7. The seventh chapter critically examines the standardization of preprocessing in neuroimaging. osf.io/42bsu #Neuroimaging #fmriprep #BIDS #mriqc #reproducibility Oscar Esteban

DeepPrep can complete both s/fMRI preprocessing within 40 mins for a single scan and low to 8.5 mins/scan in a batch setup, over 10-fold faster than #fMRIPrep. Outputs of DeepPrep are similar or superior to those of fMRIPrep.

Recently, more and more #deeplearning models have been developed to accelerate individual steps of the preprocessing process. Yet, there's still lacking an end2end pipeline that integrates these modules in a manner similar to #fMRIPrep. Thus, the birth of DeepPrep is timely.

🚀Excited to share our #preprint on DeepPrep: a high-speed, scalable preprocessing pipeline for s/fMRI, empowered by SOTA #deeplearning algorithms. What takes #fMRIPrep 7 hrs, DeepPrep achieves in just 40 mins! Dive into the details🧵 Hesheng Liu Ted Satterthwaite shorturl.at/eBEP8

5. Tidy up "junk" files Energy is needed to store and backup data. you can reduce your footprint by deleting files you don't need. For #fMRIPrep, junk files make up to 96% of total size. Our tool fMRIPrepCleanup, automatically finds/removes this data: github.com/NickESouter/fM…

3. Preprocess conservatively Reduce compute required for your research by only performing preprocessing steps that are necessary. In our recent study, we provide tips on how to minimise emissions from #fMRIPrep while still getting good quality data: osf.io/preprints/osf/…

Perfecting #fMRIPrep 🙌

Attending Russ Poldrack's live talk! Looking forward to gaining insights from the #fmri preprocessing pipeline #fMRIPrep, applying some ideas with our EEGManyPipelines project, especially exploring the "glass box" visualization! #OHBMBrainhack #workflows

We apply this reproducible benchmark to investigate the robustness of the conclusions across two OpenNeuro.org datasets and two LTS versions of #fMRIPrep. 5/8

We created a fully reproducible denoising benchmark featuring a range of denoising strategies and evaluation metrics for connectivity analyses based on the classic paper Ciric 2017, built on the #fMRIPrep and @nilear software packages. 4/8

We love #fmriprep, but the confound documentation is a bit long and difficult to navigate. It’s not a trivial job to get the right regressors implemented in the benchmarks done on non-fMRIPrep workflow. 2/8

Our work “A reproducible benchmark of resting-state #fMRI denoising strategies using #fMRIPrep and Nilearn” is now officially on the reproducible preprint service CONP / PCNO and bioRxiv Neuroscience 🎉 neurolibre.org/papers/10.5545… 1/8

Anyone collecting #multiecho #MRI data (including phase reversed images) and using #fMRIPrep? How are you setting up the AP/PA pairs in distortion correction? OHBM Trainees

Thomas Yeo We encourage users to copy #fMRIPrep's boilerplate verbatim to minimize errors in reporting. The boilerplate is released with public domain licensing terms to facilitate that.

#DeepMReye now has a wrapper for #BIDS data! pypi.org/project/bidsmr… This is great for example to decode gaze position in #fMRI datasets processed with #fMRIprep Thank you Remi Gau for this amazing contribution to our package! w/Markus Frey

Without Twitter, I never would have applied for 𝚔𝚢𝚕𝚎 𝚜 𝚋𝚞𝚛𝚐𝚎𝚛's reproducible neuroscience workshop or Neurohackademy... Since then, my PhD and now postdoc work involves using or helping others use #fMRIPREP on a weekly basis, & I recently worked on a manuscript with Oscar Esteban 🥲