.png){kind=icon}
.png){kind=icon}
Return
Excitable dynamics simplify neural connectomes
.png){kind=icon}
OA
AI
DOI：10.1016/j.xcrp.2025.102510.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Fiber networks that link brain regions form the structural basis of neural dynamics and function. Weighted networks offer detailed connectivity information, but their treatment poses methodological challenges. Here, we report that excitable dynamics-a common mechanism in biological and artificial networks- simplify network representation by making weighted and binary networks dynamically equivalent for an appropriate network threshold. Application of the framework to empirical brain connectivity shows that binary networks can reproduce functional connectivity patterns observed in human brain data, suggesting that neural activity patterns are predominantly shaped by the strongest structural connections. Moreover, the approach significantly reduces memory and processing time for network representation, analysis, and simulations. In artificial neural networks, binarized networks maintain performance while drastically reducing the number of parameters, making them highly efficient. These findings simplify empirical network analyses and support efficient artificial neural network design.
Keywords:
FUNCTIONAL CONNECTIVITY
NETWORK
BRAIN
MODELS
MRI
DIFFUSION
NEURONS
ORGANIZATION
TRACTOGRAPHY
SPECIFICITY
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.