A neural model of amygdala playing an essential role in formation of brainmaps for accomplishing spatial tasks

To clarify the neuronal mechanism of navigation behavior, we propose a mode l of total system of neural network for a rodent to accomplish a navigation task. The total system consists of a head direction map, a position recogn ition map, a goal position map, a task dependent map, a value judgement sys tem, and a motor system. We show the neuronal mechanism by which these maps are self-organized during the navigation task, and the functional roles of these maps in the achievement of the task. In the task dependent map, the important places in the task field are memorized as dynamical attractors of the network activity, and the task is accomplished by learning temporary t he correlation between dynamical attractors, depending on the input from th e value judgement system and on the strength of attention. (C) 2001 Elsevie r Science B.V. All rights reserved.