Cracking the barcode of fullerene-like cortical microcolumns
Arturo Tozzi1, James F Peters2, Ottorino Ori3
1Center for Nonlinear Science, University of North Texas, 1155 Union Circle, #311427 Denton, TX 76203-5017, USA; Computational Intelligence Laboratory, University of Manitoba, Winnipeg, MB, R3T 5V6, Canada.
Neuroscience Letters
|March 1, 2017
Summary
This study proposes a new model for brain function, suggesting neural activity relies on topological transformations within cortical microcolumns, not just logic circuits. This fullerene-like lattice model could inspire robust artificial neural networks.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Artificial Intelligence
Background:
- Current artificial neural systems often model the neural code as logic circuits, which may not fully explain complex brain functions.
- The cortical microcolumn is considered the fundamental unit of brain structure and function.
Purpose of the Study:
- To propose an alternative model for neural activity based on topological transformations and symmetry constraints within cortical microcolumns.
- To explore the potential of fullerene-like lattices to represent neural computations and inspire artificial neural networks.
Main Methods:
- Modeling cortical microcolumns as flattened, fullerene-like two-dimensional lattices with nodes representing pyramidal neurons.
- Analyzing topological transformations and symmetry constraints within these lattices.
- Investigating the resemblance of activated neuron combinations to barcodes.
Main Results:
- Nervous activity may depend on topological transformations and symmetry constraints at the microcolumn level, rather than solely logic circuits.
- The fullerene-like lattice model provides a framework for understanding neural computations as barcode-like patterns.
- This model shows potential for developing plastic, robust, and fast artificial neural networks.
Conclusions:
- The proposed model offers a novel perspective on the neural code, moving beyond traditional logic circuit assumptions.
- Fullerene-like lattices derived from microcolumn topology can represent complex neural activity.
- This research paves the way for advanced artificial neural systems in robotics and other applications.
