Chapter 3 Ultra-high field spinal cord MRI in multiple sclerosis: Where are we standing? A literature review.

Kreiter, D. J., van den Hurk, J., Wiggins, C. J., Hupperts, R. M., & Gerlach, O. H. (2022). Ultra-high field spinal cord MRI in multiple sclerosis: where are we standing? A literature review. Multiple Sclerosis and Related Disorders, 57, 103436.

Abstract

Magnetic resonance imaging (MRI) is a cornerstone in multiple sclerosis (MS) diagnostics and monitoring. Ultra-high field (UHF) MRI is being increasingly used and becoming more accessible. Due to the small diameter and mobility of the spinal cord, imaging this structure at ultra-high fields poses additional challenges compared to brain imaging. Here we review the potential benefits for the MS field by providing a literature overview of the use UHF spinal cord MRI in MS research and we elaborate on the challenges that are faced. Benefits include increased signal- and contrast-to-noise, enabling for higher spatial resolutions, which can improve MS lesion sensitivity in both the spinal white matter as well as grey matter. Additionally, these benefits can aid imaging of microstructural abnormalities in the spinal cord in MS using advanced MRI techniques like functional imaging, MR spectroscopy and diffusion-based techniques. Technical challenges include increased magnetic field inhomogeneities, distortions from physiological motion and optimalisation of sequences. Approaches including parallel imaging techniques, real time shimming and retrospective compensation of physiological motion are making it increasingly possible to unravel the potential of spinal cord UHF MRI in the context of MS research.

3.1 Introduction

Multiple sclerosis (MS) is an inflammatory demyelinating disease that can involve all parts of the central nervous system [156]. Magnetic resonance imaging (MRI) is central in establishing the diagnosis and monitoring disease activity. In the past years, there has been a vast amount of ultra-high field MRI research addressing different aspects of MS pathology, mostly in the brain. For the spinal cord, however, the body of literature is much more modest, owing to the technical challenges of imaging this small and mobile structure [169]. It is reported that approximately 80% of MS patients have spinal cord lesions [33, 170, 171], making improving imaging techniques for assessment of MS spinal cord pathology at least as important as for the brain.

Not only are spinal cord lesions common in MS, they are prognostic for conversion from a clinically isolated syndrome to clinically definite MS [170, 172, 173], for progression of relapsing-remitting MS (RRMS) to secondary progressive MS (SPMS) and also strongly associated with clinical disability on the long term [174]. Spinal cord lesions are the result of a variety of pathological processes like demyelination, axonal loss and gliosis [175]. Besides focal lesions, spinal cord atrophy can also be used as a biomarker for axonal loss and in addition demyelination, which both reduce cord volume. Atrophy measurements bring additional value, as these pathological processes occur not only in focal lesions but also in normal-appearing spinal cord tissue in MS patients. Spinal cord atrophy has been shown to be associated with disability progression in MS independent of brain atrophy and lesion volume [132, 148, 176, 177]. However, despite the usefulness on a group level, the limited sensitivity, reproducibility and the degree of measurement noise in quantifying these subtle changes in cord volume, still restrict the clinical usability [135, 169]. Achieving higher spatial resolutions is an important point of improvement to be able to better the study the pathophysiological processes of MS in vivo using MRI, given the small diameter of the spinal cord. Employing ultra-high field MRI could facilitate this and besides contributing to anatomical imaging, help improve imaging of microstructural abnormalities using advanced imaging techniques. For example, diffusion-based techniques can provide quantitative measures of neuroaxonal integrity, magnetization transfer imaging of myelin content, functional MRI (fMRI) of functional connectivity and MR spectroscopy can reveal metabolic changes in nervous tissue [88]. All these techniques can benefit from increased signal-to-noise ratio (SNR) and additionally some techniques will also benefit from other changes at higher static field strengths (like increased susceptibility effects for fMRI and greater chemical shift for MR spectroscopy [178]). Ultimately, better understanding of spinal cord pathology can in turn contribute to identifying therapeutic targets and measuring treatment response in MS patients with spinal lesions.

