MRI Brain – Multiple Sclerosis Dedicated Child Protocol

1. Executive Summary

Multiple sclerosis MRI occupies a unique position in clinical neuroradiology: it is simultaneously the most standardised and the most rapidly evolving area of brain MRI practice. No other neurological disease has generated a more extensively validated, internationally agreed-upon MRI protocol — the MAGNIMS-CMSC-NAIMS consensus [1, 2] — and yet the protocol has undergone major updates twice in four years (2021, 2024), driven by the formal incorporation of susceptibility-based biomarkers (central vein sign, paramagnetic rim lesions) and the revision of the McDonald diagnostic criteria to include the optic nerve as a fifth anatomical DIS location [3, 4].

The fundamental challenge that makes a dedicated MS protocol necessary — and distinct from the generic brain MRI — is the following: MS lesions are small (minimum 3 mm), distributed across specific anatomical locations that have defined diagnostic weight, and must be described in a structured way that directly informs whether the criteria for dissemination in space (DIS) and dissemination in time (DIT) are met. A generic brain MRI report that counts "multiple periventricular white matter lesions" is diagnostically inadequate. A correct MS MRI report must classify each lesion by its topographic category — periventricular, cortical/juxtacortical, infratentorial, spinal cord, or optic nerve — and declare which DIS and DIT criteria are fulfilled [3, 4].

1.1 Added Value Over the Generic Protocol

Protocol additions over the generic brain MRI:

• 3D FLAIR replaces 2D axial FLAIR — enabling coronal and sagittal reformats critical for callososeptal lesion detection and infratentorial lesion characterisation
• Sagittal FLAIR is mandatory (not optional as in the generic protocol) — corpus callosum and callososeptal interface are primary periventricular sites
• FLAIR* (combined FLAIR and T2* — also called EV or Echo Vector) or SWI with veins — enables central vein sign (CVS) detection for diagnostic specificity
• Post-contrast T1 — retained with specific MS-protocol timing (10 minutes post-injection for maximising enhancing lesion detection)
• Subcallosal plane orientation — axial slices must be parallel to the subcallosal line, not the generic body axial — essential for consistent periventricular lesion localisation
• Black holes (chronic T1-hypointense lesions) documentation — requires non-enhanced T1 specifically assessed for this purpose

Protocol removals or deferrals vs generic brain MRI:

• DWI: remains in protocol for the first MS MRI (to exclude alternative diagnoses) but is not required for routine monitoring scans
• SWI: upgraded from optional to the primary susceptibility sequence for CVS and PRL assessment

1.2 Limits of the Dedicated Protocol

Despite the robust MAGNIMS-CMSC-NAIMS consensus, several clinically important MS questions remain inadequately addressed by the standard brain protocol:

• Cortical lesion detection: cortical (intracortical and leukocortical) lesions — which carry strong diagnostic and prognostic weight — are underdetected on 3T FLAIR and T2 sequences (sensitivity approximately 20–30% compared to 7T MRI). DIR improves detection but is still insufficient compared to 7T research imaging [1].
• Spinal cord: brain protocol alone does not assess spinal cord, which is required for full DIS assessment and for evaluating clinical symptoms referrable to the cord.
• CVS and PRL at 1.5T: the central vein sign and paramagnetic rim lesions are substantially harder to visualise at 1.5T; dedicated 3T susceptibility sequences or FLAIR* provide the best performance [5, 6].
• Brain atrophy and grey matter involvement: global and regional brain atrophy is a strong predictor of long-term disability, but no validated automated volumetry tool has been approved for routine clinical use across all vendor platforms.

2. Clinical Context and Pre-Test Information

2.1 Clinical Presentation Relevant to MRI

The MS protocol is requested across several distinct clinical scenarios that require different protocol emphasis:

Clinically isolated syndrome (CIS): first demyelinating episode — optic neuritis, transverse myelitis, brainstem syndrome, or hemispheric event. The MRI question is whether the criteria for DIS are met at this first event (with or without DIT fulfilment) [3, 4]. The protocol must be comprehensive — including susceptibility sequences for CVS and PRL assessment — because meeting DIS criteria at the first event permits an MS diagnosis under the 2024 McDonald criteria.

Established MS, baseline scan or treatment monitoring: the primary MRI question is lesion activity — new or enlarging T2/FLAIR lesions and new enhancing lesions compared to the most recent prior scan. The scan must be directly comparable to prior studies in terms of protocol and geometry. A new enhancing lesion fulfils DIT under McDonald criteria; 3 or more new T2/FLAIR lesions indicates high radiological activity that may prompt treatment escalation.

MRI-negative suspected MS: the patient has clinical signs consistent with MS but one or more prior standard brain MRIs are negative. The 3T dedicated protocol with 3D FLAIR, DIR for cortical lesion detection, and CVS/PRL assessment is indicated — potentially converting an apparent MRI-negative case to MRI-positive.

