Connection-strength estimation of neuronal networks by fitting for Izhikevich model

Abstract

Recently, a lot of research using multi-neuron recording has been done, however, there are many problems with extracting the features from obtained spike time series which is huge amount and complex. Here we introduce a new method to estimate synaptic connection strengths between neurons by fitting for Izhikevich model using maximum likelihood estimation. We demonstrate that our method is able to estimate connection strengths from spike time series given by simulated neural ensemble and able to estimate non-connectivity between two independent cultured neuronal networks. These results suggest that our method can apply for network and plasticity analysis of neuronal networks. © 2012 The Institute of Electrical Engineers of Japan.