Multimap formation in visual cortex.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Developmental Neuroscience

Background:

• Extrastriate visual areas (V2, V4) possess neurons selective for complex shapes with topographic organization.
• Simultaneous development of multiple interdigitated maps (multimaps) is challenging due to competing neuronal demands for local response diversity and global coverage uniformity.

Purpose of the Study:

• To investigate the computational mechanisms underlying the development of multiple, interdigitated cortical maps (multimaps).
• To identify key plasticity rules that enable the formation of diverse response types within a shared topographic space.

Main Methods:

• Utilized computer simulations to model cortical map development.
• Implemented a novel hybrid developmental rule combining spatial and activity-dependent learning criteria within learning eligibility regions (LERs).

Main Results:

• Demonstrated that plasticity confined to LERs, with learning rates dependent on cell activity, successfully produces multimaps for multiple feature types.
• Showed that a single feature type results in a V1-like map with a "salt-and-pepper" structure under the same developmental rules.

Conclusions:

• Cortical maps with fine mixtures of response types are a product of a generic developmental scheme, not a breakdown of mechanisms.
• This scheme allows for mapping diverse cell response properties across shared topographic spaces in various cortical areas.