Related Experiment Video
Updated: Jul 14, 2026
08:51
Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms
Published on: November 1, 2019
Imposing biological constraints onto an abstract neocortical attractor network model
Christopher Johansson1, Anders Lansner
1School of Computer Science and Communication, Royal Institute of Technology, SE-100 44 Stockholm, Sweden. cjo@nada.kth.se
Neural Computation
|May 25, 2007
Summary
This study presents a neocortex model, revealing that specific neuron aggregation in minicolumns and hypercolumns enhances storage capacity. The model demonstrates efficient scaling under biological constraints, offering insights for both neuroscience and engineered systems.
Area of Science:
- Computational Neuroscience
- Systems Neuroscience
- Artificial Intelligence
Background:
- The neocortex exhibits modular organization into minicolumns and hypercolumns.
- Understanding the relationship between biological constraints and network performance is crucial.
Purpose of the Study:
- To develop and analyze an abstract model of the neocortex based on its modular structure.
- To connect network properties to biological neuronal properties and storage capacity.
- To evaluate the scalability of the model under biological constraints.
Main Methods:
- Development of an abstract neocortical model featuring minicolumns and hypercolumns.
- Analysis of network properties and their relation to biological constraints (activity, connectivity).
- Simulation of a full-scale instance to assess storage capacity and scalability.
Main Results:
- The model predicts that only a small percentage of hypercolumns are active at any given time.
- High storage capacity is achieved when 20-30 pyramidal neurons form a minicolumn.
- Grouping 50-60 minicolumns into a hypercolumn is necessary for optimal storage capacity.
Conclusions:
- The proposed neocortical model scales effectively with biologically constrained parameters.
- The findings provide insights into optimal neuronal aggregation for high storage capacity in the neocortex.
- The model's scalability makes it relevant for engineered systems inspired by brain architecture.
