Approach to Brain MRI
- A brain MRI isn't one picture — it's a stack of sequences, each lit differently so the same brain confesses different secrets.
- Your first job is to figure out which sequence you're looking at, because "bright" means nothing until you know the rules of that picture.
- The classic combo: T1 for anatomy, T2/FLAIR for "where's the trouble," diffusion for "is it acute," and post-gadolinium for "does it light up."
- Read it the same way every time, and always compare a finding across multiple sequences before you believe it.
A head CT is like one fast Polaroid of the brain. A brain MRI is more like a photo shoot — the same subject, shot under a dozen different lights, and your job is to walk through the contact sheet and notice who keeps showing up looking suspicious. If the head CT is the quick triage glance, MRI is the slow, gossipy, deeply nosy follow-up.
The thing that trips everyone up at first: on MRI, "bright" and "dark" don't have fixed meanings. The same fluid can be jet-black on one image and glowing white on the next. So before you interpret anything, you have to answer one question.
Step one: what sequence am I even looking at?
This is the whole game. The earlier physics pages cover why — see MRI basics: T1, T2 & weighting and the common sequences — but for reading, you mostly need to recognize each one by its "tell."
The fastest tell is the cerebrospinal fluid (CSF) sitting in the ventricles. Water is your reference object, the way a known coin gives scale in a photo.
| Sequence | CSF looks | Mostly used for |
|---|---|---|
| T1 | Dark | Anatomy; fat and blood products that go bright |
| T2 | Bright | Finding pathology — most "bad stuff" holds water and goes bright |
| FLAIR | Dark (water suppressed) | T2 problems next to the ventricles, where bright CSF would hide them |
| Diffusion (DWI) | Variable | Acute stroke and a few dense lesions |
Spot the ventricles first. Dark CSF with gray-white contrast = T1. Bright CSF = T2. Dark CSF but the brain still looks "T2-ish" = FLAIR.
Step two: T2 and FLAIR — the "where's the water" pictures
Most brain pathology — swelling, inflammation, tumor, old injury — does one rude thing: it holds onto extra water. On T2 that water shines bright like a wet spot on a gray carpet, so T2 is your wide net for catching trouble.
The problem is the ventricles are also bright water, so a bright lesion hugging the ventricle wall can vanish into the glare like a candle held up to the sun. That's what FLAIR fixes: it's a T2 image that mutes the free CSF, leaving abnormal water still glowing. This is exactly why periventricular lesions in conditions like multiple sclerosis are a FLAIR specialty.
Step three: diffusion — the "is this acute?" picture
Diffusion-weighted imaging (DWI) watches how freely water molecules can wiggle around. In healthy tissue they roam like kids at recess. In an acute ischemic stroke, the cells swell and trap that water — the recess bell rings, everyone's crammed indoors — and the trapped water lights up bright on DWI.
DWI alone can fool you, because some bright spots are just T2 brightness bleeding through ("T2 shine-through"). Always pair DWI with its partner map (the ADC map): true restricted diffusion is bright on DWI and dark on ADC. One without the other is a tease, not a stroke.
The deeper diffusion, perfusion, and spectroscopy tools get their own home in advanced MRI techniques; here, just learn to ask DWI the one question it answers beautifully: did something happen recently?
Step four: T1 and the gadolinium question
T1 is your anatomy map — crisp gray-white distinction, CSF dark. It's also where a couple of specific things go bright on their own: fat, and certain blood breakdown products. File that away so a bright T1 spot doesn't get mistaken for enhancement.
Then comes contrast. We give intravenous gadolinium and repeat the T1. Gadolinium can only light up tissue where the blood-brain barrier — the brain's bouncer that normally keeps stuff out of the tissue — has broken down. So enhancement means "the bouncer failed here": tumors, infection, active inflammation, and the like. The trick is comparing the before and after T1 side by side. Something that was already bright before contrast (fat, blood) isn't enhancing — it was just bright to begin with.
The classic rookie error is calling enhancement off the post-contrast image alone. Without the pre-contrast T1 for comparison, bright fat or subacute blood will happily masquerade as a lighting-up lesion. Always read the pair, not the single.
Putting it together: read it the same way every time
A reliable routine beats raw talent here. Mine, roughly:
- Confirm it's the right patient and the sequences/planes you expect are there.
- Identify each sequence by the CSF tell before interpreting it.
- Sweep T2/FLAIR for anything abnormally bright.
- Check DWI/ADC for true restricted diffusion (acute trouble).
- Compare pre- and post-gadolinium T1 for real enhancement.
- Localize each finding and characterize it across sequences — never trust a finding seen on only one.
That last rule is the heart of MRI. A spot is only believable when its story stays consistent as the lighting changes: bright on T2, behaving on FLAIR, restricting or not on DWI, enhancing or not on T1. One sequence is a rumor. Several agreeing sequences are a diagnosis.
Before any of this, the patient has to safely get into the magnet at all. MRI's giant magnet is always on, so screening for implants and metal is a genuine safety step, not a formality — see MRI safety & zones.
If you remember nothing else: name the sequence first, then let each light tell its part of the story, and only believe the parts that agree.