Related Experiment Video
Updated: May 5, 2026
Mapping Cortical Dynamics Using Simultaneous MEG/EEG and Anatomically-constrained Minimum-norm Estimates: an Auditory Attention Example
Published on: October 24, 2012
A neurocomputational model of the mismatch negativity
Falk Lieder1, Klaas E Stephan, Jean Daunizeau
1Translational Neuromodeling Unit (TNU), Institute of Biomedical Engineering, University of Zurich & ETH Zurich, Zurich, Switzerland ; Laboratory for Social and Neuronal Systems Research, Dept. of Economics, University of Zurich, Zurich, Switzerland ; Helen Wills Neuroscience Institute, University of California at Berkeley, Berkeley, California, United States of America.
This study models auditory cortex prediction errors to explain mismatch negativity (MMN). The model simulates neuronal dynamics, generating realistic MMN waveforms and predicting effects of probability and magnitude.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Auditory Processing
Background:
- Mismatch negativity (MMN) is an electrophysiological response to auditory regularity violations.
- The neural basis of MMN generation remains incompletely understood.
- Generative models offer a framework for understanding predictive coding in the brain.
Purpose of the Study:
- To model the neuronal dynamics underlying MMN generation.
- To investigate the role of prediction errors in auditory cortex.
- To develop a computational model for MMN based on generative principles.
Main Methods:
- A computational model of auditory cortex was developed using generalized (Bayesian) filtering.
- The model simulates continuous updating of a generative model to predict sensory input.
- Neuronal activity reporting prediction errors was modeled as the source of MMN.
Main Results:
- The model successfully generated realistic MMN waveforms.
- The model explained how deviant probability and magnitude influence MMN latency and amplitude.
- Quantitative predictions were made regarding the interaction between deviant probability and magnitude.
Conclusions:
- MMN can be understood as the superposition of electric fields from prediction error signaling neurons.
- The study provides a formal, computationally informed framework for understanding MMN.
- This approach highlights the potential for dynamic causal modeling of electromagnetic responses.