When 3T MRI scanners entered the clinic, this meant a theoretical doubling of the SNR compared to 1.5T and it quickly became the standard for both clinical and research setting. While for brain MRI scanning at 3T yielded higher MS lesion sensitivity [179, 180], comparative studies of spinal cord MRI at 1.5T and 3T never really showed obvious improvement [181]. Now with 7T scanners becoming increasingly available, improvements in lesion detection are expected with its further increase of SNR, contrast-to-noise ratio (CNR) and subsequent capability to achieve higher spatial resolution [88]. Presently, for cerebral imaging, scanning at 7T has already proven some of its capabilities for the MS field. Here we will review whether going for higher field strengths will possibly also be beneficial for spinal cord MRI in MS, and to provide an overview of where we are currently standing. For this purpose, we present and discuss: (i) the potential advantages of spinal UHF-MRI over conventional field strengths in MS, (ii) the technical challenges that are posed by moving towards higher field strengths for spinal cord MRI and (iii) an overview of current literature on spinal cord UHF research in MS.

3.2 Benefits and challenges

Scanning at ultra-high field strengths carries certain pros and cons. To start with the potential benefits, in general, increasing the static magnetic field (B0) causes a greater amount of proton spins to align with B0. This means an increase in potential signal and therefore a greater SNR and CNR. Having more signal allows for acquiring images at higher spatial resolutions. Submillimeter in-plane resolutions can be achieved, making it possible to acquire a high level of anatomical detail even of structures with a small cross-sectional area like the spinal cord (figure 3.1). In anatomical imaging, this can improve MS lesion sensitivity and delineation. The increase in CNR in general and more specifically the consequent increase in white/grey matter contrast, can contribute to differentiating and quantifying pathology between the white- and grey matter compartments in MS.

Figure 3.1 - Example of axial slice from a multi-echo T2* sequence at 7T with 0.18 mm2 in-plane resolution with high level of anatomical detail (nerve roots: blue, ligaments: purple, blood vessels: red, dura mater: green and pia mater: yellow). Reproduced from Massire et al. [182]
Figure 3.1 - Example of axial slice from a multi-echo T2* sequence at 7T with 0.18 mm2 in-plane resolution with high level of anatomical detail (nerve roots: blue, ligaments: purple, blood vessels: red, dura mater: green and pia mater: yellow). Reproduced from Massire et al. [182]


Additionally, higher field strengths result in increased susceptibility effects. The susceptibility of a tissue is its ability to become magnetized when placed in a magnetic field. The enhancement of differences in susceptibility between tissues at UHF benefits conspicuity of microvasculature and of (para)magnetic substances. With the arrival of UHF imaging of the brain, interest spiked in the central vein sign in MS lesions as a more specific MS lesion biomarker. The combination of higher spatial resolution and increased susceptibility effect gave way to achieve exquisite visualization of cerebral venules using T2*/SWI sequences. There is pathological evidence of the existence of central veins in spinal cord lesions, however, to date this has not yet been shown in vivo [183]. Also, paramagnetic rims, which are increasingly studied as a hallmark of chronic MS lesions, can be better visualized at UHF due to the greater susceptibility effects [37, 184, 185].

Not only can UHF MRI benefit anatomical imaging, but also a variety of advanced MRI techniques, which are important in studying functional and microstructural changes in pathology. In the case of functional imaging, scanning at UHF has two important advantages: (1) an increase in sensitivity for blood-oxygenation level dependent (BOLD) signal and (2) the high in-plane resolution that allows to more accurately distinguish spatial properties of the BOLD signal across the spinal cord [178].

Furthermore, as B0 becomes larger, chemical shift (differences in resonance frequency between different molecules) proportionally increases, a property which can be exploited using MR spectroscopy or chemical exchange saturation transfer imaging (CEST) to quantify certain tissue metabolites (examples for the nervous system include glutamate as marker of chronic neuroaxonal degeneration, N-acetyl-aspartate for neuronal mitochondrial metabolism and myo-inositol for glial cell activation) [88, 186].

The discussed benefits are not without drawbacks: There is an increase in image artifacts that arise as a result of inhomogeneities in the static magnetic field, radiofrequency transmit/receive fields and due to motion.

Magnetic field inhomogeneities

The use of stronger static B0 field results in an increase in magnetic field inhomogeneities. Large variation in susceptibility between regions, e.g. air-tissue interfaces, causes a major local gradient resulting in signal loss and distortions due to increased spin dephasing [177], affecting contrast and spatial localization. These distortions scale with the strength of the static field, making it an important challenge in UHF MRI.

