35th CiNet Monthly Seminar: Tomohiko Takei “Neural mechanisms for sophisticated feedback motor control”


CiNet Monthly Seminar

June 5, 2019
15:00 ~ 16:00
CiNet 1F Conference Room

“Neural mechanisms for sophisticated feedback motor control”

Tomohiko Takei
Hakubi Centre, Kyoto Univerisity

Host : Nobuhiro Hagura (Ikegaya Group)


A hallmark of our motor system is the ability to coordinate the timing and magnitude of motor commands to attain behavioural goals. Recent studies have explored this process by using small mechanical disturbances to observe the speed and complexity of rapid motor corrections. Neurophysiological studies implicate a broad network in fronto-parietal cortices in these feedback corrections, but the specific role of each region is unknown. Here we investigated the function of dorsal premotor cortex (PMd) and parietal area 5 (A5) in feedback control by combining a neural deactivation (cooling deactivation) in non-human primates with a model simulation. To give functional implications to the behavioral results, we generated an optimal feedback control model to observe how deactivations (i.e. reductions) of model parameters impacted feedback responses. Results showed that deactivation of the “feedback controller” impaired both response speed and accuracy, whereas deactivation of “state estimator” impaired only accuracy but not speed of the response. Next, we trained a rhesus monkey to perform a unilateral arm postural task, in which the monkey was required to maintain arm posture while responding to mechanical perturbations. Under normal conditions, the monkey made a quick and accurate perturbation response to return to the original position. When we deactivate PMd, the monkey showed impairments in both response speed and accuracy. On the other hand, when we cooled A5, monkey showed impairment of response accuracy, but not response speed. These results suggest that PMd and A5 have different functions in feedback control: feedback controller and state estimation, respectively. This study demonstrates for the first time that feedback processing for voluntary control involves cortical circuits beyond primary motor cortex.