MRI and Multiple Sclerosis Diagnosis

The most important paraclinical tool for the diagnosis of multiple sclerosis (MS) is MRI, which is included in all iterations of the McDonald criteria, although the magnetic field strength is not specified. Conventional 1.5T MRI is most frequently used in the diagnostic process and has high sensitivity but lower specificity for the detection of white matter (WM) abnormalities. Beyond diagnosis, MRI is used to evaluate the response to disease-modifying therapies (DMT), as an adjunct in the decision to switch to another DMT and, at times, to support the selection of a particular treatment.

Despite technologic advances and updated McDonald criteria improving the accuracy of MS diagnosis, there are complex or atypical cases for which the diagnostic value of 1.5T MRI is limited, especially for nonspecific lesions, very few small lesions, or lesions restricted to the supratentorial area, particularly in the context of comorbid conditions. Differential diagnosis between MS and MS mimics remains challenging, and misdiagnosis is highly prevalent in MS (~20%), leading to treatment delay or unnecessary initiation of DMTs.1,2 Clinicians commonly encounter a clinicoradiologic paradox. For example, a person may be experiencing clinical decline despite having stable lesions on MRI, or in contrast, may have increasing disease activity on MRI without clinical worsening. There remains an unmet need for sensitive biomarkers to facilitate accurate diagnosis and guide treatment decisions. Development of high-field (≥3T) and ultrahigh-field MRI (≥7T) can potentially overcome limitations of the 1.5T MRI and further strengthen the role of MRI in the diagnosis and management of MS.

Pros and Cons of High-Field MRI

High- and ultrahigh-field MRI have been widely used in research and are becoming more available in the clinical setting. Although 3T MRI is approved for clinical use, 1.5T MRI remains most available in clinical practice. In 2017, the Food and Drug Administration (FDA) cleared the first 7T-MRI device for limited use in clinical practice, restricted mainly to head and extremity examination. The number of 7T MRI machines is increasing and even higher-field MRI scanners (9.4T, 10.5T or, 14T) are being developed for research use.

Conventional 1.5T MRI is limited because of a low signal-to-noise ratio (SNR) and modest spatial resolution. High-field MRI has the advantages of a higher SNR and a higher contrast-to-noise ratio (CNR) to provide enhanced spatial resolution up to 100 µm. This increase in spatial resolution improves visualization of small anatomic structures, separation of gray and white matter, and accurate detection of subtle abnormalities, including small lesions.3-7 In ultrahigh-field MRI, susceptibility effects, subtle field perturbations generated by local variations in magnetic properties, are boosted, especially in proximity to tissues with high iron content8 and densely distributed myelinating axons,9 allowing examination of microscopic veins, iron in the brain, microbleeds, and densely myelinated structures (eg, optic radiations).

The routine use of MRIs with field strength ≥7Tmay be limited by factors, some of which can be mitigated by sophisticated scan techniques, sequence selection, motion correction, and the application and optimization of multiple transmit coils that offer improved transmission efficiency.

Although a number of adverse reactions to 7T MRI (eg, vertigo in 5% of people who had 7T MRI, visual disturbances, or transient muscle contractions) have been reported, all of these events were transient without long-term sequelae.

Use of ≥7T MRI may be limited in individuals with stents, dental implants, pacemakers, or surgical clips. Typically, this group represents older patients who may be in higher need of imaging. For these individuals, lower-field MRI scan could be a better fit. Although there is no evidence that exposure to magnetic field at any strength can be harmful to humans, the effect of 7T MRI on DNA has not been studied on a large scale as yet.7 Another consideration is the high cost of installation and operation of ultrahigh-field MRI, although this is decreasing with technologic advances.