Progressive MS (PPMS or SPMS): the clinical question shifts toward quantifying lesion burden, brain atrophy, and infratentorial/spinal cord involvement, which are the anatomical substrates of disability accumulation independent of relapse activity.

2.2 Pre-Test Information the Radiologist and Technologist Must Know

Prior MS MRI: the most recent prior brain MRI is the single most important pre-test datum. Every MS scan must be compared scan-to-scan for new or enlarging lesions. If prior MRI is not available in the reading system, the neurologist should be asked to provide it. Do not interpret an MS monitoring scan without comparison.

Current disease-modifying therapy (DMT): the DMT determines which lesion patterns to prioritise and which conditions require special attention:

• Natalizumab, fingolimod, dimethyl fumarate, alemtuzumab: increased PML risk — DWI must be included in all monitoring scans for patients on these agents
• All DMT monitoring: note the specific agent in the report as it frames the expected lesion activity level

JC virus antibody status and index: for natalizumab-treated patients, JC antibody index > 1.5 defines high PML risk. Dedicated PML surveillance protocol (FLAIR + DWI) should be specified when risk is high.

Timing of contrast injection relative to prior MRI: new enhancing lesions indicate active inflammation and qualify as DIT evidence. The temporal distance from the prior scan determines the clinical significance of new enhancing lesions.

Duration of symptoms: rapid lesion accumulation within weeks of symptom onset suggests a highly inflammatory phase; very stable imaging over years suggests either effective treatment or less aggressive disease.

Clinical phenotype: relapsing-remitting vs. progressive MS — in progressive disease, spinal cord atrophy and grey matter involvement predominate over new white matter lesions.

Patient age and vascular comorbidities: non-specific white matter hyperintensities from vascular disease (particularly in patients > 50 years, hypertension, diabetes, migraine) are the dominant differential diagnosis. The 2024 MAGNIMS-CMSC-NAIMS guidelines specifically note that stricter diagnostic thresholds should be applied in patients > 50 years with vascular risk factors [4].

2.3 Differential Diagnosis Landscape

The protocol must be designed to discriminate MS from its principal mimickers, because many of these produce similar T2/FLAIR white matter lesion patterns. The key discriminators — topographic distribution, CVS, PRL, enhancement pattern, clinical context — are specifically addressed by the dedicated MS protocol.

Primary mimickers requiring specific protocol consideration:

• Migraine-related white matter lesions: non-specific periventricular, subcortical, and deep white matter lesions; almost never at the callososeptal interface; absent CVS and PRL — CVS/PRL assessment is the decisive discriminator
• Small vessel cerebrovascular disease: periventricular and subcortical lesions, often with lacunes; basal ganglia and thalamic involvement atypical for MS; absent CVS; lesion morphology different (less ovoid, less perpendicular to ventricles)
• NMOSD (Neuromyelitis Optica Spectrum Disorder): area postrema lesions, longitudinally extensive spinal cord lesions; absence of typical MS callososeptal lesions; different enhancement pattern; aquaporin-4 antibodies
• MOGAD: often tumefactive or extensive cortical/leptomeningeal involvement; different lesion morphology; anti-MOG antibodies
• Acute disseminated encephalomyelitis (ADEM): typically monophasic; large, bilateral, poorly defined lesions; thalamic involvement; post-infectious context; age usually younger
• Vasculitis (CNS or systemic): variable distribution; leptomeningeal enhancement; clinical context

3. Indications, Appropriateness and Imaging Pathway

3.1 When the Dedicated Protocol Is Indicated

The dedicated MS brain MRI protocol is indicated for [1, 2, 3, 4]:

• All patients with CIS or first demyelinating event in whom an MS diagnosis is being considered
• All patients with established MS for treatment monitoring (active disease follow-up)
• All patients with suspected MS and a prior MRI that was non-dedicated or of insufficient quality (1.5T without specific MS sequences, no sagittal FLAIR, no subcallosal plane)
• All natalizumab-treated patients for PML surveillance (modified salvage version with FLAIR + DWI)

The 2021 and 2024 MAGNIMS-CMSC-NAIMS recommendations endorse the dedicated protocol for diagnosis and monitoring of MS as the standard of care [1, 2, 4].

3.2 When the Generic Master Protocol Is Sufficient

The generic brain MRI is sufficient when:

• The clinical question is acute neurological deficit with no specific MS suspicion (stroke, encephalitis, mass lesion)
• MS is not on the differential diagnosis
• Routine "check MRI" in a patient with no MS diagnosis or clinical suspicion

