Effect on information transfer of synaptic pruning driven by spike-timing-dependent plasticity.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Information Theory

Background:

• Spike-timing-dependent plasticity (STDP) drives self-organization in neural systems.
• The functional role of STDP-induced synaptic pruning requires further investigation.

Purpose of the Study:

• To explore the functional role of synaptic pruning mediated by STDP from an information-theoretic perspective.
• To investigate how synapse removal impacts information transfer in neural networks.

Main Methods:

• Utilized temporally correlated stimuli delivered to an input layer of neurons.
• Refined synapses on an output neuron's dendrite, thereby modifying its receptive field, using STDP.
• Calculated mutual information between input and output spike trains via the context tree method.

Main Results:

• Synapse removal was shown to enhance information transfer capabilities.
• The principle of "less can be more" was observed under specific conditions.
• Results were contingent on the balance between potentiation and depression in STDP and the temporal scale of input correlations.

Conclusions:

• Synaptic pruning via STDP can optimize neural information processing.
• The efficacy of pruning depends on STDP parameters and input signal characteristics.
• This study provides insights into the self-organization principles governing neural receptive fields.