Central Vein Sign a Case for >1.5T Field Strength

Early histopathologic studies reported that most demyelinating lesions are centered on small parenchymal veins and this is confirmed by high-field MRI (3T and 7T) using T2-weighted sequences. The central vein sign (CVS), refers to a vein visualized inside a white matter lesion on T2 MRI sequences that appears as a hypointensity relative to the surrounding lesion. The vein appears as a dot or thin line that is located centrally, running partially or entirely through the lesion. The CVS has been observed in all clinical phenotypes of MS, including relapsing and progressive forms of the condition. The CVS has been proposed as an imaging biomarker of great diagnostic value for distinguishing between MS and MS mimics. Studies have examined the presence or absence of central vein sign in nonMS pathology, including inflammatory vasculopathies, neuromyelitis optica spectrum disorder (NMOSD), and small vessel disease (SVD) (Figure). The current understanding is that the presence of a CVS can accurately differentiate MS from similar nonMS pathology provided a minimal cut-off between 40% to 50% of lesions with the CVS is reached.10

<p>Figure. 3T Fluiid attenuated inversion recovery* (FLAIR*) MRI images in individuals who did and did not receive an MS diagnosis. Representative sagittal FLAIR* images of a woman who received a diagnosis of systemic lupus erythematosus (SLE), a woman who received a diagnosis of SPG4-spastic-paraparesis (SPG4 HSP), a woman who received a diagnosis of Sjögren disease (Sjögren), a man who received a diagnosis of relapsing-remitting MS (RRMS), a woman who received a diagnosis of primary progressive MS (PPMS;), and a man who received a diagnosis of RRMS. A central vein running through the lesion (arrows) is visible in the majority of MS lesions but is not typical in nonMS lesions. Reproduced with permission from Maggi P, Absinta M, Sati P, et al. The “central vein sign” in patients with diagnostic “red flags” for multiple sclerosis: a prospective multicenter 3T study. <i>Mult Scler</i>. 2019;Sep 19:1352458519876031.</p>

Click to view larger

Figure. 3T Fluiid attenuated inversion recovery* (FLAIR*) MRI images in individuals who did and did not receive an MS diagnosis. Representative sagittal FLAIR* images of a woman who received a diagnosis of systemic lupus erythematosus (SLE), a woman who received a diagnosis of SPG4-spastic-paraparesis (SPG4 HSP), a woman who received a diagnosis of Sjögren disease (Sjögren), a man who received a diagnosis of relapsing-remitting MS (RRMS), a woman who received a diagnosis of primary progressive MS (PPMS;), and a man who received a diagnosis of RRMS. A central vein running through the lesion (arrows) is visible in the majority of MS lesions but is not typical in nonMS lesions. Reproduced with permission from Maggi P, Absinta M, Sati P, et al. The “central vein sign” in patients with diagnostic “red flags” for multiple sclerosis: a prospective multicenter 3T study. Mult Scler. 2019;Sep 19:1352458519876031.

In a study comparing detection of a CVS with 3T vs 7T MRI in 7 people with MS with a total of 358 lesions and 7 healthy volunteers, a CVS was identified in 87% of visible lesions using T2 sequences on 7T MRI vs 45% of visible lesions using 3T MRI. In healthy volunteers the CVS was present in only 8% of discrete white matter lesions found.11 A prospective longitudinal cohort study assessed the diagnostic value of the CVS in people with clinically isolated syndrome (CIS). Using 7T MRI with a cutoff of 40% (positive CVS), the study followed 29 people with CIS. Eventually, 22 received a final diagnosis and 13 had MS. Those diagnosed with MS all had a CVS at baseline; none with a nonMS diagnosis did.12

In a prospective study of 51 people with suspected MS, initial 3T MRI evaluations had a median 86% perivenular for those who eventually received an MS diagnosis compared with 21% in those who eventually received a nonMS diagnosis. Having 40% of lesions be perivenular was associated with 97% accuracy and 96% positivity, with a 100% negative predictive value.13 Analysis of 4,447 lesions in 487 people with MS who had 3T MRI at 1 of multiple sites showed sensitivity of 68.1% and specificity of 82.9% for distinguishing MS from nonMS, using a 35% cutoff.14

In a meta-analysis of 501 people with MS, pooled incidence of the CVS on T2-weighted images was 74%. Subgroup analysis showed an 84% incidence in studies using 7T MRI, 79% on 3T, and 56% on 1.5T, although only the difference between 7T and 1.5T imaging was statistically significant. In those with a nonMS diagnosis, pooled incidence of the CVS was 33%, with 26% on 7T and 38% on 3T. Pooled incidence of the CVS was 33% in NMOSD and 36% in SVD. The optimal cut-off for the proportion of lesions with a CVS was 45%.15

Taken together these studies suggest that finding the CVS sign in >45 % of visualized white matter lesions favors MS diagnosis. Although further work in larger prospective cohorts is needed to evaluate the value of the CVS in MS, data to date support the value of higher field strength imaging.