3.3 When Further Sub-Specialised Protocols Are Required

• Spinal cord involvement: add dedicated spinal cord protocol (cervical or thoracic) when: (a) clinical symptoms are referrable to spinal cord; (b) brain DIS criteria are not met and spinal cord lesion would fulfil DIS; (c) progressive MS with suspected cord atrophy
• Optic nerve MRI: orbital T2-FS and post-contrast orbital T1-FS when optic neuritis is suspected and as the fifth anatomical DIS location under 2024 McDonald criteria (only when orbital MRI quality meets standards for optic neuritis diagnosis — see MAGNIMS 2024 criteria [4])
• PML suspicion: dedicated PML surveillance protocol (FLAIR + DWI without contrast); do not apply standard enhancing lesion protocol
• fMRI/DTI for clinical trials: specialist research settings only; not for routine clinical protocol

3.4 Red Flags Modifying Urgency or Protocol

4. Dedicated Protocol Design

4.1 Protocol Delta vs the Master Protocol

4.2 Mandatory Dedicated Sequences

4.3 Conditional and Advanced Sequences

4.4 Rationale per Disease-Specific Sequence

3D Isotropic FLAIR — the Structural Backbone

3D FLAIR replaces 2D axial FLAIR for a specific reason that goes beyond mere resolution: the callososeptal interface — the posterior undersurface of the corpus callosum adjacent to the septum pellucidum — is the most MS-specific periventricular location, producing the classic "Dawson's fingers" pattern that is the hallmark of periventricular demyelination. This pattern is essentially invisible on axial FLAIR and requires sagittal FLAIR reformats. A 2D axial FLAIR cannot produce a diagnostic sagittal reformat of acceptable quality. The 3D isotropic acquisition provides both planes from a single acquisition.

The 2021 MAGNIMS-CMSC-NAIMS guidelines [1] specify that a single 3D FLAIR acquisition is the preferred approach when the scanner supports it, and that both axial and sagittal FLAIR must be evaluated for MS lesion topography assignment.

For MS-specific interpretation, the FLAIR must be scrutinised in three planes sequentially: (i) axial oblique (subcallosal plane) for periventricular lesions at the lateral walls and the callososeptal interface; (ii) sagittal for Dawson's fingers and corpus callosum undersurface lesions; (iii) coronal for infratentorial lesions (brainstem, middle cerebellar peduncles, floor of the fourth ventricle) that are often better seen in coronal than axial orientations.

Pitfall: 3D FLAIR is motion-sensitive. Motion artefacts produce banding across the white matter that can simulate lesions, or can obscure true periventricular lesions through k-space-related blurring. Placing 3D FLAIR early in the protocol reduces this risk.

SWI / FLAIR* for Central Vein Sign and Paramagnetic Rim Lesions

The central vein sign (CVS) is a small medullary vein coursing through the centre of an MS lesion, visible as a hypointense linear structure on T2*-weighted or susceptibility-weighted imaging within a hyperintense lesion on FLAIR. Its presence reflects the perivenular pattern of MS demyelination — one of the pathological hallmarks of MS. Systematic meta-analytic data confirm a CVS threshold of ≥ 3 CVS-positive lesions achieves pooled sensitivity 0.87 and specificity 0.75 for distinguishing MS from non-MS white matter lesion etiologies [5].

The paramagnetic rim lesion (PRL) is a T2* or QSM-visible paramagnetic rim at the outer edge of a chronic active lesion, produced by iron-laden activated microglia/macrophages at the lesion rim. PRLs have high specificity for MS (essentially absent in migraine, stroke, and vascular disease) but low sensitivity — typically only a minority of lesions in a given patient are rim-positive [6]. Under the 2024 McDonald criteria [3], both CVS and PRLs are now formally recognised as high-specificity biomarkers that can substitute for the waiting period required to demonstrate DIT under specific algorithmic scenarios.

FLAIR vs. SWI*: FLAIR* — a sequence combining FLAIR and T2* contrast simultaneously via multi-echo or interleaved acquisition — provides superior CVS detection compared to standard SWI in most published comparisons, with FLAIR*-based CVS detection achieving pooled sensitivity 0.89 and specificity 0.79 [5]. Standard SWI with phase images is an acceptable alternative when FLAIR* is not available.

Technologist note: SWI phase images must be retained and reviewed by the reporting radiologist alongside the magnitude images. CVS assessment is best performed on the phase image (where the central vein appears as a paramagnetic hypointensity within the lesion). If only magnitude images are saved, CVS assessment is substantially degraded.

Post-Contrast T1 — Timing is Disease-Specific

For MS lesion enhancement detection, the 10-minute delay between contrast injection and T1 acquisition is the disease-specific optimisation. Standard oncological protocols often image within 5 minutes of injection; in MS, active demyelinating plaques have a disrupted but not maximally permeable blood-brain barrier. The 10-minute window allows sufficient gadolinium accumulation in the plaque interstitium to produce detectable enhancement. Studies comparing 5-minute versus 10-minute post-contrast T1 in MS demonstrate higher lesion detection rates at 10 minutes, with the difference most significant for small (< 5 mm) lesions [1, 7].

A single standard dose (0.1 mmol/kg) of macrocyclic GBCA is the current standard. Triple-dose or high-relaxivity GBCA strategies are no longer recommended due to gadolinium deposition concerns.