To cope with the static field inhomogeneities one can apply external fields that offset local field distortions (shimming). At lower fields, these can usually be easily compensated with linear shims. Due to the geometry of the (human) sample, at higher fields there are more demands on the shim system, possibly also involving a need for shims with higher spatial order. Furthermore, very local field distortions, such as at the intervertebral junctions, can often not be completely compensated. To address these issues, the use of dynamic or real-time shimming is being increasingly investigated, i.e., updating the shim field on a slice-by-slice basis [187–189]. These techniques can also help address the issue of time-varying static field fluctuations mostly caused by respiration (see ‘Physiological noise’) and are showing promising results [31]. For a more in-depth review of developments in the field of shimming we refer to the reviews by Stockmann [190] and Winkler [187] and colleagues.

Additionally, the transmit radiofrequency B1 field also suffers from increased field inhomogeneities at higher field strengths. Here developments in the use of multiple independent transmission channels (parallel transmission) are forming promising solutions to improve B1 homogeneity, resulting in a more homogeneous signal across the scanned volume. On the receive side, some advantage can be brought by using parallel acceleration techniques [191] which can both reduce distortion and decrease scan time. This is, however, at cost of the SNR, which is a profitable trade-off at UHF given the increase in signal.

Physiological noise

Increasing spatial resolution causes images to become more prone to motion artifacts. Besides movement by the patient self, there are also physiological sources of motion that challenge imaging of the spinal cord, like respiration, cardiac action, swallowing and CSF flow [192]. Also, movement outside the imaging volume (e.g., due to respiration) can cause changes in the static magnetic field in the region of interest. At higher magnetic field strengths this effect amplifies, resulting in an increase of distortions.

Respiratory action is the largest source for physiological noise in spinal cord MRI. In addition to causing motion of the diaphragm and thorax, it cyclically changes the distribution of air and tissue in the thorax and abdomen, causing periodic variations in surrounding field distribution. This results in incorrect localization of signal, which can manifest as image distortion, presenting as motion, blurring or ghosting of the image [193, 194]. Figure 3.2 shows single echoes at different echo times (TE) and combined images before and after breathing correction, giving an example of distortions resulting from breathing-induced field fluctuations, which, as expected, become more prominent at longer TE’s.

Figure 3.2 - Left axial slices at C3 at increasing echo times are shown. Top row shows images before respiratory trace breathing correction and bottom row after correction. Right the root-sum-of-squares (RSS) echo combined images are shown. Left column shows the combined image before breathing correction and right after correction. Blue arrows indicate examples of distortions due to breathing-induced field inhomogeneities. Adapted from Vannesjo et al. [193]
Figure 3.2 - Left axial slices at C3 at increasing echo times are shown. Top row shows images before respiratory trace breathing correction and bottom row after correction. Right the root-sum-of-squares (RSS) echo combined images are shown. Left column shows the combined image before breathing correction and right after correction. Blue arrows indicate examples of distortions due to breathing-induced field inhomogeneities. Adapted from Vannesjo et al. [193]


Possible inconsistencies of the signal due to physiological motion can be (partly) mitigated by using gating techniques. These can be based on physiological monitors (e.g., electrocardiograms, respiratory bellows, pulse oximetry), optical tracking systems or using navigator echoes (a MR-based technique which produces 1D images of e.g., the air-diaphragm interface). Prospective gating, where intervals of minimal movement are isolated for signal acquisition, are rarely used anymore because it increases acquisition time by factor 2 to 3. Retrospective techniques use the collected motion data to correct the acquired signal afterwards. Lastly, real time shimming techniques can be applied to correct changes in magnetic susceptibility due to breathing [33], but requires additional specialized hardware and software.

Vannesjo and colleagues investigated the correction of breathing-induced field fluctuations using a respiratory trace for spinal cord MRI at 7T [193] (figure 2). They found that it can significantly reduce ghosting, improve quantitative metrics in multi-shot and functional T2*-weighted imaging of the spinal cord. This method brings the advantage that it does not rely on sequence modifications and does not require additional specialized hardware. The development of such techniques is necessary for making thoracic spinal cord MRI feasible and of sufficient quality for MS imaging studies.

Sequence design and power deposition

Unfortunately, sequences which work well on 1.5T and 3T are not readily translatable to higher field strengths. Some sequences are more difficult than others to make practicable at UHF. Spin-echo-based sequences for example are challenging to optimize for UHF. Although, they have the advantage of being less prone to static B0 field inhomogeneities, spin echo pulses have high radiofrequency (RF) power deposition (due to the 180° refocusing pulses). Because the specific absorption rate (SAR; measure for absorption of electromagnetic radiation) is proportional to the square of the RF frequency and this frequency scales with static B0 field. This means that moving from 3 to 7T increases the SAR more than fivefold. Therefore, running a spin echo sequence within safety limits leads to a longer acquisition time. This is why gradient echo (GRE) sequences, having a low RF deposition (due to lower flip angles), are easier to implement and are especially well suited for UHF.

