Location, shape and the company they keep decide it: periventricular ovoid lesions perpendicular to the ventricles, juxtacortical, infratentorial and spinal cord lesions point to demyelination; small punctate deep lesions in an older vascular patient point to small-vessel disease.
Orient first
- White-matter hyperintensity is a sign, not a disease. The differential is broad: small-vessel disease, demyelination (MS, NMOSD, MOGAD), migraine, vasculitis, infection, toxic-metabolic.
- MS lesions are ovoid, sit at the ventricle wall (Dawson fingers along the medullary veins), involve the corpus callosum and the U-fibres, and appear in the brainstem, cerebellum and cord.
- Small-vessel disease spares the callosal undersurface and the U-fibres early, is symmetric and deep, and comes with lacunes and microbleeds.
Acquire the study
- 3D FLAIR (sagittal best for the corpus callosum), axial T2, DWI, SWI, and post-contrast T1; spinal cord MRI when demyelination is suspected; the central vein sign on susceptibility-weighted FLAIR where available.
The manoeuvre
- Count and place the lesions in the MS locations: periventricular, cortical/juxtacortical, infratentorial, spinal cord (McDonald 2017 dissemination in space: at least 1 lesion in at least 2 of 4 — verify the current revision).
- Sagittal FLAIR: callososeptal interface lesions perpendicular to the ventricle (Dawson fingers).
- Shape and size: ovoid over about 3 mm versus punctate or confluent.
- Post-contrast T1: enhancing (active) and non-enhancing lesions together = dissemination in time.
- SWI: central vein sign (favours MS) versus microbleeds (favours small-vessel or amyloid).
- Company: lacunes, prominent perivascular spaces, age and vascular risk (small-vessel); longitudinally extensive cord lesions or area postrema lesions (NMOSD).
What confirms it
- Demyelination: lesions in the typical locations meeting dissemination in space (and time) with no better explanation. Small-vessel disease: symmetric deep and periventricular caps and bands with lacunes in an older patient.
What licenses you to exclude it
- A few non-specific punctate lesions do not establish MS; say so — "non-specific, most often small-vessel or migraine-related" is a legitimate conclusion.
The classic misread
- Over-calling MS on a handful of non-specific subcortical dots.
- Missing the corpus callosum lesions by not reading sagittal FLAIR.
- Missing NMOSD (long cord lesion, optic neuritis, area postrema) and labelling it MS — the treatments differ.