Exp Brain Res (2012) 218:1–8 DOI 10.1007/s00221-011-2995-2 123 RESEARCH ARTICLE People post-stroke perceive movement Xuency in virtual reality Liesjet van Dokkum · Denis Mottet · Huei-Yune Bonnin-Koang · Julien Metrot · Agnès Roby-Brami · Isabelle Hauret · Isabelle LaVont Received: 28 October 2011 / Accepted: 28 December 2011 / Published online: 11 January 2012  Springer-Verlag 2012 Abstract We investigated the visual perception of bio- logical movement by people post-stroke, using minimal kinematic displays. A group of twenty patients and a group of twelve age-matched healthy controls were asked to judge movement Xuency. The movements to judge were either displayed as an end-point dot or as a stick-Wgure of the arm and trunk. It was found that the perception of movement Xuency was preserved post-stroke, however, with an increase in the variability of judgment. Moreover, the end- point dot representation ameliorated what was perceived and judged, presumably by directing attention to the impor- tant kinematic cues: smoothness and directness of the tra- jectory. We conclude that, despite perception of actions is inXuenced by the ability of the observer to execute the observed movement, hemiparesis has a mild eVect on the perception of biological movement. Yet, a valuable virtual learning environment for upper-limb rehabilitation should be implemented to provide the observer with neither too much, nor too little information to maximize learning. Keywords Stroke rehabilitation · Perception of action · Upper-extremity · Kinematic cues · Virtual reality Introduction Post-stroke rehabilitation is based on the general principles of motor learning (e.g. Shadmehr and Wise 2005) and relies on the basic idea that sensory-motor training can harness cortex plasticity (Hallett 2001). Addressing it as a learning problem does raise questions not only on practical motor training component, for example, training schedules, task selection, context dependency, and way of learning (Huang and Krakauer 2009), but as well on the perceptual side of motor learning and the role of observation. Observational motor learning is known to depend on (a) the information that is available and (b) its interpretation at time of action in the context of the ongoing skill, in such a way that the acquisition of skill will be diminished when there is too much or too little information provided (Guadagnoli and Lee 2004). Stated diVerently, too little information would impair action recognition and corrective processes, and too much information would blur the useful cues into the sur- rounding noise. An emerging rehabilitation technique proposed to enhance motor learning is virtual reality (Henderson et al. 2007). In this, avatars (a representation of the ‘player’ in the virtual world) were introduced to establish a link between the movement executed in the real world and the L. van Dokkum · D. Mottet (&) · J. Metrot · I. LaVont Movement to Health Laboratory (M2H), Montpellier-1 University, EuroMov, 700 Avenue du Pic Saint Loup, 34090 Montpellier, France e-mail: denis.mottet@univ-montp1.fr H.-Y. Bonnin-Koang Centre Medical Grau du Roi, University Hospital of Nîmes, Nîmes, France A. Roby-Brami Laboratoire de Neurophysique et Physiologie CNRS UMR 8119, Paris Descartes University, Paris, France I. Hauret Physical Medicine and Rehabilitation Unit, University Hospital of Clermont-Ferrand, Clermont-Ferrand, France I. LaVont Physical Medicine and Rehabilitation Unit, University Hospital of Montpellier, Montpellier, France