Also, inversion recovery sequences, like STIR or PSIR, commonly used in spinal cord MRI in MS, do not directly translate to higher field strengths because of changes in relaxation times at different field strengths. Furthermore, the inversion pulses also pose difficulties in staying within SAR limits. To our knowledge, there are no publications on attempts to employ inversion recovery sequences for UHF spinal cord imaging to date. However, it is expected that solutions that have been developed for making FLAIR feasible in UHF brain MRI (like magnetization-preparation pre-pulses [51] and variable angle long length echo trains [50, 195]), can be adopted for other IR-sequences [196].

Contrast enhancement

Another important part of MRI in MS is the use of gadolinium-based contrast agents (GBCA) to detect recently developed lesions. Lesion contrast enhancement can aid the clinician to prove dissemination in time. At higher field strength, there is a reduction in r1 relaxivity (impact of contrast agent on T1 relaxation rate) of T1-shortening agents like gadolinium, while the r2 relaxivity (impact of contrast agent on T2 relaxation rate) increases [197], affecting the degree of contrast enhancement. Consequently, this theoretically requires an increase in dosage of the contrast agent and a longer sequence repetition time to achieve the same maximal contrast enhancement as on lower field strengths. On the other hand, T1 relaxation times of tissue are longer at higher field strengths and may potentially mitigate the lower r1 relaxivity due to background suppression being improved when using short repetition times [198]. To date, there are no studies published investigating usage of GBCA for spinal cord MRI at 7T and what T1-weighted techniques in that setting would be optimal to achieve maximum contrast enhancement. Therefore, this should be subject for future study. However, experience from UHF brain MRI in MS learns that in practice the effect of decreased r1 relaxivity is offset by the increased SNR and CNR achievable at 7T [195, 199].

Side effects

Also important for the feasibility of UHF MRI for research or clinical purposes is the tolerability of the investigation itself. The static magnetic fields of MRI scanners can cause transient physiological effects like dizziness, nausea or metallic taste [178]. The most common side effect is vertigo when the subject is being moved into the scanner [200]. Studies comparing lower field strengths to UHF report that these sensations do increase at higher field strengths, however, they generally remain mild and do not seem to significantly impact the acceptance of UHF MRI [178, 201]. Heilmaier et al. scanned 577 patients at 7T, in which in only 10 cases the scanning was not started or terminated due to side effects, showing that UHF MRI is generally well tolerated [200].

3.3 Literature overview

A literature search was performed through MEDLINE, Embase and the proceedings of the International Society for Magnetic Resonance Imaging for literature on in vivo UHF-MRI of the spinal cord in MS up until 19th of April 2021. The search strategy was (7T OR 9.4T OR (“ultra” AND “high” AND “field”) OR UHF) AND (spine OR spinal OR cord) AND (“Multiple Sclerosis”[Mesh] OR MS OR “multiple sclerosis”) AND (“Magnetic Resonance Imaging”[Mesh] OR “magnetic resonance” OR MRI). For conference abstracts we checked PubMed for corresponding journal publications. We identified in total four journal publications and four conference abstracts that satisfied inclusion criteria. All studies were performed at 7T, there were no studies at field strengths beyond 7T. An overview of current literature is shown in Table 3.1.

Table 3.1 - Overview of published journal articles and conference abstracts employing ultra-high field spinal cord MRI in multiple sclerosis

Reference Title Published in Vendor Patients Sequences Main findings
Ouellette et al. (2020) [55] 7 T imaging reveals a gradient in spinal cord lesion distribution in multiple sclerosis Brain Siemens 20 RRMS, 15 SPMS, 11 healthy-controls FLASH T2*

  • SC lesions involve both WM+GM from the early MS stages and occur mostly independent from brain pathology.

  • SC lesions were localized nearest the subpial surfaces for those with RRMS and the central canal CSF surface in PPMS/SPMS, possibly implying CSF-mediated pathogenic mechanisms in lesion development that may differ between MS subtypes.
Dula et al. (2016) [202] Magnetic resonance imaging of the cervical spinal cord in multiple sclerosis at 7T Multiple Sclerosis J Philips 14 RRMS, 1 PPMS, 13 healthy-controls T1, T2* GRE

  • High-resolution images at 7T exceeded resolutions reported at lower field strengths.

