Related Experiment Videos
"What" and "where" in visual working memory: a computational neurodynamical perspective for integrating FMRI and
Gustavo Deco1, Edmund T Rolls, Barry Horwitz
1Institucion Catalana de Recerca, Estudis Avançats (ICREA) and Universitat Pompeu Fabra.
Journal of Cognitive Neuroscience
|May 29, 2004
Summary
This study models prefrontal cortex (PFC) function in working memory using an integrate-and-fire network. It shows how different inhibition levels in a homogeneous network can explain distinct object and spatial memory processing in dorsal and ventral PFC.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Cognitive Neuroscience
Background:
- The prefrontal cortex (PFC) is implicated in working memory and cognitive processes.
- Existing models propose PFC organization based on stimulus domain or process.
- Understanding PFC functional organization during working memory tasks remains a challenge.
Purpose of the Study:
- To model single-neuron and fMRI data of short-term memory in the PFC.
- To investigate mechanisms underlying working memory activity in delay tasks.
- To explain topographical and functional organization of the PFC.
Main Methods:
- Utilized an integrate-and-fire neural network model.
- Modeled "what"-then-"where" delay tasks with object and spatial information.
- Incorporated attractor networks with hierarchical pools and global inhibition.
- Simulated AMPA and NMDA synaptic dynamics for realistic spiking.
Main Results:
- A biased competition model of attention explained the selection of relevant neuronal populations.
- Higher inhibition levels in the model reproduced dorsal PFC fMRI data (spatial processing).
- Lower inhibition levels reproduced ventral PFC fMRI data (object processing).
- Network topography was homogeneous, yet functional differences emerged.
Conclusions:
- The model explains how spatial versus object short-term memory is implemented in the PFC.
- fMRI BOLD signals in different PFC regions may reflect varying inhibition levels, not necessarily strict spatial/object specialization.
- Functional differentiation in the PFC can arise from network dynamics rather than fixed spatial topology.