Model-Based Approaches to Investigating Mismatch Responses in Schizophrenia

Area of Science:

• Computational cognitive neuroscience
• Psychiatry
• Neuroscience

Background:

• Alterations in mismatch responses (neural activity to unexpected stimuli) are potential biomarkers for schizophrenia.
• Computational methods provide deeper insights into neural mechanisms and cognitive functions underlying these alterations.

Purpose of the Study:

• To provide an overview of model-based approaches for studying mismatch responses in schizophrenia.
• To explore four complementary computational perspectives: connectivity, decoding, neural network, and cognitive models.

Main Methods:

• Connectivity models: Inferring effective connectivity patterns in brain regions.
• Decoding models: Using spatiotemporal patterns for classifying sensory violations or participants.
• Neural network models: Employing deep convolutional neural networks for classification and data analysis.
• Cognitive models: Quantifying mismatch responses in terms of perceptual prediction signaling and updating.

Main Results:

• Review of recent computational psychiatry studies on mismatch responses in schizophrenia.
• Demonstration of how different model types offer complementary insights.
• Identification of methodological advancements and findings from applying these models.

Conclusions:

• Model-based approaches are valuable for advancing the study of mismatch responses in schizophrenia.
• Future research should leverage these techniques to further elucidate underlying neural and cognitive disruptions.
• Suggests directions for future work applying model-based techniques.