  • GM+WM were sharply demarcated and MS lesions were more readily visualized at 7T compared to clinical acquisitions, with lesions apparent at both fields.

  • WM lesion counts mean 52% increase in detection per patient at 7T vs 3T
Conrad et al. (2018) [203] Multiple sclerosis lesions affect intrinsic functional connectivity of the spinal cord Brain Philips 22 RRMS, 56 healthy-volunteers BOLD fMRI
  • Spinal cord functional networks are generally intact in relapsing-remitting multiple sclerosis but that lesions are associated with focal abnormalities in intrinsic connectivity.
Lefeuvre et al. (2016) [204] MRI of the thoracic spinal cord in multiple sclerosis at 7T ISMRM Proc (abstract) Siemens 1 RRMS, 1 healthy-controls T1-MPRAGE, T2* GRE

  • Anatomical structures, incl. GM+WM, more clearly identified at 7T than at 3T on axial T2*.

  • MS lesion better delineated, its location more clearly identified, on axial T2* 7T compared to 3T.
Witt et al. (2019) [205] 7T MRI Shows Enlarged Anterior Vein in the Spinal Cord of Multiple Sclerosis Patients ISMRM Proc (abstract) Philips 32 RRMS, 19 healthy-controls T2* FFE
  • Significant difference anterior spinal cord vein CSA of spinal cord of MS patients compared to healthy controls.
Massire et al. (2019) [206] High-resolution multiparametric quantitative MRI of the cervical spinal cord at 7T: preliminary results at the early stage of multiple sclerosis ISMRM Proc (abstract) Siemens 7 RRMS, 7 healthy-controls T2* MEDIC, T1 mapping/3D MPRAGE, DTI
  • Great potential for high-res. mp-qMRI for MS investigations at UHF. Especially using T1 mapping, specific anatomical features and microstructural characterization were enhanced (atrophy with enlargement of the median fissure, dilatation of central canal and septum).
Conrad et al. (2015) [207] Measuring Cross Sectional Area of the Spinal Cord at 7T: Validating Fully Automated Segmentation ISMRM Proc (abstract) Philips 20 RRMS, 18 healthy-controls T1, T2* FFE
  • Multi-atlas label fusion to automatically calculate CSA from T2*w images at 7T, achieving a strong relationship with T1w semi-auto estimations
Dula et al. (2016) [186] Chemical exchange saturation transfer of the cervical spinal cord at 7 T NMR Biomed Philips 10 RRMS, 10 healthy-controls CEST, T1, T2* FFE
  • CEST results indicate differences between healthy and pathological spinal cord tissue over much of the spectral range in the corrected z-spectra.

Dula and colleagues [202] were the first to show the potential benefits for the cervical cord in MS patients using a 7T scanner. They showed an improved white matter (WM) and grey matter (GM) SNR and WM:GM CNR for 7T compared to 3T MRI of the cervical cord. They scanned 15 MS patients and 13 healthy volunteers using high resolution axial T1 + T2* weighted and sagittal T2*/spin-density-weighted sequences. Notable in this study was that they also compared field strengths using a scan-time-matched as well as a resolution-matched approach in healthy volunteers. In the MS patients, there was a 52% increase in lesion detection when comparing 3T axial T2 turbo spin-echo (TSE) to 7T T2* acquisitions. However, scan parameters of the 3T T2 TSE against which comparison of lesion sensitivity was performed, were not stated in the article. Also, this does not automatically prove superiority of 7T over 3T MRI for MS, because only a single sequence was compared. A spinal cord imaging study, to be optimal for MS imaging at 1.5T/3T should include a short tau inversion recovery (STIR) or phase-sensitive inversion recovery (PSIR) sequence as is recommended for clinical practice in the MAGNIMS guidelines [208]. Inversion recovery sequences, for now, are still one of the strong points of lower field strengths since they are challenging to optimize for UHF MRI. This shows one of the difficulties in MRI research: the challenge to design a fair comparison between field strengths, since changing field strengths inherently implies changing a variety of other variables (sequence’ scan parameters, gradients, hardware etc.). Additionally, lesion detection in this study, was only compared from C2 to C5 due to the diminishing sensitivity profile of the used spinal cord array of the 7T scanner. Despite these limitations, the authors did show that an excellent resolution and discrimination between the butterfly-shaped GM and surrounding WM can be achieved using 7T MRI of the cervical spine, which can potentially contribute to improving sensitivity for MS lesions.

