Zusammenfassung
A biologically based, multi-cortical computational model was developed
to investigate how ensembles of neurons learn to execute a three-dimensional
reaching task. The model produces outputs of spike trains that can
be analyzed using a variety of multivariate analysis tools. Simulations
show that after learning, the model neurons exhibit broad directional
tuning that depend on the de ned muscle directions of the simulated
arm, and that these neurons form functional clusters within cortical
areas. The utility of the model is demonstrated by testing arm movement
prediction strategies using ensemble activity
Nutzer