Cortical Gray Matter Lesions

Cortical gray matter (GM) lesions are of particular interest, because they occur frequently in MS, may be present at the first clinical demyelinating event, and are incorporated in the demonstration of dissemination in space (DIS) in the 2017 revised McDonald criteria for the diagnosis of MS.16 Cortical GM lesions are associated with significant pathology, including cognitive impairment, epilepsy, depression, fatigue, disability, and disease progression. A specific form of MS, myelocortical MS, with a predominance of cortical and spinal cord lesions and without WM involvement has been described.17

Detection of cortical GM lesions with 1.5T MRI can be particularly difficult because of their small size and low degree of inflammation. As a result, most cortical lesions (~95%) are not detected.18 Application of double inversion recovery (DIR) techniques can increase detection of intracortical lesions by a factor of 1.5, increasing the detection rate to as high as 18%.19 On 1.5T MRI, visible cortical lesions are associated with a higher number of total cortical lesions, and as such just just a small portion of the total.20 Postmortem studies have shown a detection rate of 41% of mixed cortical-subcortical lesions and only 5% of intracortical lesions with 1.5T MRI. In a postmortem comparative study between 7T and 3 T MRI, 7T fluid-attenuated inversion recovery (FLAIR) sequences increased detection of cortical lesions by a factor of 2.25 compared with 3T FLAIR sequences. With T2 sequences, 7T MRI detected twice as many cortical lesions vs 3T MRI. On 7T MRI, a significant portion of cortical lesions (~40%) remained undetected. Most undetected lesions represented very small intracortical lesions and subpial cortical pathology, a type of cortical lesion highly specific for MS. It is unfortunate that ultrahigh-field MRI is still not able to detect all subpial lesions.21

With a multicontrast 7T-MRI protocol, higher GM lesion detection rates of 91%, 75% and 238% were seen using 3DT1w, 2D-T1w, and FLAIR sequences, respectively, compared with 3T.22 In a comparative study between 1.5T and 7T MRI, 42% of the study participants had additional lesions at 7T and 44% of the subcortical lesions imaged with 7T MRI had cortical involvement.23 In a preliminary study of 32 participants, it was found that a 3T MRI FLAIR sequence showed a higher cerebral lesion load with significant correlation with disability and cognitive domains compared with a 1.5T MRI FLAIR sequence, possibly because of improved detection of small lesions, particularly in the cortical and juxtacortical areas.24

Cortical GM lesions correlate significantly with accrual of disability. Lower field MRI underappreciates the degree of cortical involvement. The need for more comprehensive assessment of disease effects makes visualization and quantification of GM lesions on routine scan a priority.

Deep Gray Matter Lesions

Involvement of the deep GM and increased iron deposition in the basal ganglia can occur in MS. In a comparative, volumetric study of 14 people with MS, it was found that thalamic, caudate and pallidal volumes were higher at 3T vs 1.5T. In addition, 3T MRI was more sensitive in detecting thalamic and pallidal atrophy and had higher scan/rescan reliability than 1.5T.25 Using 4.7T MRI in 22 individuals with early MS, widespread abnormalities were found in the deep GM nuclei, particularly the pulvinar nucleus, relative to healthy controls.26 Ultrahigh-field MRI can detect deep GM lesions and changes of iron distribution in the brain, such as increased iron in the basal ganglia and reduced iron content in chronic inactive MS lesions.27 Recent evidence suggests that atrophy of deep GM is more prominent than cortical atrophy, can occur independently of WM atrophy, and progresses more rapidly than WM lesions or whole-brain tissue loss in MS patients.

