CNS Infection (Abscess, Encephalitis)
- "CNS infection" mostly splits into two imaging stories: a brain abscess (a walled-off pocket of pus) and encephalitis (the brain tissue itself inflamed).
- The single most useful trick for an abscess is diffusion-weighted imaging (DWI): pus is thick and traps water, so an abscess glows bright on DWI — most tumors and metastases don't.
- A classic abscess on contrast MRI is a smooth, thin, ring-enhancing lesion with surrounding swelling, often with the thinnest part of the ring pointing toward the ventricles.
- Encephalitis is subtler: look for swelling and bright T2/FLAIR signal in a typical territory — herpes loves the temporal lobes and insula, often on both sides but lopsided.
- These diagnoses can look exactly like stroke and tumor. The clinical story (fever, headache, confusion, immune status) is half the read.
Infection in the brain is the rare situation where the immune system's usual cleanup crew shows up to a fight inside a sealed, padded box that has no room to swell. Everything about how CNS infection looks on imaging comes back to that one cruel fact: the skull doesn't stretch. So whether it's a pocket of pus or inflamed tissue, the brain's response — swelling, pressure, and a desperate attempt to wall things off — is what we end up hunting for.
Let me split this into the two characters you'll meet most: the abscess and the encephalitis.
The brain abscess: a pimple with a fortress wall
Think of an abscess as the body building a fortress around an invader. First there's a messy, ill-defined infected zone (the radiologists call this cerebritis — basically "brain on fire, no wall yet"). Over days, the body lays down a wall of inflamed tissue around the pus, and now you have a mature abscess: a liquid center, a tough rim, and a moat of swelling around it.
On a contrast-enhanced scan, that wall lights up — this is ring enhancement. The center stays dark because pus doesn't take up contrast, and beyond the ring is a wide zone of edema (swelling) that looks dark on CT and bright on T2.
Here's the catch: a ring-enhancing lesion is one of the great mimics in neuroradiology. A high-grade glioma rings. A brain metastasis rings. So how do you tell pus from tumor?
DWI: the abscess's tell
This is where diffusion-weighted imaging earns its keep. Pus is thick, gloppy, protein-rich soup, and water molecules can't wander freely through it — they're stuck. DWI is exquisitely sensitive to that trapped water, so the center of an abscess lights up bright on DWI (with a correspondingly dark ADC map).
The necrotic center of a tumor, by contrast, is usually watery and lets molecules drift — so it stays dark on DWI. That bright-center-on-DWI sign is one of the most reliable ways to nudge "abscess" up your differential.
"Reliable" is not "perfect." Some tumors restrict, some abscesses behave oddly, and the immunocompromised brain breaks all the rules. DWI shifts your odds; it doesn't close the case. The clinical picture and follow-up still vote.
| Feature | Brain abscess | Necrotic tumor (glioma/met) |
|---|---|---|
| Enhancing rim | Thin, smooth, often regular | Thick, irregular, nodular |
| Thinnest wall | Tends to face the ventricle | No consistent pattern |
| Center on DWI | Bright (restricts) | Usually dark |
| Clinical story | Fever, rapid course, source of infection | Slower, often no fever |
Steroids are routinely given to shrink the swelling around brain lesions — and they quietly suppress enhancement. A treated abscess (or lymphoma) can look deceptively tame on the next scan even while it's still very much present. Always read the chart for steroids before you trust a faint-looking ring.
Encephalitis: when the tissue itself is inflamed
Encephalitis is the quieter, sneakier sibling. There's no tidy pocket to point at — the brain tissue is inflamed, so you're looking for swelling and bright signal on T2 and FLAIR sequences in a pattern that fits.
The one every trainee must know is herpes simplex encephalitis. It has a strong appetite for the temporal lobes, the insula, and the inferomedial frontal lobes — frequently on both sides but asymmetric. On MRI you'll see swelling and bright FLAIR signal in those regions, sometimes with a bit of restricted diffusion or a dash of hemorrhage.
Herpes encephalitis is a true emergency: untreated, it's devastating, and treatment (antiviral therapy) is started on clinical suspicion — you do not wait for perfect imaging. If the picture fits, the medicine goes in. Imaging supports the diagnosis; it should never delay treatment.
A useful mental anchor: if a confused, febrile patient has bright temporal lobes lighting up like a struck match, herpes is at the top of the list until proven otherwise.
Why CT alone often isn't enough
A head CT is usually the first study because it's fast, but early infection can look nearly normal on it — a little swelling, maybe a faint dark patch. MRI is the workhorse here because T2/FLAIR catches subtle edema and DWI sorts pus from tumor. If infection is on the table and the CT looks unconvincing, that's an argument for MRI, not reassurance. (If you want the bigger map of how these sequences fit together, see the approach to brain MRI.)
The one trap to carry with you
The biggest mistakes in CNS infection come from pattern-matching too fast. A ring-enhancing mass gets called a tumor; bright temporal lobes get mistaken for a posterior ischemic stroke. The fix is the same every time: pair the picture with the patient. Fever, headache, altered mental status, and immune status turn an ambiguous blob into a confident diagnosis.
Abscess = thin, smooth ring with a DWI-bright center; encephalitis = swelling and bright FLAIR in a typical territory (think herpes in the temporal lobes). When the imaging is ambiguous, the clinical story breaks the tie — and for suspected herpes, treatment starts before imaging is ever certain.