Deviance detection via competitive inhibition between local neocortical ensembles

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Deviance detection (DD) is crucial for processing unexpected stimuli.
• Previous theories proposed specialized cell types or brain area routing for DD.
• The neural mechanisms underlying DD remain incompletely understood.

Purpose of the Study:

• To propose and test a novel theory where deviance detection emerges from local network interactions within a neocortical column.
• To investigate the role of competitive inhibition and selective (dis)inhibition in neural dynamics.
• To computationally model phenomena related to neural tuning shifts and sensory context storage.

Main Methods:

• Developed a theoretical framework based on competitive inhibition between neuronal ensembles.
• Utilized computational modeling with two neural network types of varying biophysical detail.
• Simulated neural tuning shifts across neurons, time, and stimulus history.
• Tested hypotheses on emergent dynamics and connectivity parameter robustness.

Main Results:

• Demonstrated that deviance-driven neural dynamics can emerge from local network interactions.
• Showed that ensemble priming via competitive inhibition under selective (dis)inhibition can store sensory context.
• Validated the model against experimentally observed phenomena, including shifts in neural tuning.
• Found that DD does not necessitate specialized input from other brain regions.

Conclusions:

• Deviance detection can be explained by local network mechanisms within neocortical columns.
• Competitive inhibition and selective (dis)inhibition provide a local framework for sensory context storage.
• This novel paradigm challenges existing theories and resolves confounding aspects of sensory encoding and predictive processing.