Virtual environments to assess perceptuomotor factors that influence obstacle circumvention in the post-stroke population

Locomotor adaptations to circumvent moving obstacles are initiated with respect to obstacle and self-motion perception, in order to predict and avoid an impending collision. We have developed a virtual reality (VR) based obstacle circumvention task to assess locomotor adaptation (experiment [Exp] 1) and the ability to predict an impending collision (Exp 2) as well as detection of the onset of obstacle motion (Exp 3). The locomotor (Exp 1) and seated perceptuomotor [Exp 2-3] behaviors were contrasted between post-stroke (n=12) and healthy participants (n=12). Three red cylindrical obstacles approaching from head-on (0°) and diagonally (30° left/right) were seen in a virtual room along with a blue central target inside a helmet-mounted display. In Exp 1, participants walked towards the target while avoiding collision with the approaching obstacles. The delay in initiation of an avoidance strategy after the initiation of obstacle motion was assessed. In Exp 2, obstacles moved diagonally (left/right) at a fixed speed of 0.75 m/s while the virtual scene moved at 0.5, 0.75 and 1 m/s, thus creating `collision' and `no collision' conditions. Participants were instructed to indicate their prediction of such using the joystick. In Exp 3, participants were instructed to indicate as soon as they perceived motion of obstacle in the virtual scene. The delay in pressing the joystick button after initiation of obstacle motion was assessed. Individuals with stroke initiated locomotor avoidance strategies significantly later than healthy individuals for all obstacle approaches. This delay, however, was not associated with the delay in obstacle detection. No significant difference was found in either the accuracy of collision prediction or the delay in detection of obstacle motion. However, two post-stroke individuals who collided frequently with obstacles demonstrated low accuracy in predicting an impending collision. One of them also showed significant delays in detecting initiation of obstacle motion when compared with others. Thus, the seated perceptuomotor tasks may reveal underlying sensory-perceptual and cognitive deficits post-stroke. VR based locomotor obstacle circumvention along with seated perceptuomotor assessments provide a comprehensive understanding of obstacle circumvention behavior and the underlying perceptual and cognitive factors shaping them in post-stroke and healthy individuals.