then constant rate self-motion was simulated for the remaining 20 min. Before and after the treadmill locomotion adaptation period subjects performed five stepping trials. In each trial they marched in place to the beat of a metronome at 90 steps min(-1) for a total of 100 steps while blindfolded in a quiet room. The subject's final heading direction (deg) and final X (fore-aft, cm) and final Y (medio-lateral, cm) positions were measured for each trial. During the treadmill locomotion adaptation period subjects' 3D torso position was measured. We found that subjects in the Rotating Room group, as compared with the Infinite Hallway group: (1) showed significantly greater deviation during post-exposure testing in the heading direction and Y position opposite to the direction of optic flow experienced during treadmill walking; and (2) showed a significant monotonically increasing torso yaw angular rotation bias in the direction of optic flow during the treadmill adaptation exposure period. Subjects in both groups showed greater forward translation (in the +X direction) during the post-treadmill stepping task that differed significantly from their pre-exposure performance. Subjects in both groups reported no perceptual deviation in position during the stepping tasks. We infer that viewing simulated rotary self-motion during treadmill locomotion causes adaptive modification of sensorimotor integration in the control of position and trajectory during locomotion, which functionally reflects adaptive changes in the integration of visual, vestibular, and proprioceptive cues. Such an adaptation in the control of position and heading direction during locomotion, because of the congruence of sensory information, demonstrates the potential for adaptive transfer between sensorimotor systems and suggests a common neural site for processing and self-motion perception and concurrent adaptation in motor output.