White Matter Lesions

Before inclusion of cortical lesions in the 2017 revised McDonald criteria, MS diagnosis was mainly based on detection of WM lesions. Therefore, the question of a higher detection rate with ultrahigh-field MRI is of paramount importance.

The first comparative study of 7T vs 1.5T MRI showed only a modest (23%) increase in the detection of WM lesions with 7T MRI.23 In a comparative study between 1.5T and 4T MRI, a mean of 88 more lesions were seen at 4T, a difference that was statistically significant. All of these lesions were smaller than 5 mm and typically aligned along perivascular spaces.28 Significant differences in the lesion count between 7T and 3T MRI T2 sequences have been seen,11 as has a higher detection rate on 3T MRI for infratentorial and juxtacortical lesions with FLAIR and infratentorial, juxtacortical, and periventricular lesions with T2TSE sequences.26

Surprisingly, 3D-FLAIR 3T MRI detected significantly more lesions than 3D-FLAIR 7T MRI; otherwise no difference was found in other sequences. Nevertheless, certain sequences, such as 3T FLAIR and 7T magnetization-prepared rapid gradient echo imaging (MPRAGE), have been found more sensitive in detecting white matter lesions than 7T T2w or FLAIR.22

Additionally, 7T T1w sequence MRI was found highly sensitive for detection of black holes that are suggestive of irreversible WM lesion damage. Every T2w hyperintense lesion was visible as distinct hypointense plaque on 7T MRI T1WMPRAGE images, whereas only 68% to 78% of T2w lesions were delineated by 1.5T MRI.30

Spinal Cord Lesions

Spinal cord involvement occurs more frequently in the cervical area (59%) and less frequently in the lower thoracic area (20%). According to the revised McDonald criteria, the spinal cord is 1 of the 4 areas of the central nervous system (CNS) where demyelinating lesions are scored as confirmation of dissemination in space (DIS). Overall, studies have shown a weak correlation between spinal cord lesion volume and disability status; a stronger correlation has been found between spinal cord atrophy and Expanded Disability Status Scale (EDSS) scores.31

Very few studies have compared the detection of spinal cord lesions on 1.5T vs 3T MRI; all of them failed to show a significant improvement in lesion detection on 3T MRI vs 1.5T MRI. A preliminary study detected a similar lesion volume and number on both 1.5T and 3T MRI, although there was a nonsignificant trend towards a higher-volume load in participants with progressive MS at both field strengths. Moreover, the correlation between spinal cord lesion load and measures of disability was still weak and did not change with use of higher-field MRI, although it is possible that technical limitations might have interfered with the results in this early study.32 Another study showed no difference in the detection rate of spinal cord lesions and gadolinium-enhancing lesions between 3T and 1.5T.33 Owing to technical limitations, no 7T MRI spinal cord studies in people with MS have been performed yet.

Diagnosis of Multiple Sclerosis

Because early and accurate MS diagnosis is of major significance, interest is growing in the use of 3T and higher-field strength MRI to increase diagnostic yield, raising the question of whether high-field MRI leads to earlier MS diagnosis. In a multicenter trial of 66 individuals who presented with CIS, 3T MRI detected 15% more T2 brain lesions, although there was no difference in the detection rate for spinal cord or gadolinium-enhancing lesions. Scanning with 3T MRI did not lead to an earlier diagnosis of MS per the 2017 revised McDonald diagnostic criteria.33 At 6-month follow up, it was found that the MS lesions that were better visualized on 3T vs 1.5T MRI were located in the deep white matter, but the 3T MRI did not prove DIS or dissemination in time more than 1.5T MRI.34

Given the higher sensitivity of 7T MRI in detecting cortical lesions and the CVS, 7T MRI can play a key role in improving MS diagnostic accuracy. With 7T MRI, cortical lesions, undetectable on 1.5T or 3T MRI, can be seen; this can increase diagnostic yield, now that cortical lesions are included in the MS diagnostic criteria.


What is the future of high- and ultrahigh-field MRI? Is it just a matter of time to find widespread application and use in everyday clinical practice? At this point, the evidence supporting the use of 7T or even higher-field MRI in early diagnosis of MS is limited. Most of the studies involved a small number of participants, and only few of them were prospective. However, high-field MRI can play a major role in the differential diagnosis of MS, given its high sensitivity in lesion detection, particularly as cortical and deep GM lesions are concerned.

