Improvements to multi-echo denoising in tedana

[tsalo]

Improvements to multi-echo denoising in tedana

• Katherine Bottenhorn, FIU (@62442katieb)
• Cesar Caballero Gaudes, BCBL (@CesarCaballeroGaudes)
• Elizabeth DuPre, McGill (@emdupre)
• Javier Gonzalez-Castillo, NIH (@javiergcas)
• Daniel Handwerker, NIH (@handwerkerd)
• Ross Markello, McGill (@rmarkello)
• Stefano Moia, BCBL (@smoia)
• Taylor Salo, FIU (@tsalo)
• Joshua Teves, MUSC (@jbteves)
• Eneko Uruñuela, BCBL (@eurunuela)

---

Brief description of what was accomplished with this project

We have (1) implemented GIFT PCA decomposition in Python, (2) improved how TE-dependence is evaluated, (3) restructured and enhanced automated testing, and (4) improved BIDS Derivatives compatibility of workflow outputs (which will also make the outputs more stable and easier for AFNI to ingest). Additionally, we have starting working on a large refactor of the package that will allow researchers to implement their own denoising pipelines within tedana. These changes will make it easier for researchers to utilize multi-echo fMRI data and to use the tedana package specifically.

---

Resources

• https://github.com/ME-ICA/tedana

---