Post-contrast T1 in monitoring scans: for routine monitoring of stable patients on disease-modifying therapy, the 2021 MAGNIMS-CMSC-NAIMS guidelines [1] state that contrast can be omitted from routine monitoring scans when T2/FLAIR alone demonstrates new lesions. However, contrast remains standard for: initial scans, first monitoring scan on new therapy, any scan with clinical relapse, and when the clinical question specifically includes active lesion documentation for treatment escalation decisions.

DIR — Cortical Lesion Detection

DIR (double inversion recovery) suppresses both white matter and CSF signals, leaving grey matter cortex visible against a dark background. In MS specifically, DIR improves the detection of cortical lesions — both intracortical (confined entirely to the cortex) and leukocortical (spanning the cortex and subcortical white matter) — by a factor of 2–4 compared to T2 and FLAIR alone [1, 8]. Cortical lesions count toward the DIS "cortical/juxtacortical" category under both 2017 and 2024 McDonald criteria and carry strong prognostic weight (associated with cognitive impairment, disability progression, and grey matter atrophy).

The MAGNIMS guidelines include DIR as an optional sequence, with the acknowledgment that it substantially increases cortical lesion detection [1]. At tertiary MS centres with dedicated neuroradiology expertise, DIR should be considered standard.

4.5 Dedicated Planes, FOV, Resolution and Coverage

Subcallosal line plane orientation: the axial sequences must be angled along the subcallosal line — a reference line drawn from the anteroinferior margin of the corpus callosum to the inferoposterior margin of the corpus callosum on the midsagittal image. This alignment is the MAGNIMS-CMSC standard [1, 2] because it: (i) cuts the periventricular white matter perpendicular to the ventricular walls, maximising lesion-to-background contrast; (ii) creates reproducible axial sections across examinations for lesion comparison.

This is a key departure from the generic brain MRI protocol, which uses the standard body axial plane (or the ACPC line). At most departments, this requires manual prescription from the sagittal scout on every MS examination — it cannot be automated without scout review.

3D FLAIR coverage: whole brain, 1 mm isotropic. Sagittal reformat from the 3D FLAIR is the primary plane for Dawson's finger and corpus callosum lesion identification.

Slice thickness: all T2 and FLAIR sequences ≤ 3 mm with no gap, per MAGNIMS-CMSC specifications [1, 2]. The minimum lesion size for McDonald DIS count is 3 mm. A 5 mm slice thickness with 1–2 mm gap would miss the majority of small lesions.

SWI coverage: whole brain axial at 1.5–2 mm slice thickness. Phase images must be preserved. The Minimal MS-relevant CVS detection requires slice thickness ≤ 2 mm to resolve the central vein. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page Spin Echo DWI / Non-EPI DWI Sequence.

4.6 Contrast Strategy Specific to MS

Gadolinium administration is indicated for:

• Initial diagnostic scan (CIS or first MS evaluation): mandatory for active lesion (DIT) documentation
• First monitoring scan on a new DMT: assesses baseline activity on the new treatment
• Any scan accompanying a clinical relapse
• Scans where new activity is suspected and clinical decisions (treatment escalation) depend on confirming or excluding active lesions

Gadolinium is not required for:

• Routine monitoring scans in clinically stable patients on established effective DMT, when T2/FLAIR change alone answers the clinical question [1]
• PML surveillance scans (FLAIR + DWI only)

Macrocyclic agents are mandatory: given the evidence for gadolinium deposition in the CNS with repeat administrations, the 2021 MAGNIMS-CMSC-NAIMS guidelines explicitly recommend macrocyclic GBCAs (gadobutrol, gadoteridol, gadoterate meglumine) for all MS MRI with contrast [1].

Timing: 10 minutes post-injection before T1 acquisition. The post-contrast T1 is the last sequence in the protocol (after 3D FLAIR, T1, SWI, DWI — see Section 8.2).

4.7 Sequence Matching, Reproducibility and Follow-Up

For disease monitoring, the MAGNIMS-CMSC 2021 guidelines [1] state: "Brain MRI should be performed in an identical manner at each follow-up according to the above-mentioned protocol." This is the cardinal rule — identical field strength, identical protocol, identical scanner whenever possible.

For reliable lesion change detection:

• Same field strength (1.5T and 3T produce different lesion counts for the same scan — mixing invalidates comparison)
• Same scanner (different scanner models produce different lesion conspicuity even at the same field strength)
• Same protocol (3D FLAIR vs 2D FLAIR are not directly comparable for lesion counts)
• Subcallosal plane reproducibility — document the angulation for each examination

The baseline scan: the post-treatment baseline scan (performed 3–6 months after starting DMT) is the primary reference for all subsequent monitoring. New lesions on subsequent scans are counted relative to this baseline, not relative to the pre-treatment diagnostic scan.

