Simulating the response of biological neurons to noise through the leaky integrate-and-fire model
Abstract
Given opposing literature regarding the effect of noise on neurons, there is a need to investigate what specific noise parameters are beneficial. This paper thus simulates the response of the Leaky Integrate-and-Fire (LIF) neuron to three different inputs: (i) a constant, noise-free input, (ii) a low-noise time-varying input, and (iii) a high-noise synaptic input, using the Python simulation software Brian 2. Comparing the simulations based on spiking reliability reveals that synaptic input produces the most reliable spiking behavior. Here, spiking reliability refers to the quickness of the response of the neuron to the input. Our results support the hypothesis that noise may improve the reliability of spiking and also suggest that the usefulness of noise may be dependent on the characteristics of the noise itself. Lastly, we briefly outline other noise parameters, as a recommendation, that can be varied to gain deeper insight into the functional role of noise in neurons.