Stroke Mimics
- A "stroke mimic" is anything that walks and talks like a stroke but isn't one — same droopy face or weak arm, totally different cause.
- They matter because the clock-and-pressure of a code stroke tempts everyone to treat first and think later — and some mimics get worse with stroke treatment.
- Your imaging job is twofold: confirm there's no bleed, and look for the tell that says "this territory doesn't behave like a clot."
- Diffusion-weighted MRI is the great referee: most true strokes light up, and most mimics don't.
- The classics to keep on a sticky note: seizure, low blood sugar, migraine aura, and a faded old stroke that flares back up.
Imagine a fire alarm goes off in a building. Everyone evacuates, the trucks roll up, hoses out — and it turns out someone burned toast. The alarm wasn't lying; the smoke was real. But the cause was not a fire. Stroke mimics are burnt toast: the deficit is real, the panic is justified, but the plumbing in the brain is mostly fine. Sorting the toast from the fire, fast, under a ticking clock, is one of the genuinely hard jobs in neuro.
Why this is a trap and not just trivia
Ischemic stroke treatment is built for speed. The whole workflow is engineered to shave minutes, which is wonderful for real clots and slightly terrifying for everything else, because speed and "let's be sure" pull in opposite directions. Some stroke treatments thin the blood or bust clots, and if the real problem was, say, a brain that's actively seizing or a sugar that's bottomed out, you've now added a side of risk to a problem that didn't need it.
So the radiologist isn't just hunting for the stroke. You're the person in the room quietly asking, "are we sure this is a fire?"
The first job: rule out blood
Before anything clever, the non-contrast head CT exists to answer one blunt question — is there intracranial hemorrhage? A bleed can produce a stroke-like deficit and is the one thing you absolutely must not treat like a clot. This is the floor, not the ceiling, of the read.
A bleed mimics a stroke clinically but is the opposite emergency. Confirming "no blood" is the non-negotiable first move of any code stroke read — everything below assumes you've already cleared it.
The usual suspects
Most mimics fall into a short, learnable cast. Memorizing fourteen of anything is a fool's errand, so anchor on the common ones and the logic.
| Mimic | What gives it away |
|---|---|
| Seizure / post-ictal state (Todd paralysis) | Weakness that follows a seizure and tends to fade; symptoms may not respect a single artery's territory. |
| Hypoglycemia | Low blood sugar starves neurons; a fingerstick fixes both the patient and the puzzle. Cheapest test in the building. |
| Migraine with aura | Symptoms that march and spread over minutes, often with visual zigzags, in someone with a migraine history. |
| Old stroke "re-emerging" | A healed deficit flares with fever, infection, or fatigue. The CT shows a chronic scar, not a fresh insult. |
| Functional (non-organic) deficit | Real distress, but exam findings that don't map onto neuroanatomy. A diagnosis of exclusion, handled with care. |
| Tumor or infection | A mass or encephalitis can present acutely and masquerade as vascular. |
Notice the recurring theme: mimics often ignore the map. A clot lives in one artery's delivery zone. Seizures, migraines, and low sugar don't read the vascular atlas, so their deficits smear across territories in ways a single blocked vessel can't.
The sneakiest mimic is the old stroke that lights back up. A patient with a years-old deficit gets a urinary infection, and the weak side gets weaker. It looks brand new. The non-contrast CT saves you here: chronic infarcts are dark, shrunken, and sharply marginated — an old crater, not a fresh bruise. Always ask, "is this scar old?"
The referee: diffusion-weighted MRI
Here's the single most useful tool for this exact problem. Diffusion-weighted imaging (DWI), part of the advanced MRI toolkit, is exquisitely sensitive to the cellular swelling of acutely dying brain. In plain terms: when neurons are truly starved of blood, water inside them stops moving freely, and DWI screams about it as a bright spot.
Most genuine acute strokes glow on DWI. Most mimics — seizure, migraine, functional deficits, low sugar — do not produce that bright, territory-shaped restriction. That single contrast does enormous work: a clean DWI in someone who's clinically improving points hard away from a stroke.
The honest caveat: it's "most," not "all." A handful of mimics (a seizing brain, for one) can occasionally produce DWI changes, and very early or tiny strokes can hide. Imaging informs the clinical picture; it doesn't replace it.
How to actually tell them apart
Three questions carry most of the weight:
- Is there blood? Non-contrast CT first, always.
- Does the deficit respect a vascular territory? Clots honor the map; many mimics don't.
- Does the DWI light up where it should? Bright, territorial restriction favors a real stroke; a clean scan in an improving patient favors a mimic.
Before the scanner, the bedside team should have already checked a fingerstick glucose and a clear timeline. Hypoglycemia and a known seizure are mimics you can sometimes catch before imaging even starts — and the cheapest "scan" in stroke care is a drop of blood.
The whole art here is holding two truths at once: move fast enough to treat a real clot, and stay skeptical enough not to treat the toast. When in doubt, the picture and the patient should agree — and if they don't, that disagreement is the most useful finding on the study.