5. MRI Semiotics — Disease-Specific Imaging Findings

5.1 Direct Signs

T2/FLAIR Lesions — MS-Specific Semiotic Features

MS lesions are characterised by their topographic distribution — not their T2 signal alone, which is non-specific. The five McDonald lesion locations (2024 criteria [3]) are:

Periventricular lesions: abutting the lateral ventricular surface, with the callososeptal interface (posterior undersurface of corpus callosum) as the most MS-specific periventricular site. The classic "Dawson's fingers" pattern — perpendicular T2/FLAIR hyperintense projections from the callososeptal interface into the white matter, following the orientation of medullary veins — is virtually pathognomonic of MS. Visible only on sagittal FLAIR or sagittal reformats of 3D FLAIR.

Cortical/juxtacortical lesions: lesions within the cortex (intracortical), abutting the cortex (leukocortical), or confined to the subcortical U-fibres (juxtacortical) without intervening normal white matter between the lesion and the cortex. On FLAIR, cortical lesions are often isointense or only mildly hyperintense due to the CSF partial volume effect at the cortical surface — the primary reason DIR was developed.

Infratentorial lesions: involving the brainstem, cerebellar peduncles (especially middle cerebellar peduncle), and cerebellum. Typical sites: floor of the fourth ventricle, dorsal pons, medullary pyramids. Infratentorial lesions in MS are typically small (< 10 mm), ovoid, and do not produce mass effect.

Spinal cord: beyond the scope of the brain protocol — requires dedicated spinal cord sequences.

Optic nerve: requires dedicated orbital MRI — now the fifth DIS location under 2024 McDonald criteria [3].

Lesion Morphology Characteristics

• Ovoid shape: the characteristic ovoid or ellipsoid morphology of MS lesions, with the long axis perpendicular to the ventricular surface (in periventricular lesions) or cortical surface (in juxtacortical lesions), reflects the perivenular distribution of demyelination
• Minimum size for McDonald counting: 3 mm longest dimension on T2/FLAIR
• Homogeneous T2/FLAIR signal: unlike tumour or ischaemia, most MS lesions show relatively homogeneous signal

Acute/Active Lesions vs. Chronic Lesions

Active lesions (enhancement on post-contrast T1):

• Ring or nodular enhancement, typically closed ring or open ring pattern
• Enhancement persists for approximately 4–8 weeks on average (range 2–16 weeks)
• Active lesion may appear as new T2/FLAIR hyperintensity before enhancement is visible

Chronic lesions:

• T2/FLAIR hyperintense, non-enhancing
• May evolve to T1 hypointensity ("black holes") — persistent T1 hypointensity reflecting irreversible axonal loss and tissue destruction
• The T1-hypointense black hole — visible on non-enhanced MPRAGE or IR-GRE — is a marker of severe tissue damage and correlates with clinical disability better than T2 lesion volume alone

Paramagnetic rim lesions (PRLs):

• T2* or QSM-visible rim of paramagnetic signal at the outer lesion margin
• Indicates a "chronic active" lesion with an ongoing inflammatory rim despite lack of gadolinium enhancement
• Highly specific for MS [6]; absent in vascular disease, migraine

Central vein sign (CVS):

• Small central hypointensity on T2* or FLAIR* within the lesion, corresponding to a central medullary vein
• Present in MS lesions due to perivenular demyelination; absent or rare in vascular lesions

5.2 Indirect and Secondary Signs

Brain atrophy: progressive cerebral atrophy — particularly cortical grey matter and deep grey matter (thalamus, caudate) — is a secondary sign of MS disease severity. In the absence of automated volumetry, the radiologist should qualitatively note disproportionate atrophy for age, particularly of the corpus callosum (thinning of the body and splenium on sagittal T1) and the pericallosaul white matter.

"Dirty white matter": diffuse T2 signal change in the normal-appearing white matter (NAWM) — a subtle, poorly defined diffuse hyperintensity not meeting the threshold of a focal lesion — reflects widespread microstructural damage and is associated with disability in progressive MS. It is under-appreciated in routine reporting.

Thalamic atrophy: the thalamus is an early site of MS grey matter involvement, and bilateral thalamic volume reduction is a prognostic imaging finding in MS.

Corpus callosum involvement: involvement of the body and splenium of the corpus callosum on sagittal FLAIR — in addition to the callososeptal interface — is characteristic of MS and uncommon in vascular disease.

5.3 Severity, Extent and Activity Assessment

MS MRI severity assessment integrates:

• Total lesion volume/count: established as a clinical trial endpoint; routine qualitative estimate sufficient in clinical practice (low/moderate/high T2 lesion burden)
• Active lesion count: number of enhancing lesions on post-contrast T1 (current activity) and new T2/FLAIR lesions compared to prior scan (interval activity)
• Black hole burden: T1-hypointense lesions on non-enhanced T1; correlates with irreversible tissue damage
• Brain atrophy: qualitative global atrophy assessment; corpus callosum thinning
• Location of lesions: infratentorial and spinal cord lesions carry greater disability risk than equivalent-volume supratentorial lesions

5.4 Validated Classification and Grading Systems

McDonald Criteria (2024 revision) [3]

The McDonald criteria are the primary diagnostic classification system for MS. Under the 2024 revision, DIS requires lesions in ≥ 2 of 5 anatomical locations: periventricular, cortical/juxtacortical, infratentorial, spinal cord, optic nerve.

