The Locus Concept and Its Application to Neural Analogs

A new approach to analog simulation and study of the neuron is proposed. This approach is based on recent physiological evidence which indicates that the individual nerve cell is functionally much more complex than the classical view of a synaptic region coupled directly to a spike or impulse-generating region. At least two different intermediate regions have been found. One provides a reliable low-frequency timing or pacemaker function; the other provides nonlinear amplification of both the synaptic and the pacemaker potentials. In addition, the synaptic regions have been found to provide a large variety of complicated interneural transfer functions. In the view presently held by many physiologists, the spatial distribution of these functionally distinct regions within a single neuron would determine its information-processing capabilities. The behavior of each of the functionally distinct regions of the neuron is discussed in this paper. Simple transistor circuits which may be used to simulate individual regions are also described. Groups of these circuits may be connected to form analogs of the entire neuron or any part thereof. Special emphasis is placed here on the synaptic functions, with only a cursory discussion being given for the other regions. It is hoped that networks of the type described in this paper will be of considerable use in future studies of the information processing capabilities of single nerve cells.