Alexander Gail and Florentin Wörgötter

We have shown that the posterior parietal cortex of primates encodes the motor-goals of planned reach movements in a context-specific manner [Gail et al. 2006, Brozovic et al. 2007, Gail et al. 2009, Westendorff et al. 2010]. Our results suggest that PPC represents explicit reach plans based on abstract visuomotor associations. Control signals from such highly adaptive brain areas denote a major advantage for learning to control neuroprosthetic devices [Gail 2007, Musallam et al. 2004]. Aim of this project is to investigate the neuronal mechanism underlying the adaptive, implicit learning of visuomotor transformations and its applicability to robotic control.

In psychophysical tests with humans we will develop a novel adaptation paradigm. While subjects conduct goal-directed reaches, we will induce systematic reach errors by perturbing the sensory input. The visuomotor system adapts in such situations to compensate the error.

Further, we will conduct multi-channel microelectrode recordings in trained monkeys during visuomotor adaptation. These adaptation experiments will allow us to analyze visuomotor learning on the neuronal level, and help to identify effects of implicit motor-goal learning. We will test, in cooperation with project D1, if neural computations in the parietal cortex could contribute to predictions of the sensory consequences of a planned movement. Understanding such mechanisms will help to improve corresponding learning algorithms in robot control.