Signal propagation and separation in a self-organized three-dimensional neural network with activity-dependent conduction

Abstract Activity-dependent myelination makes conduction delays and connection strengths depend on the history of transmitted signals, but how this shapes signal propagation in spatially embedded networks is poorly understood. We study spike propagation in three-dimensional neural networks grown by an energy-based self-organization model. The delay of a connection is set by the lengths and thicknesses of the axon and the dendrite, and its weight by the thinner of the two and by dendritic attenuation; thickness grows with passing spikes up to a maximum and decays without them, and short-term synaptic depression limits transmission. In networks of 100–500 neurons, the transition from local activity to activation of the whole network is sharp as the connection weight increases. Without depression, thickness plasticity drives self-sustained activity, and a dimensionless synaptic load separates a relay regime from an onset-detector regime. We propose to extract the signal pathway from the growth of connection weights and introduce measures for comparing pathways: pathways from different inputs barely overlap, and nested stimuli produce nested pathways. Community detection by modularity maximization does not recover the signal pathway, because the pathway is an elongated tree rather than a dense group; we also show that the spatial null model of modularity must be normalized. Comparison with random networks built without growth shows that growth makes connections local, and a random network whose connection probability decays with distance as in the grown network and whose axon capacity is limited reproduces the functional properties of the grown network.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-12
DOI
https://doi.org/10.21203/rs.3.rs-11535023/v1
Primary Topic
Neural dynamics and brain function
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Signal propagation and separation in a self-organized three-dimensional neural network with activity-dependent conduction

Victor Yu. Trubitsin
Research Square
Neural dynamics and brain function
preprint

Signal propagation and separation in a self-organized three-dimensional neural network with activity-dependent conduction

Victor Yu. Trubitsin
preprint en

Abstract

Abstract Activity-dependent myelination makes conduction delays and connection strengths depend on the history of transmitted signals, but how this shapes signal propagation in spatially embedded networks is poorly understood. We study spike propagation in three-dimensional neural networks grown by an energy-based self-organization model. The delay of a connection is set by the lengths and thicknesses of the axon and the dendrite, and its weight by the thinner of the two and by dendritic attenuation; thickness grows with passing spikes up to a maximum and decays without them, and short-term synaptic depression limits transmission. In networks of 100–500 neurons, the transition from local activity to activation of the whole network is sharp as the connection weight increases. Without depression, thickness plasticity drives self-sustained activity, and a dimensionless synaptic load separates a relay regime from an onset-detector regime. We propose to extract the signal pathway from the growth of connection weights and introduce measures for comparing pathways: pathways from different inputs barely overlap, and nested stimuli produce nested pathways. Community detection by modularity maximization does not recover the signal pathway, because the pathway is an elongated tree rather than a dense group; we also show that the spatial null model of modularity must be normalized. Comparison with random networks built without growth shows that growth makes connections local, and a random network whose connection probability decays with distance as in the grown network and whose axon capacity is limited reproduces the functional properties of the grown network.

Research Square
Udmurt Federal Research Center, Ural Branch of the Russian Academy of Sciences (RU)
Ministry of Science and Higher Education of the Russian Federation
Neural dynamics and brain function
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon