Related Experiment Video
Updated: May 3, 2026
08:51
Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms
Published on: November 1, 2019
5.0K
An artificial network model for estimating the network structure underlying partially observed neuronal signals.
Misako Komatsu1, Jun Namikawa2, Zenas C Chao2
1RIKEN Brain Science Institute, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan; Graduate School of Science and Engineering, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8502, Japan.
Neuroscience Research
|February 18, 2014
Summary
This study introduces a new method to map brain networks by accounting for unobserved neural activity. This approach reveals insights into cortical dynamics and neuronal interactions beyond directly recorded signals.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Network Science
Background:
- Quantifying neuronal interactions is crucial for understanding brain function.
- Previous methods often analyze recorded signals in isolation, neglecting unobserved network components.
Purpose of the Study:
- To develop a novel approach for estimating neuronal interactions by incorporating unobserved network structures.
- To model brain activity using a recurrent network with both observable and unobservable units.
Main Methods:
- Proposed a recurrent network model with observable (recorded activity) and unobservable (unobserved structures) units.
- Estimated connective weights representing interaction intensities from recorded multi-channel brain signals.
- Applied the model to macaque monkey brain recordings.
Main Results:
- Successfully obtained robust and physiologically relevant network structures.
- Identified inversely correlated interactions between excitatory and inhibitory neuronal populations, reflecting cortical dynamics.
- Demonstrated consistency with established models of cortical local circuits.
Conclusions:
- Incorporating unobserved structures into network estimation offers theoretical advantages.
- The novel approach provides deeper insights into brain dynamics than methods analyzing only observable signals.
- This method has potential for advancing our understanding of complex neural systems.
