Want to blindly get robust estimates of neuronal-related activity on your resting-state, naturalistic or clinical fMRI data?

I am excited to share our latest preprint with you👇

Whole-brain multivariate hemodynamic deconvolution for multi-echo fMRI with stability selection

It has been a pleasure to collaborate with @fMRI_today, @javiergcas, Charles Zheng, and @CaballeroGaudes.

You can read the preprint in the following link: https://www.biorxiv.org/content/10.1101/2022.09.30.510299v2

Or you can keep reading to learn about the highlights👇

Context:

Deconvolution algorithms can blindly estimate neuronal-related activity without any information about the timings of the BOLD events.

This makes them especially useful to analyze resting-state, naturalistic, and clinical fMRI data, where timings are unavailable.

https://twitter.com/eurunuela/status/1420036314066341888

‼️ However, these methods have 3 main limitations ‼️

• they don't take advantage of the spatial information in the data
• their efficacy strongly depends on the selection of the regularization parameter
• they offer no measure of statistical certainty of the estimates

Our novel approach:

• a multivariate formulation that operates at the whole-brain level and adds spatial information via a mixed-norm regularization
• the use of stability selection to avoid the choice of the regularization parameter
• the AUC of the stability paths

What makes the AUC so interesting?

• depicts the probability of having a neuronal-related event at each voxel and time-point
• we make our estimation of activity based on these probability values
• robust estimates against the selection of the regularization parameters

The estimation of neuronal-related activity with this multivariate formulation and stability selection is

👍 MORE ACCURATE 👍

than that of multi-echo Paradigm Free Mapping (3dMEPFM) as shown by the dice coefficient, sensitivity, and specificity metrics.

The Python library will be soon available as part of the splora package.

In the meantime, you can use the univariate (voxel-wise) version that is already available in pySPFM 👇
https://github.com/eurunuela/pySPFM

Whole-brain multivariate hemodynamic deconvolution for multi-echo fMRI with stability selection

Graduate students and researchers, are you struggling to keep your academic literature organized and easily accessible?

The Academic Literature Review template for @tana_inc can help streamline your process and make research a breeze.

https://t.co/AaaawALGYa

With this template, you can easily import metadata from Zotero, tag papers with "read"/"unread" status, and highlight key quotes in Zotero.

Then, simply copy and paste your highlights into @tana_inc, where you can organize them by topic, author, and more.

But the real magic happens when you start writing your literature notes.

With @tana_inc's attributes, you can assign a topic, link the note to other notes, and see its origin.

And with the "quote" supertag, you'll never lose the context of your literature notes.

Best of all, your literature notes can be copied over to any manuscript you're writing.

And with the papers, permanent notes, literature notes, list of topics, and authors pinned to the sidebar, you can easily find your notes and stay on top of your literature review.

If you're ready to streamline your academic literature review process using @tana_inc, download the template at the link below.

And feel free to reach out with any feedback and questions you may have.

https://t.co/AaaawALGYa

Organizing Academic Literature Review with Tana Template

Scientific writing can be made easier with the right note-taking tools and workflows.

Unlock the power of @tana_inc and other #TfT tools like Notion or Obsidian to easily read, highlight, and synthesize academic literature.

First, you want to import the metadata from Zotero.

Once you add the gist linked below as an option in Zotero:

• ⌘ ⇧ C on a Zotero item to copy the metadata
• ⌘ V on a @tana_inc node to paste

https://github.com/eurunuela/zotero_tana_export

You can edit the metadata to follow a template of your liking.

Here's the one I use to keep track of topics, authors, journals, etc.

I even have a "read"/"unread" status to filter the papers I have in @tana_inc.

Reading and highlighting happen in Zotero.

Once done, I extract the highlights from Zotero and paste them into @tana_inc.

In @tana_inc, I tag the highlights with the "quote" supertag.

And here is where the magic happens.

I write literature notes of relevant highlights and their attributes get automatically filled up, so I never lose the context of my literature notes.

Once I'm done reading papers and writing literature notes, I write permanent notes.

These are going to be the pieces that I copy over to any manuscript that I'm writing.

With @tana_inc's attributes, I can assign a topic, link the note to other notes, and see its origin.

Finally, whenever it is time to write, all I have to do is go through my permanent notes on a given topic and bring them over to the manuscript.

I also have the papers, permanent notes, literature notes, list of topics, and authors pinned to the sidebar.

This way, I can find my notes and their context very easily and I can stay on top of my literature review.

I took one of @cortexfutura’s frameworks and adapted it to my needs.

