Shadmehr Laboratory
Laboratory for Computational Motor Control · Johns Hopkins University
Jun Izawa

Jun Izawa

Postdoctoral Fellow (2005-2008)

Research

Jun joined the lab as a postdoc in 2005 after completing his PhD from University of Tokyo. Jun played a leading role in a number of new directions for the lab. He demonstrated that adaptation was not merely a process of canceling a perturbation, but rather re-optimization, i.e., finding a new way to maximize probability of acquiring reward. This implied that there could not be a ‘desired trajectory’ in the brain, but rather that all learning was ultimately a search for reward. He discovered that in cerebellar disease, there was intact learning from reward prediction errors, but a specific impairment in learning from sensory prediction errors. In this way, it was possible to improve performance of cerebellar patients, something that had been very difficult to do before Jun’s work. He discovered that in autism, children exhibited an abnormally strong sensitivity proprioception during learning. Perhaps the most important of his many contributions is the idea that the brain maintains multiple learning systems, and that during motor learning, two systems cooperate: a reward based learning system, and a sensory prediction based learning system. The former is spared in people with cerebellar damage. He discovered a way to behaviorally examine the consequence of learning with each system. He is currently Professor in the Faculty of Engineering, Information, and Systems, University of Tsukuba, Japan.

Ideas developed by Jun

  • Distinguishing learning from sensory prediction errors and reward prediction errors. Jun developed a framework in which motor learning can arise from different error signals. Sensory prediction errors drive adaptation of an internal model, whereas reward prediction errors can alter actions through a distinct learning process. This work helped connect theories of motor adaptation with reinforcement learning.
  • Showing that motor adaptation involves both learning an internal model and reoptimizing behavior. Jun's work demonstrated that adaptation is not simply the correction of movement errors. As the nervous system learns new dynamics, it can also change the movement that it chooses to produce, reoptimizing behavior in response to the newly learned internal model.
  • Clarifying the cerebellum's role in learning the sensory consequences of action. Jun helped show that cerebellar damage impairs the ability to learn the sensory consequences of motor commands, supporting the idea that the cerebellum contributes to constructing predictive internal models that estimate what an action should feel like.
  • Revealing how uncertainty, sensory information, and reward shape movement. Jun studied how the motor system integrates uncertain sensory information when controlling movements and how this process differs across neurological conditions. His work also helped show that reward changes movement variability, linking the value of an outcome to exploration during motor learning.
  • Selected publications

    Pekny SE, Izawa J, Shadmehr R. Reward dependent modulation of movement variability. Journal of Neuroscience. 2015;35:4015–4024.

    Izawa J, Criscimagna-Hemminger SE, Shadmehr R. Cerebellar contributions to reach adaptation and learning sensory consequences of action. Journal of Neuroscience. 2012;32:4230–4239.

    Izawa J, Pekny SE, Marko MK, Haswell C, Shadmehr R, Mostofsky SH. Motor learning relies on integrated sensory inputs in ADHD, but over-selectively on proprioception in Autism spectrum conditions. Autism Research. 2012;5:124–136.

    Orban de Xivry JJ, Marko MK, Pekny SE, Pastor D, Izawa J, Celnik P, Shadmehr R. Stimulation of the human motor cortex alters generalization patterns of motor learning. Journal of Neuroscience. 2011;31:7102–7110.

    Izawa J, Shadmehr R. Learning from sensory and reward prediction errors during motor adaptation. PLoS Computational Biology. 2011;7:e1002012.

    Haswell C, Izawa J, Dowell L, Mostofsky SH, Shadmehr R. Representation of internal models of action in the autistic brain. Nature Neuroscience. 2009;12:970–972.

    Izawa J, Shadmehr R. Online processing of uncertain information in visuomotor control. Journal of Neuroscience. 2008;28:11360–11368.

    Izawa J, Rane T, Donchin O, Shadmehr R. Motor adaptation as a process of reoptimization. Journal of Neuroscience. 2008;28:2883–2891.

    Gallery

    ← Return to lab members