A Neural Network Model of Cortical Information Processing in Schizophrenia II - Role of Hippocampal-Cortical Interaction: A Review and a Model

Abstract

This paper considers the relevance of hippocampal dysfunction to symptom production in schizophrenia with the help of an integrative computational model of cortical processing. First, evidence for involvement of the hippocampus was systematically reviewed. Then, a neural network model of hippocampal-cortical interaction was proposed. The effects of a dysfunction in hippocampal-cortical interaction were explored through simulation experiments. Neural network simulation demonstrated that when the hippocampal function of classification and reduction of overlap in input is sub-optimal, increasing correlation among input patterns compromised storage capacity of the cortical network. Hippocampal dysfunction predisposes to the emergence of spurious retrieval states in cortical information processing. This provides a model for psychotic symptom formation in schizophrenia.