Congenital Brain Malformations
- Congenital brain malformations are blueprint problems, not break-it-later problems: the brain was built differently from the start.
- Group them by when the build went sideways — brain didn't split, neurons didn't migrate, the back of the brain didn't form, or the wiring (corpus callosum) never connected.
- MRI is the workhorse; you're reading brain architecture, not looking for a mass or a bleed.
- A few have classic shapes you can name on sight — the "molar tooth," the "bat wing" ventricles, the single front-of-brain ventricle.
- Many travel with hydrocephalus, seizures, or a neurocutaneous syndrome, so the malformation is rarely the whole story.
Most of neuroradiology is about something going wrong with a normal brain — a clot, a tumor, a bleed. Congenital malformations are the strange, fascinating opposite: the brain didn't get damaged, it got assembled wrong. Think of a house where the plumbing leaks (acquired disease) versus a house where the architect drew the staircase going into a wall (congenital). You're not hunting for an intruder. You're inspecting the original blueprint.
That reframing is the single most useful thing I can give you here, because the developing brain is built in a rough sequence, and these malformations sort neatly by which construction phase hiccupped.
The build order (and where it can break)
Picture the fetal brain assembling itself on a schedule, like a flat-pack bookshelf with very strict instructions. Roughly:
- The front of the brain divides into two hemispheres.
- Neurons are born deep near the ventricles and migrate outward to build the cortex.
- The back of the brain (cerebellum and brainstem) forms its own way.
- The hemispheres get wired together by big white-matter cables, the largest being the corpus callosum.
Skip or fumble any step and you get a recognizable family of malformations. Let me walk the phases.
When the brain doesn't split: holoprosencephaly
Early on, the forebrain is supposed to cleave down the middle into two hemispheres. In holoprosencephaly that split is incomplete, so instead of two tidy lateral ventricles you get one merged midline chamber, and the brain looks fused across the front.
The radiology tell is the missing midline. The structures that normally separate left from right — the front of the falx, the septum pellucidum — just aren't there, and a single horseshoe of brain wraps over a monoventricle. It comes in a spectrum from severe (alobar) to subtle, and because the face and forebrain are built from the same neighborhood, facial anomalies often tag along.
When neurons don't migrate: cortical malformations
If neurons are born but never finish their commute to the surface, the cortex is built out of order. The cleanest example is lissencephaly — literally "smooth brain." The folds (gyri and sulci) that should give the cortex its walnut texture never form, leaving an eerily smooth surface. Other migration problems leave clumps of gray matter stranded in the wrong place (heterotopia — gray matter where white matter should be), or a cortex that's too lumpy and disorganized.
A useful instinct: gray matter is gray matter no matter where it ends up. Stranded neurons (heterotopia) follow gray-matter signal on every MRI sequence. If a "lesion" tracks cortex on every single sequence and doesn't enhance or swell, you should be thinking misplaced brain, not tumor.
These malformations are heavily linked to epilepsy, which is often what brings the patient to imaging in the first place.
When the back of the brain misbuilds: posterior fossa malformations
The cerebellum and brainstem have their own construction crew, and a few classic results show up on board exams and in clinic.
| Malformation | What you see | Quick hook |
|---|---|---|
| Chiari I | Cerebellar tonsils sag below the foramen magnum | The tonsils are "peg-like," dipping out the bottom door of the skull |
| Dandy-Walker spectrum | Cystic, enlarged fourth ventricle; small/up-rotated cerebellar vermis | The back of the brain looks scooped out and filled with fluid |
| Joubert syndrome | "Molar tooth sign" on axial images | The midbrain looks like a tooth — thick straight cerebellar peduncles |
Low-lying cerebellar tonsils are a spectrum, and a few millimeters of "sag" can be a normal variant rather than a true Chiari I. The measurement matters, but so does the shape — peg-like, pointed tonsils are more convincing than rounded ones that merely sit a touch low. Don't overcall a mild droop in an asymptomatic patient.
The Dandy-Walker spectrum is a good reminder that these often come bundled: that big cystic fourth ventricle frequently obstructs CSF flow, so it travels with hydrocephalus.
When the wiring never connects: corpus callosum anomalies
The corpus callosum is the thick cable bundle joining the two hemispheres — the brain's main internet trunk line. In agenesis of the corpus callosum, that cable was never laid down (or only partly).
Because the crossing fibers have nowhere to go, the architecture rearranges in a way that's genuinely satisfying to recognize: the lateral ventricles get pulled into parallel, widely-spaced lines, and on a coronal view the frontal horns point up and outward like the wings of a bat — the classic "bat-wing" appearance. On a midline sagittal image, you simply don't see the normal arc of the callosum, and the gyri around the third ventricle splay out in a radiating, spoke-like pattern.
How to actually approach one
When a study looks structurally weird rather than damaged, slow down and run the blueprint checklist: Did the midline split (is there a falx, a septum pellucidum)? Is the cortex appropriately folded? Is the corpus callosum a complete arc on the sagittal? Does the posterior fossa look normally built? MRI is where this lives — if you want a refresher on which sequence shows what, detour through the approach to brain MRI, and if the midline anatomy feels shaky, neuroanatomy essentials is the place to firm it up.
Don't try to memorize every malformation as a flat list. Memorize the four build phases — split, migrate, form the back, wire it together — and let each malformation slot into the phase it broke. The shapes will start naming themselves.
The honest caveat: this is a deep, evolving field with its own dense classifications, and a single page can't make you a pediatric neuroradiologist. But the framework above gets you 80% of the way to recognizing that you're looking at a malformation, naming the famous ones, and knowing what question to ask next.