Computing with neural synchrony.

Area of Science:

• Computational neuroscience
• Neural coding
• Sensory processing

Background:

• Traditional neural computation theories focus on firing rates, but experimental data suggest spike timing is important.
• Neural synchrony, characterized by correlated spike timing, is sensitive to input correlations and may play a key computational role.
• The precise function and advantages of neural synchrony in computation remain unclear.

Purpose of the Study:

• To theoretically investigate the role and functional advantages of neural synchrony in computation.
• To explore how neural synchrony relates to stimuli in heterogeneous neural populations.
• To demonstrate how synchrony-based computation can emerge and extract relevant sensory information.

Main Methods:

• Theoretical modeling of neural computation.
• Introduction of the concept of 'synchrony receptive field' to link stimuli and synchrony.
• Analysis of emergent neural circuitry using spike-timing-dependent plasticity.
• Application of the theory to examples across different sensory modalities.

Main Results:

• Neural synchrony gains computational relevance in populations with heterogeneous neuronal properties.
• Synchrony patterns represent structural or invariant features within stimuli.
• Postsynaptic neurons can detect these synchrony patterns.
• Spike-timing-dependent plasticity can facilitate the spontaneous emergence of the necessary neural circuitry.
• The model successfully extracts relevant information from realistic sensory stimuli, such as identifying odors amidst distractors.

Conclusions:

• Relative spike timing, or neural synchrony, has significant computational relevance beyond simple firing rates.
• Synchrony-based computation enables neural circuits to extract complex information from sensory inputs.
• This framework suggests novel neural network models for sensory processing with enhanced computational capabilities.