Reinforcement Learning  Learning Through SuccessBased Feedback
We can acquire knowledge of new motor behavior through different forms of learning. One of the most common forms is known as reinforcement learning, which gives success-based feedback. In this study, Dr. Celnik, the Director of the Department of Physical Medicine and Rehabilitation, and his lab evaluated physiological markers from the cerebellum and primary motor cortex using noninvasive brain stimulations while healthy participants trained with finger-reaching tasks. They had participants perform a task of trying to reach the “target” position from the “home” position on a computer screen while not being able to see their hands. The only feedback they received was the visual feedback on the screen—the screen would go green if the target was hit and red if the target was missed. Since the participant cannot physically see his hand move, he needs to explore the movements in order to hit the target. This leads to reinforcement learning—if he is correct he will repeat that action and if incorrect he will avoid that same movement in the future. Although they still need to be tested, the findings of this research show that, instead of trying to teach someone to learn with a partially damaged area of the brain, there’s a possibility of having them relearn motor behaviors by using areas of the brain that have not been damaged. If true, this can lead to the development of new rehabilitation interventions to help patients with neurological conditions.