The superior WM/GM discrimination opened the way for improved segmentation of these compartments. This allows, for example, for better calculation of WM and GM volumes separately [207] and, moreover, detailed determination of the spatial organization of lesions. The study by Ouellette et al. [55] elegantly employed the capabilities of 7T cervical cord MRI in MS patients to investigate lesion distribution of white as well as grey matter lesions. They found that spinal cord lesions have a propensity to manifest nearer the central canal and subpial CSF interfaces, with the gradient shifting from the subpial surface in RRMS to the central cord later in the disease as it progresses to SPMS. They applied an axial and sagittal T2* sequence of the whole cervical spine (C1-C7) at a high in-plane resolution of 0.4 x 0.4 mm and 3 mm slice thickness within an acquisition time of approximately 4 minutes. This demonstrated that with the current state of improvements in coil and sequence design, it is feasible at 7T to achieve whole cervical spine coverage at high resolution within a reasonable scan time. Figure 1 illustrates an example of the potential level of detail and spatial resolution that can be achieved at 7T.

In regard to the thoracic segment of the spine, at the time of this review, there is only one report of a thoracic spine of a MS patient being scanned at 7T. This results from the limited availability and challenges in coil design suitable for thoracic imaging at UHF. This group scanned one RRMS patient and 1 healthy volunteer. Here too, anatomical structures, including the GM and WM, and MS lesions were more clearly identified on axial T2* at 7T compared to 3T [204].

Conrad and colleagues [203] used this to assess functional connectivity at 7T within the spinal cord GM of 22 RRMS patients and 56 healthy controls. The presence of lesions was associated with local alterations in connectivity, which also depended on columnar location. They concluded that in RRMS spinal cord functional networks are generally intact, but that there are local alterations associated with lesions. Additionally, Massire et al. [206] did an exploratory study using diffusion tensor imaging (DTI) and T1 relaxometry to investigate the use of these quantitative measures at 7T in MS patients. Their study showed the potential of using a multiparametric quantitative MRI approach at 7T and that using T1 relaxometry can enhance specific anatomical features and microstructural characterization like atrophy with enlargement of the median fissure, dilatation of central canal and septum. Finally, initial efforts have been made applying chemical exchange saturation transfer (CEST) imaging at 7T to the spinal cord patients in MS [186]. The study concluded that the increased frequency separation and SNR in combination with the prolonged T1 relaxation at 7T provides the necessary improvement to detect subtle changes in MS patients not detected at lower field strengths. CEST can provide an indirect measurement of biochemicals which that are altered in MS.

3.4 Summary and future perspectives

Usage of UHF MRI for the spine, for research is still in an early stage. However, the recent studies discussed in this review, applying spinal imaging at UHF on groups of MS patients are showing that it does have added value for MS research and that, while spinal imaging at 3T did not really improve over 1.5T, these studies already show that 7T MRI can improve spinal MS lesion detection over lower field strengths. An exquisite level of anatomical detail can be achieved due to the higher spatial resolutions and within an acceptable acquisition time. This could, in the future, benefit MS diagnosis as a result of the higher lesion sensitivity. Secondly, UHF spinal cord imaging will likely be able to more accurately show and quantify the different aspects of spinal cord pathology like atrophy, functional connectivity on fMRI, loss of neuroaxonal integrity using diffusion-based techniques (e.g., DTI and NODDI) and nervous system metabolites with MR spectroscopy. Furthermore, with its excellent capabilities for imaging microvasculature and consequent conspicuity for small veins, maybe UHF MRI will make it possible to find in vivo evidence of a perivenous organization of MS lesions in the spinal cord and to identify paramagnetic rims in spinal cord lesions which we are already able to image in chronic MS brain lesions. Most advanced imaging techniques will stay limited to research use. However, since still a lot is unknown about the pathophysiology of spinal pathology in multiple sclerosis, improvements that UHF MRI can bring to the mentioned techniques are important to enriching the toolbox of instruments to further unravel the workings of MS spinal cord pathology. Additionally, as can be seen for UHF brain imaging, is that useful diagnostic or prognostic biomarkers like the central vein sign and paramagnetic rims are drivers of bringing UHF MRI to the clinic as well. If spinal cord UHF MRI continues to prove a great leap in lesion sensitivity and such biomarkers can be identified in the spinal cord as well, this could in the future help identifying MS patients with a poorer prognosis and inform earlier initiation of high-efficacy disease-modifying treatment, making the arrival of UHF spinal cord MRI to the clinic a realistic possibility.

