A neural network of obsessive- compulsive disorder: modelling cognitive disinhibition and neurotransmitter dysfunction.

While significant advances have been made in documenting both neurotransmitter and neuropsychological dysfunction in obsessive-compulsive disorder (OCD), there remains a need for theoretical models to account for their relationship. A neural network model of OCD was developed to provide a rigorous simulation of the relationship between the cognitive disinhibition and serotonin/dopamine dysfunction that characterize this disorder. An architecture-specific recurrent neural network of the Elman type was able to model the cognitive disinhibition that is apparent when OCD patients are compared with other anxiety disorder patients on a modified Stroop (Temporal Stroop) test, with OCD patients showing reduced negative priming (shorter reaction times to previously ignored stimuli). Lesions to either the color gain parameter (reflective of serotonergic dysfunction) or to the context gain parameter (reflective of dopaminergic dysfunction) resulted in decreased semantic negative priming. Neural network modelling provides a surprisingly coherent perspective of the psychobiology of OCD, simulating both reduced cognitive disinhibition as well as neurotransmitter dysfunction. Copyright 2000 Harcourt Publishers Ltd.