Larger clinical studies using 7T MRI for diagnosis and treatment decisions are needed before introducing ultrahigh-field MRI into everyday clinical practice. Even if the evidence for use of high-field MRI grows, installation and operation costs present an obstacle. Safety considerations should be also addressed and resolved. Although high-field MRI will be more widely used in research, and there are unparalleled possibilities for the future, 1.5T MRI is still the main diagnostic tool for the MS diagnosis and treatment decisions but its limitations raise the question whether MS patients should be evaluated on high- and ultrahigh-field MRI scanners.

1. Solomon AJ, Naismith RT, Cross AH. Misdiagnosis of multiple sclerosis: impact of the 2017 McDonald criteria on clinical practice. Neurology. 2018;92(1);26-33.

2. Kaisey M, Solomon AJ, Luu M, Giesser BS, Sicotte NL. Incidence of multiple sclerosis misdiagnosis in referrals to two academic centers. Mult Scler Relat Disord. 2019;30:51-56.

3. Trattnig S, Springer E, Bogner W, et al. Key clinical benefits of neuroimaging at 7T. Neuroimage. 2018;168:477-489.

4. Absinta M, Reich DS, Filippi M. Spring cleaning: time to rethink imaging research lines in MS? J Neurol. 2016;263(10):1893-1902.

5. Sati P. Diagnosis of multiple sclerosis through the lens of ultra-high-field MRI. J Magn Reson. 2018;291:101-109.

6. Inglese M, Fleysher L, Oesingmann N, Petracca M. Clinical applications of ultra-high field magnetic resonance imaging in multiple sclerosis. Expert Rev Neurother. 2018;18(3):221-230.

7. Sinnecker T, Kuchling J, Dusek P, et al. Ultrahigh field MRI in clinical neuroimmunology: a potential contribution to improved diagnostics and personalised disease management. EPMA J. 2015;6(1):16.

8. Haacke EM, Cheng NY, House MJ, et al. Imaging iron stores in the brain using magnetic resonance imaging. Magn Reson Imaging. 2005;23(1):1-25.

9. Lodygensky GA, Marques JP, Maddage R, et al. In- vivo assessment of myelination by phase imaging at high magnetic field. NeuroImage. 2012;59(3):1979-1987.

10. Maggi P, Absinta M, Grammatico M, et al. Central vein sign differentiates multiple sclerosis from central nervous system inflammatory vasculopathies. Ann Neurol. 2018;83(2):283-294.

11. Tallantyre EC, Morgan PS, Dixon JE, et al. A comparison of 3T and 7T in the detection of small parenchymal veins within MS lesions. Invest Radiol. 2009;44(9):491-494.

12. Mistry N, Dixon J, Tallantyre E, et al. Central veins in brain lesions visualized with high-field magnetic resonance imaging: a pathologically specific diagnostic biomarker for inflammatory demyelination in the brain. JAMA Neurol. 2013;70(5):623-628.

13. Maggi P, Absinta M, Sati P, et al. The “central vein sign” in patients with diagnostic “red flags” for multiple sclerosis: a prospective multicenter 3T study. Mult Scler. 2019;Sep 19:1352458519876031

14. Sinnecker T, Clarke MA, Meier D, et al. Evaluation of the central vein sign as a diagnostic imaging biomarker in multiple sclerosis. JAMA Neurol. 2019;76(12):1446-1456.

15. Suh CH, Kim SJ, Jung SC, Choi CG, Kim HS. The “central vein sign” on T2*-weighted Images as a diagnostic tool in multiple sclerosis: a systematic review and meta-analysis using Individual patient data. Sci Rep. 2019;9(1):18188.

16. Thompson AJ, Banwell BL, Barkhof F, et al. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. Lancet Neurol. 2018;17(2):162-173.