However, to sufficiently exploit the advantages of UHF spinal cord MRI and for it to have a possible clinical future, further research is needed in coping with field inhomogeneities, physiological motion and further developments in coil (e.g., to also achieve thoracic coverage) and sequence design (e.g., STIR and PSIR at UHF) are necessary. Current developments addressing technical challenges are showing promising results. With 7T scanners becoming more accessible and the first scanners being FDA-approved for clinical use, this will further drive progress in UHF research. Additionally, recently, the first steps have been made in moving even beyond 7T for in vivo spinal cord MRI [209] with encouraging results. Finally, besides the needed technical improvements, spinal cord UHF MRI, on top of the expensive scanner (7 to 10 million US dollars [210]), requires specialized coils and hardware (e.g., shim coils) in order to achieve acceptable quality. This is not without financial cost and thus it will still need a considerable amount of time before spinal cord UHF MRI will be optimized enough to achieve a balance between the costs and the benefits of scanning at UHF that is acceptable for a clinical setting. Combined with the fact that sequences and scanners at current clinical strengths are also becoming better and increasingly optimized, this raises the expectation that UHF spinal imaging will not yet benefit MS imaging in daily clinical practice in the coming years and will for the foreseeable future mostly have a role in research.

References

33. Granella F, Tsantes E, Graziuso S, Bazzurri V, Crisi G, Curti E. Spinal cord lesions are frequently asymptomatic in relapsing-remitting multiple sclerosis: A retrospective MRI survey. J Neurol. 2019;266:3031–7. https://doi.org/10.1007/s00415-019-09526-3.
37. Maggi P, Sati P, Nair G, Cortese ICM, Jacobson S, Smith BR, et al. Paramagnetic rim lesions are specific to multiple sclerosis: An international multicenter 3T MRI study. Ann Neurol. 2020;88:1034–42. https://doi.org/10.1002/ana.25877.
50. Bot JC, Barkhof F, Polman CH, Lycklama a Nijeholt GJ, Groot V de, Bergers E, et al. Spinal cord abnormalities in recently diagnosed MS patients: Added value of spinal MRI examination. Neurology. 2004;62:226–33. https://doi.org/10.1212/wnl.62.2.226.
51. Lukas C, Sombekke MH, Bellenberg B, Hahn HK, Popescu V, Bendfeldt K, et al. Relevance of spinal cord abnormalities to clinical disability in multiple sclerosis: MR imaging findings in a large cohort of patients. Radiology. 2013;269:542–52.
55. Ouellette R, Treaba CA, Granberg T, Herranz E, Barletta V, Mehndiratta A, et al. 7 t imaging reveals a gradient in spinal cord lesion distribution in multiple sclerosis. Brain. 2020;143:2973–87. https://doi.org/10.1093/brain/awaa249.
88. Moccia M, Ruggieri S, Ianniello A, Toosy A, Pozzilli C, Ciccarelli O. Advances in spinal cord imaging in multiple sclerosis. Ther Adv Neurol Disord. 2019;12:1756286419840593. https://doi.org/10.1177/1756286419840593.
132. Lukas C, Knol DL, Sombekke MH, Bellenberg B, Hahn HK, Popescu V, et al. Cervical spinal cord volume loss is related to clinical disability progression in multiple sclerosis. J Neurol Neurosurg Psychiatry. 2015;86:410–8. https://doi.org/10.1136/jnnp-2014-308021.
135. Weeda MM, Middelkoop SM, Steenwijk MD, Daams M, Amiri H, Brouwer I, et al. Validation of mean upper cervical cord area (MUCCA) measurement techniques in multiple sclerosis (MS): High reproducibility and robustness to lesions, but large software and scanner effects. Neuroimage Clin. 2019;24:101962. https://doi.org/10.1016/j.nicl.2019.101962.