DIT requires: new T2/FLAIR or enhancing lesion on follow-up MRI, OR simultaneous presence of enhancing and non-enhancing lesions, OR positive CSF (OCB or kappa free light chains — new in 2024 criteria), OR ≥ 1 central vein sign-positive lesion combined with ≥ 1 paramagnetic rim lesion in specific algorithmic pathways.

The CVS and PRL are now formally incorporated as optional high-specificity biomarkers in the 2024 McDonald criteria — the radiologist reporting an MS MRI must be capable of assessing and reporting these features.

MAGNIMS-CMSC-NAIMS 2021/2024 Radiological Activity Criteria [1, 4]

Radiological activity is defined as:

• ≥ 1 new gadolinium-enhancing lesion, or
• ≥ 3 new T2/FLAIR lesions compared to the most recent prior scan

This definition is used in clinical trials for treatment response assessment and is applicable in clinical practice for treatment escalation decisions.

5.5 Differential Diagnosis on MRI

5.6 Mimickers, Pseudolesions and Normal Variants

FLAIR perivascular space enlargement: dilated Virchow-Robin spaces in the basal ganglia and deep white matter follow CSF signal on T1 (hypointense) and are FLAIR-attenuated (bright on T2, but relatively dark on FLAIR). These do not satisfy MS lesion criteria.

Truncation/Gibbs ringing artefact at the callososeptal interface: the high-contrast interface between white matter and CSF produces Gibbs ringing that can simulate small periventricular lesions on 3D FLAIR. These artefacts are typically thin, parallel to the ventricular wall, and reproduce with predictable geometry. They disappear or change character on lower-resolution acquisitions. The true callososeptal MS lesion is thicker, ovoid, and extends into the white matter.

Normal anatomical variants at the posterior fossa: the complex anatomy of the posterior fossa (choroid plexus, olivary nuclei, transverse sinus edges) can produce apparent FLAIR hyperintensities that mimic small infratentorial MS lesions. Correlation with T1 and the specific location eliminates most of these.

End-plate artefact on spinal cord FLAIR: not directly relevant to the brain protocol but relevant when brain and cord are reviewed together — the end-plate disc signal can simulate cord lesions on FLAIR.

6. Reporting Framework Specific to MS

6.1 Structured Reporting Template

Indication: state the clinical scenario (CIS/first demyelinating event / established MS baseline / monitoring on [specific DMT] / active relapse / PML surveillance) and the specific diagnostic question.

Technique: field strength, 3D vs 2D FLAIR, use of SWI/FLAIR* for CVS/PRL, DWI, DIR, gadolinium (agent, dose, timing post-injection), subcallosal plane orientation confirmed.

Comparison: date of prior scan, scanner, field strength, protocol. State explicitly if directly comparable or not comparable.

Findings — organised by McDonald topographic location:

Periventricular region:

• Number and description of periventricular lesions (callososeptal interface, ependymal surface)
• Dawson's fingers pattern: present / absent
• CVS-positive lesions: number (if assessed)
• New periventricular lesions vs prior (if comparison available)

Cortical / juxtacortical region:

• Number of cortical/juxtacortical lesions
• DIR findings if performed

Infratentorial region:

• Brainstem, cerebellar peduncles, cerebellum: number and location

Spinal cord (if included): state if excluded and why

Enhancement:

• Number of enhancing lesions: none / state number and location
• Enhancement morphology: ring / nodular / open ring
• New enhancing lesion vs prior

Black holes:

• T1-hypointense lesions: absent / present (number, location)
• Change vs prior

Paramagnetic rim lesions:

• Number and location (if SWI/FLAIR* performed)

Brain atrophy:

• Qualitative assessment; specific note on corpus callosum and thalamus

Impression — disease-specific:

• McDonald DIS criteria: met (specify which locations) / not met (specify what is lacking)
• McDonald DIT criteria: met (specify how) / not met
• If applicable: 2024 McDonald criteria — CVS/PRL contribution to DIS or DIT
• Radiological activity assessment: active (state number of new/enhancing lesions) / no new lesions vs prior
• Overall lesion burden: low / moderate / high
• Diagnostic conclusion: imaging consistent with MS / not consistent with MS / inconclusive