17. Trapp BD, Ontaneda D. Identifying a new subtype of multiple sclerosis. Neurodegener Dis Manag 2018;8(6):367-369.

18. Geurts JG, Bö L, Pouwels PJW, Castelijns JA, Polman CH, Barkhof F. Cortical lesions in multiple sclerosis: combined postmortem MR imaging and histopathology. AJNR AmJ Neuroradiol. 2005;26(3):572-577.

19. Tallantyre EC, Morgan PS, Dixon JE, et al. 3 Tesla and 7 Tesla MRI of multiple sclerosis cortical lesions. J Magn Reson Imaging. 2010;32(4):971-977.

20. Seewann A, Vrenken H, Kooi EJ, et al. Imaging the tip of the iceberg: visualization of cortical lesions in multiple sclerosis. Mult Scler. 2011;17(10):1202-1210.

21. Kilsdonk ID, Jonkman LE, Klaver R, et al. Increased cortical grey matter lesion detection in multiple sclerosis with 7 T MRI: a post-mortem verification study. Brain. 2016;139(Pt 5):1472-1481.

22. de Graaf WL., Kilsdonk ID, Lopez-Soriano A. et al. Clinical application of multi-contrast 7-T MR imaging in multiple sclerosis: increased lesion detection compared to 3 T confined to grey matter. Eur Radiol. 2013;23:528-540.

23. Kollia KS, Maderwald S, Putzki N, et al. First clinical study on ultra-high-field MR imaging in patients with multiple sclerosis: comparison of 1.5T and 7T. Am J Neuroradiol. 2009;30(4):699-702.

24. Stankiewicz JM, Glanz BI, Healy BC, et al. Brain MRI lesion load at 1.5T and 3T versus clinical status in multiple sclerosis. J Neuroimaging. 2011;21(2):e50-e56.

25. Chu R, Hurwitz S, Tauhid S, Bakshi R. Automated segmentation of cerebral deep gray matter from MRI scans: effect of field strength on sensitivity and reliability. BMC Neurology. 2017;17:172

26. Lebel RM, Eissa A, Seres P, Blevins G, and Wilman, AH. Quantitative high-field imaging of sub-cortical gray matter in multiple sclerosis. Mult Scler. 2012;18(4):433-441.

27. Hametner, S., Dal Bianco, A., Trattnig, S. and Lassmann, H. Iron related changes in MS lesions and their validity to characterize MS lesion types and dynamics with ultra-high field magnetic resonance imaging. Brain Pathology. 2018;28:743-749.

28. Keiper MD, Grossman RI, Hirsch JA, et al. MR identification of white matter abnormalities in multiple sclerosis: a comparison between 1.5T and 4 T. AJNR Am J Neuroradiol. 1998;19(8):1489-1493.

29. Wattjes MP, Lutterbey GG, Harzheim M. et al. Higher sensitivity in the detection of inflammatory brain lesions in patients with clinically isolated syndromes suggestive of multiple sclerosis using high field MRI: an intraindividual comparison of 1.5T with 3.0 T. Eur Radiol .2006;16(9):2067-2073.

30. Sinnecker T, Mittelstaedt P, Dörr J, et al. Multiple sclerosis lesions and irreversible brain tissue damage: a comparative ultrahigh-field strength magnetic resonance imaging study. Arch Neurol. 2012;69(6):739-745.

31. Bernitsas E, Bao F, Seraji-Bozorgzad etal. Spinal cord atrophy in multiple sclerosis and relationship with disability across clinical phenotypes. Mult Scler Relat Disord. 2015;4(1):47-51.

32. Stankiewicz JM, Neema M, Alsop DC, et al. Spinal cord lesions and clinical status in multiple sclerosis: a 1.5T and 3T MRI study. J Neurol Sci. 2009;279(1-2):99-105.

33. Hagens MHJ, Burggraaff J, Kilsdonk ID, et al. Three-Tesla MRI does not improve the diagnosis of multiple sclerosis: a multicenter study. Neurology. 2018;91(3):e249-e257.

34. Kataoka H, Kiriyama T, Taoka T, et al. Comparison of brain 3.0-T with 1.5-T MRI in patients with multiple sclerosis: a 6-month follow-up study. Clin Neurol Neurosurg. 2014;121:55-58.

EB reports no disclosures related to this article.