Schizophrenic speech as a disordered trajectory in a collapsed cognitive "Small-World".

Area of Science:

• Computational neuroscience
• Psycholinguistics
• Graph theory

Background:

• Schizophrenia is characterized by disorganized speech, including goal confusion and incoherence.
• Understanding the neuropathological basis of these speech alterations is crucial.
• Existing models may not fully capture the dynamic nature of schizophrenic discourse.

Purpose of the Study:

• To explore new theoretical instruments for understanding schizophrenic speech.
• To investigate the potential of neural network models and semantic graph representations.
• To correlate speech graph properties with disease severity and treatment efficacy.

Main Methods:

• Simulating neuropathologic disorders of schizophrenia using goal-directed neural network models.
• Generating geometric representations of speech trajectories from real patient data using semantic graphs.
• Analyzing the structural properties of speech graphs (e.g., order-to-random transitions, small-world properties).

Main Results:

• Neural models can simulate goal confusion and incoherence in schizophrenic speech.
• Geometric representations of speech trajectories can be derived from spontaneous patient speech.
• A potential transition from ordered to random graph structures with increasing disease severity is hypothesized.

Conclusions:

• Goal-directed neural networks and semantic graphs provide novel approaches to studying schizophrenic speech.
• Speech graph analysis may serve as a biomarker for disease progression in schizophrenia.
• Geometric representations of speech could aid in evaluating the effectiveness of therapeutic interventions.