Ketamine Affects Prediction Errors about Statistical Regularities: A Computational Single-Trial Analysis of the Mismatch Negativity
Journal of Neuroscience June 19, 2020 Lilian Weber, Andreea O. Diaconescu, Christoph Mathys et al. 82 citations
The auditory mismatch negativity (MMN) is reduced in schizophrenia and can also be reduced by NMDA receptor (NMDAR) antagonists, suggesting impaired predictive coding. This study tested the theory that perceptual inference depends on NMDAR-dependent hierarchical precision-weighted prediction errors (PEs). Using a hierarchical Bayesian model on single-trial EEG data from healthy volunteers given the NMDAR antagonist S-ketamine in a placebo-controlled, double-blind, within-subject design, the analysis showed that low-level PEs (about stimulus transitions) appear early (102-207 ms), while high-level PEs (about transition probability) appear later (152-199 and 215-277 ms). Ketamine significantly diminished high-level PE responses, indicating NMDAR antagonism disrupts inference on abstract statistical regularities and impairs hierarchical Bayesian inference about the world's statistical structure.