An implementation of a simple neuron model in field programmable analog arrays

A minimal neuron model, the Morris-Lecar model, is implemented on field programmable analog arrays (FPAAs). Our approach is to solve the differential equation describing the model in a similar way a computer solves the same problem: numerically integrate the differential equation by making arithmetic operations on voltage mode circuits of the FPAAs. The results demonstrate that biologically relevant dynamics can be observed from the electronic neuron despite limitations on the configurability of the FPAAs. Such models can be run accurately in real-time or many orders of magnitude faster than real-time. FPAAs are feasible candidates for implementation of neuron models using off-the-shelf software-reconfigurable analog circuit elements.