Demyelination protocol (MRI)

Last revised by Andrew Murphy on 23 Mar 2023

MRI protocol for demyelinating diseases is a group of MRI sequences put together to best approach these white matter disorders characterized by the destruction or damage of normally myelinated structures. These disorders may be inflammatory, infective, ischemic or toxic in origin. 

Historically the same protocol was used for the diagnosis and follow-up of multiple sclerosis. Over the past few years, however, the combination of much greater efficacy of disease-modifying drugs that have reduced the frequency of new lesions and increasing concerns regarding gadolinium deposition has resulted in reduced use of contrast in routine follow-up. 

An excellent resource for current MRI protocol recommendations is the Consortium of Multiple Sclerosis Centers who publish guidelines on their website, most recently updated in 2018 4

They recommend four different protocols depending on clinical context: 

  1. new diagnosis or clinically isolated syndrome
  2. routine followup
  3. progressive multifocal leukoencephalopathy (PML) surveillance 
  4. spinal cord imaging


In all instances, good quality imaging with ≤1 mm x 1 mm in-plane resolution and a high signal to noise ratio is required. Additionally, although the CMSC guidelines pragmatically remain open to 2D sequences, there is an increasing shift towards isotropic 3D sequences that enable higher resolution, multiplanar reconstructions and automated computer-aided diagnosis and morphometry 4,5

New diagnosis
  • T1
    • sequence: preferably ≤1.0 mm 3D isotropic acquisition
    • purpose: anatomical, best for assessing volume loss. This is particularly important for multiple sclerosis as brain volume loss is correlated with disease severity and disability  2,3
  • T2
    • sequence:  preferably ≤1.0 mm 3D isotropic acquisition
    • purpose: white matter signal abnormality. T2 is more sensitive than FLAIR for infratentorial lesions
    • if only 2D available then <3 mm slice thickness, no skip 4,5
    • sequence:  preferably ≤1.0 mm 3D isotropic acquisition
    • purpose: white matter signal abnormality. FLAIR is more sensitive than T2 in the detection of juxtacortical and periventricular plaques
    • if only 2D available then <3 mm slice thickness, no skip 4,5
    • sequence: routine 2D diffusion-weighted imaging 
    • purpose: active demyelinating MS plaques may demonstrate restricted diffusion
  • T1 C+ (Gd) 
    • for new diagnosis or clinically isolated syndrome, contrast is required
    • sequence:  ≤1.0 mm 3D isotropic acquisition or 2D sequence 4,5
    • purpose: active demyelinating lesions show enhancement, which is often incomplete around the periphery (open ring sign). In this setting, contrast is required to allow for dissemination in time to be established on the first scan and to assess for other alternative diagnoses. 
    • Gadolinium-based contrast agents (GBCAs) for CNS 1
    • all these GBCAs are approved by FDA at identical administered total doses of 0.1 mmol/kg body weight 1

Note: rather than the more common AC-PC axial plane preferred for most neuroimaging, the CMSC guidelines recommend the subcallosal plane 4

Optional additional sequences: 
  • SWI or T2*
  • double inversion recovery (DIR)
Routine follow-up

Routine follow-up imaging is exactly the same as new diagnosis or clinically isolated syndrome with the exception that contrast is not required in most instances. If, however, a patient is rapidly progressing, has unexpected symptoms or symptoms that are not thought to be typical of multiple sclerosis, then contrast should be administered 4

PML surveillance

In patients who are on natalizumab (TysabriTM), especially if they are positive for John Cunningham virus (JC virus) or who have been on the drug for more than 18 months, regular surveillance with 3 to 6 monthly MRI scans is recommended to assess for the development of progressive multifocal leukoencephalopathy (PML) is recommended 4. An abbreviated protocol consisting only of FLAIR and DWI suffices for these extra scans, with routine imaging every 12 months 4

Spinal cord imaging

Full spinal cord imaging takes additional time, and in many patients will require a second appointment, a pragmatic approach in clinically stable patients undergoing routine followup is to only image the cervical cord (which will typically cover down to T4 in any case). If there are symptoms of thoracic involvement, or concern regarding an alternative diagnosis is present then, of course, the whole cord can be imaged 4

There is no strong consensus on which sequences are best, and CMSC recommends performing any two of the following sagittal sequences:

  • T2
  • Proton Density
  • STIR
  • T1-PSIR

These should be <3 mm slice thickness with no skips 4

Axial T2 or T2* images should also be performed, at the very least through any lesions seen on sagittal 4. In practice, it can be difficult to be certain of the presence of lesions only on sagittal imaging and thus full coverage with axial images is prudent. 

These should be <5 mm slice thickness with no skips 4

ADVERTISEMENT: Supporters see fewer/no ads

Updating… Please wait.

 Unable to process the form. Check for errors and try again.

 Thank you for updating your details.