nilearn

Open

[nilearn @nilearn]

I’ve been building Nilearn to make neuroimaging + ML both approachable and powerful. Since my first GitHub drops, it’s evolved fast.

Here’s a quick tour of what it can do and what I built into it

https://t.co/zMPuvsuVGz

[Nicholas Fabiano, MD @NTFabiano]

Repetitive negative thinking is associated with the accumulation of Alzheimer's dementia protein in the brain.

Stay positive.

[Doc Navarrow @DocNavarrow]

🤵‍♂️47-year-old man presents with generalized weakness and fatigue. No prior neurologic or psychiatric history. Currently receiving chemotherapy for hepatocellular carcinoma. Brain imaging was performed to exclude acute pathology.

🧠👀Any thoughts?

[nilearn @nilearn]

Curious how these connectomes change with age, disease, or behavior?

Once signals are extracted, you can feed them into ML models, statistical tests, or clustering. Let’s connect brain patterns to real outcomes

[nilearn @nilearn]

Pro tip: If you skip confound regression, the correlation matrix becomes overly dense — every region seems connected.
That’s why preprocessing matters.

Bonus: Nilearn lets you flexibly load confounds from fMRIPrep outputs using load confounds, following best practices.

[nilearn @nilearn]

Here’s the core idea in Python:

YES THATS IT! ask Claude :p

[nilearn @nilearn]

Step 4: Compute a correlation matrix between regions → this is your functional connectome.
High correlation = strong functional connection.

[nilearn @nilearn]

Step 3: Regress out confounds — movement, white matter, CSF, etc.
Removing noise is critical; without confound removal, every region looks connected.

[nilearn @nilearn]

Neuroimaging isn’t just pictures it’s data you can analyze.

With nilearn, you can extract meaningful time series from a brain parcellation and build functional connectivity maps in Python.

[nilearn @nilearn]

Step 2: Use a masker (like NiftiLabelsMasker) to extract region-wise signals from 4D fMRI.
This turns complex volumes into manageable time series

[nilearn @nilearn]

Step 1:
Load a predefined brain atlas and fMRI data.
A parcellation divides the brain into regions so you can summarize activity by region instead of raw voxels.

[Dylan O'Sullivan @DylanoA4]

The neuroscientist Jaak Panksepp wrote that loving social bonds are a primal form of addiction, and the brain signals the lack of these bonds via the feeling of depression

when you're depressed, you're actually in withdrawal from love

[Take Action! @takeactionD]

Consistency is the only thing that makes you realise that results have nothing to do with intelligence.

[nilearn @nilearn]

@hubermanlab Wow.

[Andrew D. Huberman, Ph.D. @hubermanlab]

Important paper, also for adults who “enjoy” stimulants. A new class of actual focus enhancing stimulants is emerging too— and they seem to enhance REM duration that night. Most are for narcolepsy and no, not Modafinil. 2026 is going to be a banner year for Rx stimulants.

[nilearn @nilearn]

@IamProHuman Depression is a drug.

[Lorwen Harris Nagle, PhD @LORWEN108]

After analyzing 500+ anxious clients, I found the cause behind most chronic anxiety loops.

It’s not trauma, stress, or overthinking.

It’s a SINGLE morning habit almost everyone does without realizing it... 🧵
(And, the psychological cost is staggering)

[nilearn @nilearn]

code is up to date 🧠👀

[Adam Grant @AdamMGrant]

Being in nature reduces fatigue and boosts attention.

34 experiments: After walking, playing, and learning in parks, forests, or water, kids & teens show better focus and self-control.

Not all outdoor time is equal. Green and blue spaces are more refreshing than urban settings.

[Ben Smith @bensmithlive]

A sleeping brain is still at work.

Nothing is more effective, and more ignored, than good sleep.