Computational psychiatry: from synapses to sentience.

Area of Science:

• Computational psychiatry
• Neuroscience
• Psychopathology

Background:

• Psychopathology is increasingly explained through pathophysiology.
• Generative models underpin sentient brain processing and psychiatric research.
• The brain functions as an organ of inference and prediction.

Purpose of the Study:

• To review a theoretical narrative in computational psychiatry.
• To explain psychopathology via pathophysiology using generative models.
• To explore the brain as an inference and prediction organ.

Main Methods:

• Formal description of neuronal message passing and belief propagation.
• Modeling dysconnections as synaptopathy affecting uncertainty encoding.
• Empirical testing of process theories via computational modeling of neuronal circuits.

Main Results:

• Dysconnections lead to aberrant belief updating and false inference.
• Synaptopathy impacts the encoding of uncertainty and precision.
• Computational modeling aids in understanding attentional set, active inference, learning, and planning.

Conclusions:

• Generative models offer a framework for understanding brain function and psychopathology.
• Computational approaches like in silico experiments (neuropsychology, phenotyping) show promise.
• A computational nosology could advance psychiatric classification across diverse disorders.