BrainQCNet : Detection of Artifacts on Brain T1-weighted Scans

BrainQCNet is a software that automatically detects the presence of defaults on brain structural MRI scans.

It is based on a Deep Learning algorithm, you can find more details on how it was built on the paper here.

To run BrainQCNet on your data, you can either use the BIDS-app that:

• works on Nvidia GPU machines with CUDA here,
• works on CPU machines here.

A list of Nvidia GPUs with CUDA technology is provided here.

🔔 We strongly recommend users to run the app on a machine with CUDA-compatible GPU ! 🔔

Explicative videos

Videos are available on YouTube in order to guide you when using BrainQCNet:

• GPU & CUDA users
• CPU users

How to report errors

Users can get help using the brainqcnet-users mailing list.

All bugs, concerns and enhancement requests for this software can be submitted here for the GPU version and here for the CPU version.

Citation

When using BrainQCNet, please include the following citation:

BrainQCNet: A Deep Learning attention-based model for the automated detection of artifacts in brain structural MRI scans., Mélanie Garcia, Nico Dosenbach, Clare Kelly; Imaging Neuroscience 2024; 2 1–16. doi: https://doi.org/10.1162/imag_a_00300