Optimization and Wrapper for Spatio-Temporal Clustering in fMRI

[eurunuela]

Optimization and Wrapper for Spatio-Temporal Clustering in fMRI

The Spatio Temporal Clustering script has been developed for fMRI data using AFNI. It is written in Python (and uses BASH) and it aims to remove spurious, scattered neuronal-related activity in favor of clusters. The clustering is applied in a t-1 < t < t+1 sliding window to maintain the activity in groups of voxels >= SIZE. The sliding window can now be given a different size.

• César Caballero, Basque Center on Cognition, Brain and Language
• Paul Taylor, NIH
• Eneko Uruñuela, Basque Center on Cognition, Brain and Language

---

Brief description of what was accomplished with this project

A wrapper was written for the original BASH code, giving more flexibility on how clusters are calculated. We also reduced the number of BASH commands to avoid I/O computation and achieve reasonable results in fewer lines of code. This tool will be very useful to remove scattered neuronal-related activity and retain big clusters of them.

---

Resources

You can check the code in this repo.

---