WhoBPyT - Whole-Brain Modelling in PyTorch

Title

Short description and the goals for the OHBM BrainHack

WhoBPyT is a Python library for connectome-based neurophysiological modelling of neuroimaging and other macroscale brain data. At its core is a deep learning-based approach for model fitting and parameter estimation, that capitalizes on the in-built algorithmic differentiation and optimization functionality of PyTorch.

The current WhoBPyT code base is a consolidation of several projects that have used this computational technique to study mechanisms underlying brain dynamics in TMS-EEG, stereotactic EEG, resting-state fMRI.

We have completed a major refactor of the code base that generalizes several key functionalities and improves overall readability. The goal of this OHBM BrainHack project is to improve the first-time user experience with WhoBPyT, principally through docs and tutorial examples on computational modeling of EEG and fMRI data.

Link to the Project

https://github.com/GriffithsLab/whobpyt

Image/Logo for the OHBM brainhack website

https://drive.google.com/uc?export=view&id=1wJaFEEk6hDfeOCjix2v_BhmWmNak9VzO

Project lead

John Griffiths

Main Hub

Seoul

Link to the Project pitch

Pitch video will be added ahead of the event.

Other hubs covered by the leaders

[X] Seoul
[ ] Hybrid (Asia / Pacific)
[ ] Hybrid (Europe / Middle East / Africa)
[X] Hybrid (Americas)

Skills

Python PyTorch Github EEG fMRI MNE Nilearn Numerical simulation Reading+writing

Good first issues

Test the EEG/fMRI modelling examples
Run examples with alternative data
Improve data visualization in examples
Add any missing docstrings
Add missing items to the README
...

Twitter summary

No response

Short name for the Discord chat channel (~15 chars)

PyTorchBrainModelling

Please read and follow the OHBM Code of Conduct

[X] I agree to follow the OHBM Code of Conduct during the hackathon

ohbm / hackathon2024