If you think this workflow is useful to you, download my Academic Literature Review template for @tana_inc from the following link 👇

https://eneko.lemonsqueezy.com/checkout/buy/5818b754-fab8-4d38-83e1-8b07b62e45e9

PS: If you wonder what tool I use to write my tweets and be consistent, check https://typefully.com/?via=eneko

Academic Literature Review with Tana

Are you looking for an easy way to write your LaTeX manuscripts following journal templates?

MyST Markdown may be what you're looking for!

I have just attended a presentation given by @curvenote and I could not be more hyped for it.

Here are my main takeaways👇

There are already some widely-used templates available.

But if the template you need is not available, the community (or yourself) can add more templates!

I bet the list will grow quickly.

Dealing with cross-references and citations is super easy.

It supports BibTeX.

And they render perfectly on the web!

Yes, you can render your manuscript as a website that looks like a Jupyter Book.

And you can also very easily generate the PDF, the LaTeX, or the docx for your manuscript with just a simple command.

Not only can you go MyST → LaTeX or docx

You can also do LaTeX or docx → MyST.

Perfect if you want to build a website for your manuscript!

The gem of MyST may be Jupyter Notebooks though.

You can add a fully working and interactive Jupyter Notebook to your manuscript's website.

Perfect if you want to share your analysis or the code you used to generate the figures in your manuscript!

Check the MyST website for more info and how-to guides: https://myst.tools/

If you wonder what I used to write threads, I use @typefully

You can check it out here: https://typefully.com/?via=eneko

MyST Markdown: an easy and fast way to write manuscripts

Twitter is an excellent platform for researchers, especially Ph.D. students - if you know how to use it.

I've been quite active on Twitter for the last year and have learned a lot.

In this thread, I'll share what I've learned so you can too:

To get your research seen, tweet regularly.

It's tough to keep the momentum going, but seeing your activity map (like on GitHub) is a good motivator to stick with it.

To keep the momentum going, I recommend using @typefully

It's perfect to tweet without distractions, write threads, and grow on Twitter, and it's free!

They also have great analytics as part of their subscription plan.

https://typefully.com/?via=eneko

Write threads that describe your latest preprint or publication.

It's no surprise that the day impressions peaked for me was when I tweeted about our latest work.

Tag your collaborators and ask them to retweet and share your work.

Tweeting your research during conferences will make it more visible.

Use the conference hashtag and don't forget to tweet about discussions you've had with fellow researchers, tagging them to keep the conversation going!

You don't need to be an expert or influencer on Twitter.

Just be honest about how you use it and engage with others in your field, and your audience will grow!

The best part of being active on Twitter?

Your work gets noticed by experts in your field, which can lead to awesome career opportunities!

So don't hesitate—be active and get involved!

A year being active on Twitter

Do you want to read PDFs in Zotero on your tablet but don't want to pay for Zotero's cloud?

Here's a little trick 👇

1. Make sure your PDFs are on a cloud storage system of your liking but linked to Zotero.

https://twitter.com/eurunuela/status/1550526178057891840

2. Find the paper you want to read on the Zotero app and click on the "Add attachment" button.

Then find the PDF that is stored on your cloud storage.

You can now read, highlight, and takes notes in the Zotero app!

3. Once you're done reading:

• Make sure you delete the PDF on your cloud storage (you should only have the version on your tablet now)
• Right click on the paper → Manage Attachments → Rename and Move

Your highlighted PDF will be back on your cloud storage 💫

How to use Zotero on a tablet without paying for its cloud service

Brainhack Donostia 2022 was a huge success!

But we're not done yet @BHDonostia @brainhackorg.

🧵 Here's my personal opinion on what we did well, and what we need to improve:

First, congratulations to the organizing committee for hosting the best Brainhack Donostia yet.

For the first time, it felt like we had a proper, real Brainhack where we mostly focused on projects and everyone collaborated.

HUGE thanks to chairs @h_zjakic and @lmanso_ for leading & coordinating the committee.

Also, kudos to the "help desk" & "keep people happy" teams for their amazing work this year.

And last but not least, a big shoutout to the "talks" team for the incredible lineup of speakers!

In general, all the teams within the organization worked pretty well.

For next year, I would try to make sure of two things:

• Reduce the number of dropouts
• Have a predefined number of people needed in each group

Another thing we need to improve is our website: it is based on an unsupported template.

I've been making edits as we went to make it work, but I couldn't get it done this year.

Next year the website should use a new template.

I think we should collaborate with @biook_org next year to have a project for people who want to learn about neuroscience or would like to study neuroscience in the future.


It could be a very simple project in which people from outside the neuroscience field would learn what data looks like (e.g., EEG), how to program a few lines of code with Python, and how to visualize what they've done.

TL;DR: to create a new generation of brainhackers.

We had an idea a while back for a Brainhack Summer School, where we would teach neuroscience and programming to teenagers and undergraduates.

