And another thank you to the whole @bhdonostia.bsky.social organising team for inviting me to speak about the carbon footprint of neuroimaging computing!
(I think the subtitles shown on the screen here were from people all over the room, not me just mumbling incoherently)
A massive thank you to the team who actually did the hard work of putting this together:
MONIKA UTROSA
CLΓMENTINE LΓVY-FIDEL
FΓBIO LOBATO
GABRIEL PIΕ VEJC
Everything we developed is available open source on GitHub: github.com/NickESouter/...
So, a CALL TO ACTION:
Are you currently in an institution where the energy usage of various MRI sequences is already being recorded? Are you in a position to start doing this? Please get in touch! Weβd love to source some more representative input data and understand which factors matter most
A caveat, weβve taken energy consumption metrics from specifications provided by scanner manufacturers. The numbers generated as output seem to be a pretty big underestimation of the impact weβd expect from fMRI research scanning. The numbers are probably more relevant to clinical scanningβ¦
The tool takes input information including:
π’ Field strength
π’ Scanner model
π’ Country in which data was collected
π’ Year of data collection
π’ Sample size
From this, it generates an environmental impact statement which can be popped into a grant application or journal paper
This week I attended my first ever hackathon with @bhdonostia.bsky.social! We put together a functioning dashboard to estimate the carbon footprint of MRI data collection π§
You can access the neuro-impact-calculator here: nicksouter.shinyapps.io/neuro-impact...
(Not yet optimised for phones) π§΅
We just wrapped an excellent keynote from @nicksouter.bsky.social on reducing the carbon footprint of research computing. A powerful reminder that scientific progress and sustainability can β and should β go hand in hand! ππ‘ #BHDonostia2025 #sustainability #GreenScience
βΌοΈ If you're interested in getting involved with the SEA-SIG, we are accepting nominations for the roles of Chair, Secretary, and Treasurer until OCTOBER 5! βΌοΈ
β¨ π« π π± π
ποΈ Only ONE WEEK to register for the OHBM Virtual Satellite Meeting, hosted with the SEA-SIG @ohbmenvironment.bsky.social! There will be an educational course, 3 symposia, and a canβt-miss keynote from Dr. Michael Breakspear! You wonβt want to miss it!
Photo of Cambridge River and view over King's College with a group of people punting.
Photo of the Heart and Lungs Research Institute on Cambridge Biomedical Campus. The building is bronze-colour and in front of a pond.
π£ Interested in the environmental (un)sustainability of AI (and how to support positive change)? Or know someone who is? I have an open position for a postdoctoral researcher to join my group in Cambridge, UK π±
π¨Closing June 5th
www.jobs.cam.ac.uk/job/51255/
Topics below π
#AcademicSky #GreenSky
Save the date ! Co-organized with @inserm.bsky.social bLORIER, @nicksouter.bsky.social from @sussexfss.bsky.social will talk about Β« How to reduce the carbon footprint of your research computing? Β»
tinyurl.com/4hck9xbh
An event affiliated with @ohbmenvironment.bsky.social !
Som staggering statistics from @ourworldindata.org
Are you a neuroscientist interested in sustainability, but don't know who to follow? π€
You're in luck! Here's a starter pack of neuroscientists who post about climate, and researchers who work in the intersection of neuro/psych and climate change!
π±Msg if you want to be added!
go.bsky.app/E1iW12T
Weekends are a great time to let ideas percolate, so why not give @nicksouter.bsky.social's paper on "10 recommendations for reducing the carbon footprint of neuroimaging research" a read and let it sink in while you head outside and get some fresh air?
direct.mit.edu/imag/article...
What is the relationship between neuroscience and climate change? At first these two topics can seem very distinct, but this article by Rae et al. makes a very compelling case for why we should incorporate environmental awareness into our research practices.
journals.sagepub.com/doi/10.1177/...
π¨We're recruiting! Interested in making science and computing more environmentally friendly? Now's the time!
Very excited to build a research group at Cambridge Uni to expand our Green Algorithms work thanks to the support of the Wellcome Trust
www.green-algorithms.org/join-us
#AcademicSky #SciSky
βΌοΈ NEW PAPER ALERT βΌοΈ
@nicksouter.bsky.social et al. (including many of our SEA-SIG members) demonstrate the environmental impact of different fMRIprep preprocessing choices, and make recommendations for minimizing carbon emissions of your pipeline!
onlinelibrary.wiley.com/doi/10.1002/...
Are you looking to get involved with like-minded neuroimagers to make our field more sustainable?
Nominate yourself to be the new Chair, Secretary, or Treasurer of the SEA-SIG in our 2024 elections! Nominations accepted from August 15-September 15!
ohbm-environment.org/upcoming-sig...
In this short 'PosterByte' video made for the British Neuroscience Association 2024 Members' Meeting, I discuss our recent work on the carbon footprint of neuroscience data processing π±π§ π₯οΈ
shorturl.at/aguWY
#BNAMembersMeeting2024
The Problem with Doomerism:
Critics are right that doom and gloom can demoralize people into inaction.
We found that negative messages increase social media sharing, but had no effect on policy support or climate beliefs & even backfired for effortful action: powerofus.substack.com/p/how-to-tal...
Big thanks to @loiclnlg.bsky.social, Gabby Samuel, Chris Racey, @lincoln81.bsky.social, Nikhil Bhagwat, Raghav Selvan, and Charlotte Rae for all their work on this paper!
10. Talk about greener computing
Discussing the carbon footprint of data processing within your neuroimaging community will make it easier to think seriously about this issue. This will help in embedding sustainable ideas into standard practice π±
9. Use existing data
Running analysis on previously collected and preprocessed data is a great way to test hypotheses without using additional energy to preprocess it yourself. We provide a table describing different open access datasets.
8. Reflect on what needs to be shared
It's good practice to share all raw neuroimaging data on platforms like OpenNeuro. But this storage still has a carbon footprint. Consider whether just sharing preprocessed files would be enough to allow others to make use of your data.
7. Push for publicly owned centralised data storage
Large centralised data centres tend to be more efficient, but this often means relying on big tech comapnies to host data. Alternatives like EuropHPC JU may allow storage that is efficient & sustainable:
eurohpc-ju.europa.eu/index_en
6. Plan your long-term storage
For files you do need to keep, have a data management plan in place. This could include deleting data after a certain time period (e.g., 10 years) or transitioning it to tape storage.
See spectrum.ieee.org/tape-storage...
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/...
4b. Time and location matter
You can also consider choosing cloud computing services based in countries/regions with lower average carbon intensity.
2022 data for countries and US states β¬οΈ
4a. Time and location matter
The carbon intensity of electricity peaks at busy times and drops overnight and at the weekend. By scheduling preprocessing/analysis to run at low intensity times, you can use the same amount of energy while emitting less carbon.
Data for the UK β¬οΈ
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/os...
