Modeling of topology-dependent neural network plasticity induced by activity-dependent electrical stimulation

Activity-dependent electrical stimulation can induce cerebrocortical reorganization in vivo by activating brain areas using stimulation derived from the statistics of neural or muscular activity. Due to the nature of synaptic plasticity, network topology is likely to influence the effectiveness of this type of neuromodulation, yet its effect under different network topologies is unclear. To address this issue, we simulated small-scale three-neuron networks to explore topology-dependent network plasticity. The induced neuroplastic changes were evaluated by network coherence and unit-pair mutual information measures. We demonstrated that involvement of monosynaptic feedforward and reciprocal connections is more likely to lead to persistent decreased network coherence and increased network mutual information independent of the global network topology. On the contrary, disynaptic feedforward connections exhibit heterogeneous coherence and unit-pair mutual information sensitivity that depends strongly upon the network context.