The perception of Rorschach inkblots in schizophrenia: A neural network model

Schizophrenia is a psychiatric disorder characterized by a variety of cognitive deficits, including perceptual distortions and hallucinations. In recent years several studies have proposed that schizophrenia may involve a disturbance of "context". We have used a three layer neural network model constructed from an input layer followed by two computational layers to simulate responses of schizophrenic patients to the Rorschach test. In this test subjects respond to a set of ambiguous patterns created by ink blots on paper. Our model proposes that a disturbance of context caused by altered noise-to-signal ratio at the level of the single units, is responsible for schizophrenic responses to the Rorschach test. The assumption that catecholaminergic neurotransmitter systems regulate noise-to-signal ratio in cortical neurons constitutes a link between findings of altered neurotransmitter activity and deficits of cognitive functions requiring contextual integration in schizophrenia. The development of models for specific task deficits in schizophrenia could advance our insights regarding the neurological mechanisms underlying serious mental disorders such as schizophrenia.