6.2 Mandatory Disease-Specific Reporting Checklist

• Periventricular lesions: counted and classified (callososeptal vs ependymal)
• Sagittal FLAIR reviewed for Dawson's fingers and callososeptal interface
• Cortical/juxtacortical lesions: assessed (DIR reviewed if performed)
• Infratentorial lesions: reviewed (brainstem, cerebellar peduncles, cerebellum)
• Spinal cord: if imaged — reviewed; if not imaged — documented as not assessed
• CVS assessment: stated (number of CVS-positive lesions or not assessable)
• PRL assessment: stated (present/absent or not assessable)
• Enhancing lesions: number, location, morphology
• Timing of post-contrast T1: stated (10-minute delay confirmed or not)
• Black holes: documented on T1
• New/enlarging lesions vs comparison: explicitly stated
• DIS criteria: explicitly stated as met/not met with locations
• DIT criteria: explicitly stated as met/not met with basis (enhancing, new T2, CSF)
• Technical limitations: field strength, protocol deviations, motion

6.3 Critical Findings and Communication

Direct communication to the referring neurologist is required when:

• PML features are identified (new large FLAIR lesion + DWI restriction in a natalizumab patient)
• Unexpected aggressive tumefactive lesion requiring urgent management decision
• New lesion or enhancement in a clinically stable patient that significantly changes the DIS/DIT calculation at initial diagnosis
• Major unexpected new lesion count (> 10 new T2 lesions) in a patient presumed stable on treatment

6.4 Common Reporting Errors

7. Technical Pitfalls and Disease-Specific Optimisation

7.1 Technical Pitfalls Specific to MS

3D FLAIR motion artefacts simulating MS lesions: Gibbs ringing and motion banding on 3D FLAIR produce artefactual signal changes in the periventricular and callososeptal white matter — the same regions where true MS lesions predominate. Ringing produces thin linear hyperintensities parallel to the ventricular wall, while motion produces diffuse banding. True periventricular MS lesions are ovoid, project perpendicularly from the ventricular surface, and persist across motion-corrupted and motion-free scans. Practical mitigation: acquire 3D FLAIR first in the protocol; verify lesion persistence on T2 (which is less motion-sensitive).

Subcallosal plane not applied: at departments not using a dedicated MS protocol, the axial FLAIR is often prescribed along the standard body axial plane or ACPC. This orientation cuts the callososeptal interface obliquely, reducing the apparent number of periventricular lesions by 20–35% compared to the subcallosal plane. Detect this error by checking the sagittal scout or reformat — the inferior margin of the genu and splenium should both be within the first axial slices in subcallosal plane.

1.5T CVS/PRL failure: at 1.5T, the central vein within lesions (typically 0.5–1 mm diameter) is below the spatial resolution of standard SWI acquisitions. CVS assessment at 1.5T requires dedicated thin-section high-resolution T2* sequences and remains less reliable than at 3T. If CVS or PRL assessment is required at 1.5T, the report should state that field strength limitations apply.

Post-contrast T1 at wrong timing: a post-contrast T1 acquired at 5 minutes instead of 10 minutes significantly reduces enhancing lesion detection. This error is common when the MS post-contrast T1 is acquired on the same schedule as a generic brain MRI. The 10-minute delay requires an explicit timing hold — typically using the DWI, SWI, or DIR acquisitions as temporal spacers after gadolinium injection.

7.2 Sequence-Specific Disease Pitfalls

FLAIR T2 shine-through of perivascular spaces: enlarged perivascular spaces in deep white matter and basal ganglia are FLAIR-suppressed (CSF-like signal) but incompletely suppressed when the CSF within them has short T1 due to protein or cellular content. These can appear as focal FLAIR hyperintensities that mimic MS lesions. Key discriminator: CSF-signal on T1 (hypointense) and T2 (bright, FLAIR-attenuated), without the perpendicular-to-ventricle orientation of MS lesions.

T1 black hole vs T1-hypointense BPH (benign tissue): the normal dentate nucleus and internal capsule can be mildly T1 hypointense. MS black holes are specifically T1-hypointense lesions that correspond to T2/FLAIR hyperintense lesions in MS-typical locations. Correlation between T1 and FLAIR is required.

SWI blooming artefact mimicking PRL: calcifications, microhaemorrhages, and venous structures all produce T2* signal loss. A PRL is distinguished by its: (i) rim morphology (thin rim at the lesion edge, not central); (ii) location within an established MS lesion on T2/FLAIR; (iii) presence on both phase and magnitude images. Calcifications are typically diffuse or central within lesions; microhaemorrhages are round, not ring-shaped.

7.3 When the Exam Is Non-Diagnostic for This Question

• No comparison scan at same field strength and protocol: new vs. old lesion distinction is not possible; the neurologist must be advised that the report provides a cross-sectional assessment only
• 1.5T with suboptimal 2D FLAIR only: CVS and PRL cannot be reliably assessed; cortical lesions may be systematically missed; repeat at 3T recommended for definitive MS workup
• Patient motion degrading 3D FLAIR: if the primary diagnostic sequence is non-diagnostic, the exam should be repeated or supplemented with 2D FLAIR in the affected planes

8. MRI Technologist Pearls Specific to MS

8.1 Disease-Specific Positioning and Coil Tricks

The most important MS-specific technologist step is prescribing the subcallosal plane: on the midsagittal scout (obtained from the 3D FLAIR or T1 localiser), identify the anteroinferior tip of the genu of the corpus callosum and the posteroinferior tip of the splenium. Draw the prescription line connecting these two points — this defines the subcallosal line. All axial slices should be parallel to this line. This takes 30–60 additional seconds and dramatically improves callososeptal lesion visibility.

