The neurobridge project aims to investigate an approach to reconcile structural and functional neural connectivity to design dynamical models of the cerebral cortex. The goal is to use both functional and anatomical data to create a model that can simulate the dynamics of the brain and provide insights into its connectivity.
This project involves using both graph theory and machine learning tools to analyze brain connectivity data and apply dynamic modeling techniques to represent how different regions of the brain interact with each other.
- Import and pre-process biological data on structural and functional connectivity.
- Design a dynamical network model to reproduce biological data features.
- Validate the designed model.
The project will use existing tools and algorithms for connectivity analysis:
- Peixoto's clustering algorithm using the graph-tool library.
- Baruzzi's method for integrating structural data and functional data.
The dataset provided contains two matrices:
- Functional Matrix (
matrices.mat): A correlation matrix between 48 brain regions (48x48) across different patients. - Connectivity Matrix (
matrices_HNU1.mat): A structural matrix representing the number of nerve bundles connecting the regions.
- Baruzzi, V., Lodi, M., Sorrentino, F., & Storace, M. (2023). Bridging functional and anatomical neural connectivity through cluster synchronization. Scientific Reports, 13, 22430. DOI: 10.1038/s41598-023-49746-2
- Peixoto, T.P. (2022). Ordered community detection in directed networks. Phys. Rev. E, 106(2), 024305.
This project uses Python and Conda for environment management. The required dependencies are listed below.
- graph-tool (for graph-based algorithms)
-
Create a new Conda environment using the
environment.ymlfile:conda env create -f environment.yml
-
Activate the environment:
conda activate neurobridge
-
Download the data from the open-neurodata portal:
For each subject, download all 10 session files in CSV format (use the HOA atlas connectomes).
-
Organize the data locally:
- Create a folder
./data/subject-<ID>where<ID>is the subject identifier (e.g.,./data/subject-25452). - Inside, create two subfolders:
functional: put all functional connectome CSV files.structural: put all structural connectome CSV files.
- Create a folder
- The meeting logs and additional documentation can be found in the
docs/folder, including:- log.md: A log of all meeting discussions, with a detailed introduction to the project.
TBD
TBD