This new idea fits perfectly with that!

We were late getting local media coverage for our event this year.

I think @bcbl_ could help us get access to radio, papers, and TV next time if we start advertising earlier.

It would put us on the map and help us get exposure for better grants.

We need to be more active on Twitter all year round - not just a couple months before the event.

It's a lot of work to come up with tweet ideas, but if we want new attendees every year, people should see us consistently through different media.

One option is to start a newsletter that we send out 3 or 4 times a year to everyone that has ever participated in our Brainhack.

We could use MailChimp for easy editing and mailing, and people who want to unsubscribe can also do that.

Brainhack Donostia 2022 reflection

This week I learned about 3 tools that are going to change my programming life.

If you're not using these, you're missing out:

• Pre-commit
• Sourcery
• Mamba

Pre-commit is a package manager for code formatting hooks (among other things).

This means you can automatically style your code before every commit, to ensure all commits follow coding standards.

Check it out here: https://pre-commit.com/

Sourcery is a tool that feels like sorcery, to be honest.

This tool reviews your code everywhere you work and automatically suggests improvements 🤯

Check it out here: https://sourcery.ai/

If you use conda environments, check out Mamba.

It's a faster, more lightweight replacement written in C++.

Here's the link: https://mamba.readthedocs.io/en/latest/

PS: I use Typefully to write my tweets and stay consistent. Check it out: https://typefully.com/?via=eneko

3 tools that will make your programming life easier

When I started my Ph.D., I didn't know what tools would make my productivity peak.

Now, I have found the perfect set of tools for my work and I cannot imagine my daily routine without them.

I hope these tools are as useful to you as they are for me👇🧵

Raycast is where I search for everything on my computer, and what I use to open apps, start Zoom meetings, search GitHub issues and PRs, find Google Drive files, and manage my windows.

✨ Tip: create a "Quicklink" to search on Google Scholar without opening your browser.

Zotero is where I store all the papers I want to read or have in my library.

I have it set up so that my papers are stored on iCloud (can read on my iPad), with each author having their folder and each PDF named with title, year, and first author.

Here's how I did it 👇

https://twitter.com/eurunuela/status/1355909557877399552?s=20&t=Hf0-q45quamzAHp1DYYRUA

RoamResearch is where I spend most of my day taking notes and linking concepts.

I keep track of my todos, do interstitial journaling, start writing my manuscripts, take notes on papers and even find relevant papers using the amazing zoteroRoam extension by @AlixLahuec.

VSCode is where I do all my coding and LaTeX writing.

These extensions make it so much easier and enjoyable to code:

• GitHub Copilot
• GitLens
• autoDocstring
• Error Lens
• LaTeX Workshop
• NightOwl theme

GitHub is where I store and version control every piece of code or manuscript that I write.

I use GitHub every single day and I wish I had learned how to make the most of it earlier in my Ph.D.

Mastering GitHub really is a superpower ✨

Cron is the calendar app I have been using for the past few months and it is by far the best calendar out there in terms of the beauty of its design and the functionality it provides.

My favorite feature is sharing my availability with a link to Zoom.

Twitter is a source for the latest advances in the field and entertainment. That's why I use lists to choose the content I want to see.

I use @typefully to write my tweets and threads with no distractions.

And for a better search bar and some serendipity, I use @TwemexApp.

The browser is where we all spend most of the time every day.

I have been using Arc by @browsercompany for almost a month and it really is a game changer.

I'm loving the action bar, the different spaces, how I don't have to worry about open tabs, and the screenshots.

Most useful tools to work on my PhD thesis

Here are the talks and posters that I personally found very interesting on the last day (Thursday) of #ISMRM2022👇

Automatic detection of spatio-temporal patterns of interictal epileptic activity with fMRI by @tobias_cris with an oral session on her first ever ISMRM.

Really exciting results for the detection of epileptic foci using functional MRI only.

Static and dynamic connectivity are differentially affected by head motion: a healthy control and Parkinson’s disease study by @FrancescaSavio9.

We have work to do if we want aroma to be reliable!

Quality control and nuisance regression of fMRI, looking out where signal should not be found by @CelineProvins.

Data-driven heatmap clustering approach to analyzing structural connectomes by Wilburn Reddick.

Hybrid PCA denoising - improving PCA denoising in the presence of spatial correlations by @RafaelNetoHenr1.

Simulated fMRI responses using human Vascular Anatomical Network models with varying architecture and dynamics by Grant Hartung.

Brain in a virtual Petri dish: Flattening high resolution cortex patches in MRI and histology by Omer Faruk Gulban.

TensorFlow MRI: A Library for Modern Computational MRI on Heterogenous Systems by @JavierMontalt.