CONDUCTANCE VERSUS CURRENT-BASED INTEGRATE-AND-FIRE NEURONS: IS THERE QUALITATIVELY NEW BEHAVIOUR?

The framework of Hodgkin-Huxley conductance-based models provides an excellent description of the electrical behaviour of neurons. However, their non-linear voltage-dependent membrane currents are hard to handle analytically and it has been difficult to gain insight into the quantitative behaviour of ensembles of such neurons. A much-simplified model neuron, the integrate-and-fire (IF) neuron captures many of the broad features that biological neurons share and has become a standard component of models of large recurrent-networks. A cortical neuron in an active network receives a massive synaptic bombardment. This input is commonly approximated as a fluctuating current. Synaptic drive is, however, better modelled by a fluctuating conductance. Recently, the effect this conductance change has on the response of neu-rons has been brought to the attention of theorists and experimentalists [1, 2]. It was demonstrated that synaptic bombardment in vivo can lead to four-fold conductance increases, as compared to the quiescent case. Such an increase can have dramatic effects on the integrative properties of the neuron- effects that are neglected in current-based models. Here we examine the impact of this conductance change analytically and make a comparison with the case of current-based mode