Making Sense of Computational Psychiatry

Area of Science:

• Psychiatric Neuroscience
• Computational Psychiatry
• Systems Neuroscience

Background:

• The concept of "models" is central to psychiatric research.
• Defining the utility and scope of these models is crucial for scientific advancement.

Purpose of the Study:

• To clarify the definition and utility of computational models in psychiatry.
• To explore the advantages of dynamical systems over standard statistical methods.
• To propose strategies for improving model building in computational psychiatry.

Main Methods:

• Conceptual analysis of "models" in psychiatry.
• Evaluation of computational models for accuracy and power.
• Reconceptualization of statistical models as dynamical systems.
• Exploitation of neuroimaging data for systems-level analysis.

Main Results:

• Computational models offer significant added value in terms of accuracy and predictive power.
• Standard statistical methods have limitations that can be overcome by dynamical systems approaches.
• Neuroimaging data contains deep systems-level information for advancing psychiatric understanding.

Conclusions:

• A clearer understanding of model utility is essential for computational psychiatry.
• Treating statistical models as dynamical systems can expand analytical explanatory power.
• Leveraging systems-level neuroimaging data can mitigate cognitive biases in psychiatric research.