Computational Psychiatry

Abstract

Psychiatric disorders such as autism and schizophrenia arise from abnormalities in brain systems that underlie cognitive, emotional and social functions. The brain is enormously complex and its abundant feedback loops on multiple scales preclude intuitive explication of circuit functions. In close interplay with experiments, theory and computational modeling are essential for understanding how, precisely, neural circuits generate flexible behaviors and their impairments give rise to psychiatric symptoms. This Perspective highlights recent progress in applying computational neuroscience to the study of mental disorders. We outline basic approaches, including identification of core deficits that cut across disease categories, biologically-realistic modeling bridging cellular and synaptic mechanisms with behavior, model-aided diagnosis. The need for new research strategies in psychiatry is urgent. Computational psychiatry potentially provides powerful tools for elucidating pathophysiology that may inform both diagnosis and treatment. To achieve this promise will require investment in cross-disciplinary training and research in this nascent field. © 2014 Elsevier Inc. All rights reserved. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.Neither Dr. Wang nor Dr. Krystal has any financial interests relevant to the advancement of computational psychiatry. to and engage with psychiatric research. Second, it will be important to develop a cadre of young psychiatrists who learn computational modeling, lest computational psychiatry.

Visual Mismatch and Predictive Coding: A Computational Single-Trial ERP Study

“…Later, this interpretation was followed by the "regularity violation" hypothesis (Winkler, 2007), according to which the MMN signals a difference between the current stimulus and expectations based on prior information that might not only represent a sensory memory trace but also more complex or abstract rules extracted from regular relationships between preceding stimuli (e.g., conditional probabilities; Paavilainen et al, 2007;Stefanics et al, 2009Stefanics et al, , 2011for review, see Paavilainen, 2013). This interpretation is compatible with the most recent view of the MMN as an expression of pwPEs during PC (Friston, 2005;Baldeweg, 2006;Stephan et al, 2006;Wacongne et al, 2011;Lieder et al, 2013a;Stefanics et al, 2015). In fact, a PC view of MMN can be seen as mathematically formalizing ideas already inherent to the earlier "regularity violation" hypothesis.…”

“…Face stimuli, 10 female and 10 male Caucasian models, were selected from the Radboud Faces Database (Langner et al, 2010;www.rafd.nl) based on their high percentage of agreement on emotion categorization (98% for happy, 92% for fearful faces). To control low-level image properties, we used the SHINE toolbox (Willenbockel et al, 2010) to equate luminance and spatial frequency content of grayscale images of the selected happy and fearful faces.…”

“…We used the HGF Mathys et al, 2014) to simulate computational trajectories to create parametric regressors for the general linear model (GLM) analysis. The HGF is a generative (Bayesian) model of perceptual inference and learning that represents a variant of PC in the temporal domain and has been used in several recent studies to investigate hierarchical PE responses in the brain (Iglesias et al, 2013;Hauser et al, 2014;Schwartenbeck et al, 2015;Vossel et al, 2015;Lawson et al, 2017;Powers et al, 2017). It is implemented in the freely available open source software TAPAS (http://www.…”

“…The PC interpretation of MMN is supported by much, albeit mostly indirect, experimental evidence (Garrido et al, 2007Stefanics and Czigler, 2012;Phillips et al, 2015;Auksztulewicz and Friston, 2016;Chennu et al, 2016). By contrast, experimental studies based on formal trial-by-trial computational quantities are rare, almost entirely restricted to the auditory domain, and typically focused on specific sensors or time windows (Lieder et al, 2013b;Kolossa et al, 2015;Jepma et al, 2016). Here, we go beyond previous investigations and use a Bayesian model [the Hierarchical Gaussian Filter (HGF)] to examine whether visual mismatch responses reflect pwPEs, a hallmark of PC.…”

“…We used the HGF to generate pwPE trajectories and tested the implication by PC, that trial-bytrial brain activity would reflect these computational quantities. In addition, we applied a trial-wise change detection (CD) model (Lieder et al, 2013b) and evaluated the explanatory power of both hypotheses by statistical model comparison. Finally, we analyzed visual mismatch responses [i.e., visual MMN (vMMN) responses; for review, see Stefanics et al, 2014;Kremláček et al, 2016] obtained with traditional averaging and subtraction methods, and compared the results to those obtained by modeling.…”