8.2 Sequence Order Logic

Recommended sequence order for the dedicated MS brain protocol:

1. Three-plane localiser
2. 3D FLAIR ← first and most critical; patient freshest; most motion-sensitive
3. 3D T1 MPRAGE ← less motion-sensitive; pre-contrast mandatory for black hole identification
4. DWI ← before contrast; pre-contrast mandatory
5. Gadolinium injection (macrocyclic, 0.1 mmol/kg; power injector; record injection time)
6. SWI/FLAIR* ← during the 10-minute wait; fills the timing interval without wasting scanner time
7. DIR (if performed) ← second spacer for the 10-minute wait
8. Post-contrast T1 ← exactly 10 minutes after injection

The use of SWI and DIR as temporal spacers between gadolinium injection and post-contrast T1 is a highly efficient workflow strategy — these sequences provide additional diagnostic value (CVS, cortical lesions) while consuming the mandatory 10-minute waiting period.

8.3 Fast Salvage Version of the Dedicated Protocol

Three to four sequences in approximately 18–22 minutes. DWI deferred for PML-risk patients or initial scan; DIR deferred.

8.4 Disease-Specific Avoidable Errors

9. Quality Control Checklist for the Dedicated Protocol

• 3D FLAIR: sagittal, axial (subcallosal oblique), and coronal reformats all available
• Axial plane angle: subcallosal line verified on sagittal scout
• Slice thickness: ≤ 3 mm no gap for all T2/FLAIR sequences
• SWI: phase images saved and available for CVS/PRL assessment
• Gadolinium injection time documented
• Post-contrast T1: acquired ≥ 10 minutes after injection
• GBCA type: macrocyclic confirmed
• DWI: present for initial/CIS scans; or PML-risk monitoring scans
• DIR: performed or explicitly deferred (documented reason)
• Field strength documented: 3T (preferred) or 1.5T (document as limitation)
• Comparison scan: available in reading system and directly comparable
• Motion artefacts: assessed on 3D FLAIR; non-diagnostic if banding over periventricular white matter

11. Evidence Gaps and Ongoing Debate

CVS and PRL in clinical practice vs. research: the 2024 McDonald criteria formally incorporate CVS and PRLs as optional high-specificity biomarkers, but their routine clinical use requires susceptibility sequences (SWI, FLAIR*, QSM) at 3T, reader training, and standardised thresholds — infrastructure that is not universally available. The operational details of how many CVS-positive lesions (select3* vs select6*) should trigger a diagnostic impact are still being refined in real-world multicentre studies.

Cortical lesion detection gap at 3T: even with DIR, cortical lesion detection at 3T achieves only approximately 20–30% of the detection possible at 7T. DIR adds value over FLAIR alone but does not close this gap. Whether the additional clinical burden of DIR acquisition in routine practice is justified by the incremental lesion detection remains debated; the MAGNIMS guidelines classify it as optional rather than mandatory.

Contrast necessity in monitoring scans: the trend toward contrast-free monitoring (given gadolinium deposition evidence) is supported by the 2021 MAGNIMS-CMSC-NAIMS guidelines, which explicitly state that contrast is not required for routine monitoring in stable patients on effective therapy. However, clinicians continue to request contrast routinely, and the practical threshold for "stable on effective therapy" that safely permits contrast omission is not uniformly agreed upon.

AI-assisted lesion detection and change detection: deep learning models for automated MS lesion segmentation show high sensitivity for white matter lesion detection but variable performance for small juxtacortical and infratentorial lesions. AI-assisted lesion volume measurement and change detection tools are entering clinical use, but standardisation across vendors and validation for clinical decision-making are incomplete.

Brain volumetry as a clinical endpoint: brain atrophy and regional grey matter volumetry are validated as prognostic biomarkers in clinical trials but have not been adopted as standard clinical practice due to lack of normalised cross-vendor reference values and absence of clinical trial-grade software approval for individual patient decisions.

12. Evidence-Based References

A. Guidelines / Consensus / Society Recommendations

B. Systematic Reviews / Meta-analyses

C. Important Prospective / Original Studies

D. Technical MRI Papers

E. Landmark Historical References

End of document — MRI Brain MS Child Protocol — MRIninja v1.0 — May 2026 Parent page: MRI Brain Generic Standard Protocol Future child pages building on this page: spinal cord MS protocol; optic nerve MS protocol; PML surveillance protocol; progressive MS dedicated assessment; paediatric MS protocol.