Impact of Long Term Plasticity on Information Transmission Over Neuronal Networks.

The realization of bio-compatible nanomachines would pave the way for developing novel diagnosis and treatment techniques for the dysfunctions of intra-body nanonetworks and revolutionize the traditional healthcare methodologies making them less invasive and more efficient. The network of these nanomachines is aimed to be used for treating neuronal diseases such as developing an implant that bridges over the injured spinal cord to regain its normal functionality. Thus, nanoscale communication paradigms are needed to be investigated to facilitate communication between nanomachines. Communication among neurons is one of the most promising nanoscale communication paradigm, which necessitates the thorough communication theoretical analysis of information transmission among neurons. The information flow in neuro-spike communication channel is regulated by the ability of neurons to change synaptic strengths over time, i.e. synaptic plasticity. Thus, the performance evaluation of the nervous nanonetwork is incomplete without considering the influence of synaptic plasticity. In this paper, we focus on information transmission among hippocampal pyramidal neurons and provide a comprehensive channel model for MISO neuro-spike communication, which includes axonal transmission, vesicle release process, synaptic communication and spike generation. In this channel, the spike timing dependent plasticity (STDP) model is used to cover both synaptic depressiofan and potentiation depending on the temporal correlation between spikes generated by input and output neurons. Since synaptic strength changes depending on different physiological factors such as spiking rate of presynaptic neurons, number of correlated presynaptic neurons and the correlation factor among them, we simulate this model with correlated inputs and analyze the evolution of synaptic weights over time. Moreover, we calculate average mutual information between input and output of the channel and find the impact of plasticity and correlation among inputs on the information transmission. The simulation results reveal the impact of different physiological factors related to either presynaptic or postsynaptic neurons on the performance of MISO neuro-spike communication. Moreover, they provide guidelines for selecting the system parameters in a bio-inspired neuronal network according to the requirements of different applications.