DiFuMo atlases

1. We provide Dictionaries of Functional Modes “DiFuMo” that can serve as atlases to extract functional signals, eg to serve as IDPs, with different dimensionalities (64, 128, 256, 512, and 1024). These modes are optimized to represent well raw BOLD timeseries, over a with range of experimental conditions.

   - All atlases are available in .nii.gz format and sampled to MNI space
   

2. Additionally, we provide meaningful names for these modes, based on their anatomical location, to facilitate reporting of results.

   - Anatomical names are available for each resolution in .csv
   

Simple demo

Run the demo online

Datasets and Statistical model used to extract these atlases

For this, we leverage the wealth of openly-available functional images (Poldrack et al., 2013) and stochastic online matrix factorization algorithm (SOMF, Mensch et al., 2018), sparse dictionary learning.

Cite this work if you use these atlases

Dadi, K., Varoquaux, G., Machlouzarides-Shalit, A., Gorgolewski, KJ., Wassermann, D., Thirion, B., Mensch, A. Fine-grain atlases of functional modes for fMRI analysis. NeuroImage, Elsevier, 2020, pp.117126 Link to this paper

References

Mensch, A., Mairal, J., Thirion, B., Varoquaux, G., 2018. Stochastic Subsampling for Factorizing Huge Matrices. IEEE Transactions on Signal Processing 66, 113–128.

Poldrack, R.A., Barch, D.M., Mitchell, J.P., et al., 2013. Toward open sharing of task-based fMRI data: the OpenfMRI project. Frontiers in neuroinformatics 7.