Effect of Potassium Conductance Characteristics on Pattern Matching in a Model of Dendritic Spines

Abstract

Pattern matching is the ability to produce a stronger response to a previously learned pattern than to a novel pattern. Is it possible that the ability of mammals to recognize patterns is due to pattern matching by arrays of single neurons? Previous modeling studies have shown that plausible neuron models can match patterns of binary synaptic inputs. This study investigates the plausibility of analog pattern matching in a model of a dendrite with spines. Each dendritic spine includes a synaptic conductance and a calcium dependent potassium current whose properties depend on the previously learned value of synaptic conductance. The input to the model is a pattern of synaptic activation, and output of the model is the time integral of membrane potential (signal strength). Simulations show that signal strength is greatest when synaptic input equals the previously learned value, and is smaller when components of the synaptic input pattern are either smaller or larger than corresponding components of the previously learned pattern. The decrease in signal strength is proportional to the difference between input pattern and previously learned pattern. Pattern matching is robust to large changes in parameter values.