Poor context integration, the process of incorporating both previous and current information in decision making, is a cognitive symptom of schizophrenia. The maintenance of the contextual information has been shown to be sensitive to changes in excitation-inhibition (EI) balance. Many regions of the brain are sensitive to EI imbalances, however, so it is unknown how systemic manipulations affect the specific regions that are important to context integration. We constructed a multi-structure, biophysically-realistic agent that could perform context-integration as is assessed by the dot pattern expectancy task. The agent included a perceptual network, a memory network, and a decision making system and was capable of successfully performing the dot pattern expectancy task. Systemic manipulation of the agent's EI balance produced localized dysfunction of the memory structure, which resulted in schizophrenia-like deficits at context integration. When the agent's pyramidal cells were less excitatory, the agent fixated upon the cue and initiated responding later than the default agent, which were like the deficits one would predict that individuals on the autistic spectrum would make. This modelling suggests that it may be possible to parse between different types of context integration deficits by adding distractors to context integration tasks and by closely examining a participant's reaction times.
Bibliographical noteFunding Information:
This work was supported by NIH grant P50-MH119569 (ADR) and by a fellowship to OLC from the T32-DA037183 training grant. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
© 2021 Calvin, Redish. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.