Decoding time-resolved neural representation of working memory and motor learning

Unprecedented progress in machine learning have recently revolutionized many sectors of our society like self-driving cars. Over the last few years, we implemented multivariate analytical approaches (MVPA) of magnetoencephalographic activity (MEG) to track the information encoded in the brain (e.g. memory content) during performance of working memory and motor learning tasks. We first decoded the memory content and its selection process during a working memory task. Evidence is presented in favor of a role for the ventrolateral prefrontal cortex in the selection rather than the maintenance of working memory content. Our results show that the memory content was transformed from the initial visual encoding into a different and transiently reactivated memory representation in a posterior brain network Second, we recorded MEG activity during a motor learning task. Behavioral analysis showed that a large proportion of early motor skill learning occur during small time windows of rest during the practice session. MVPA analyses show that during these rest periods, the brain replays at a faster rate the sequence of the events it just learned. The amount of replay predicted the magnitude of this form of rapid consolidation in early skill learning Multivariate analyses applied to time-resolved MEG neural signal opens a unique window to study how mental representations are dynamically manipulated and transformed in the brain.For more information go tohttps://hpc.nih.gov/trainingAir date...
Source: Videocast - All Events - Category: General Medicine Tags: Upcoming Events Source Type: video