148. Daams M, Weiler F, Steenwijk MD, Hahn HK, Geurts JJ, Vrenken H, et al. Mean upper cervical cord area (MUCCA) measurement in long-standing multiple sclerosis: Relation to brain findings and clinical disability. Mult Scler. 2014;20:1860–5. https://doi.org/10.1177/1352458514533399.
156. Thompson AJ, Baranzini SE, Geurts J, Hemmer B, Ciccarelli O. Multiple sclerosis. The Lancet. 2018;391:1622–36. https://doi.org/10.1016/s0140-6736(18)30481-1.
169. Kearney H, Miller DH, Ciccarelli O. Spinal cord MRI in multiple sclerosis–diagnostic, prognostic and clinical value. Nat Rev Neurol. 2015;11:327–38. https://doi.org/10.1038/nrneurol.2015.80.
170. Bot JC, Barkhof F, Polman CH, Lycklama a Nijeholt GJ, Groot V de, Bergers E, et al. Spinal cord abnormalities in recently diagnosed MS patients: Added value of spinal MRI examination. Neurology. 2004;62:226–33. https://doi.org/10.1212/wnl.62.2.226.
171. Lukas C, Sombekke MH, Bellenberg B, Hahn HK, Popescu V, Bendfeldt K, et al. Relevance of spinal cord abnormalities to clinical disability in multiple sclerosis: MR imaging findings in a large cohort of patients. Radiology. 2013;269:542–52.
172. Brex PA, O’Riordan JI, Miszkiel KA, Moseley IF, Thompson AJ, Plant GT, et al. Multisequence MRI in clinically isolated syndromes and the early development of MS. Neurology. 1999;53:1184–90. https://doi.org/10.1212/wnl.53.6.1184.
173. Arrambide G, Rovira A, Sastre-Garriga J, Tur C, Castillo J, Rio J, et al. Spinal cord lesions: A modest contributor to diagnosis in clinically isolated syndromes but a relevant prognostic factor. Mult Scler. 2018;24:301–12. https://doi.org/10.1177/1352458517697830.
174. Brownlee WJ. Do spinal cord lesions matter in patients with clinically isolated syndrome and early MS? Mult Scler. 2018;24:430–1. https://doi.org/10.1177/1352458517749048.
175. Ciccarelli O, Cohen JA, Reingold SC, Weinshenker BG, Amato MP, Banwell B, et al. Spinal cord involvement in multiple sclerosis and neuromyelitis optica spectrum disorders. The Lancet Neurology. 2019;18:185–97. https://doi.org/10.1016/s1474-4422(18)30460-5.
176. Zurawski J, Glanz BI, Healy BC, Tauhid S, Khalid F, Chitnis T, et al. The impact of cervical spinal cord atrophy on quality of life in multiple sclerosis. J Neurol Sci. 2019;403:38–43. https://doi.org/10.1016/j.jns.2019.04.023.
177. Vargas MI, Delavelle J, Kohler R, Becker CD, Lovblad K. Brain and spine MRI artifacts at 3Tesla. J Neuroradiol. 2009;36:74–81. https://doi.org/10.1016/j.neurad.2008.08.001.
178. Ladd ME, Bachert P, Meyerspeer M, Moser E, Nagel AM, Norris DG, et al. Pros and cons of ultra-high-field MRI/MRS for human application. Prog Nucl Magn Reson Spectrosc. 2018;109:1–50. https://doi.org/10.1016/j.pnmrs.2018.06.001.
179. Stankiewicz JM, Glanz BI, Healy BC, Arora A, Neema M, Benedict RH, et al. Brain MRI lesion load at 1.5T and 3T versus clinical status in multiple sclerosis. J Neuroimaging. 2011;21:e50–6. https://doi.org/10.1111/j.1552-6569.2009.00449.x.
180. Wattjes MP, Barkhof F. High field MRI in the diagnosis of multiple sclerosis: High field-high yield? Neuroradiology. 2009;51:279–92. https://doi.org/10.1007/s00234-009-0512-0.
181. Stankiewicz JM, Neema M, Alsop DC, Healy BC, Arora A, Buckle GJ, et al. Spinal cord lesions and clinical status in multiple sclerosis: A 1.5 t and 3 t MRI study. J Neurol Sci. 2009;279:99–105. https://doi.org/10.1016/j.jns.2008.11.009.
182. Massire A, Taso M, Besson P, Guye M, Ranjeva JP, Callot V. High-resolution multi-parametric quantitative magnetic resonance imaging of the human cervical spinal cord at 7T. Neuroimage. 2016;143:58–69. https://doi.org/10.1016/j.neuroimage.2016.08.055.
183. Sati P, Oh J, Constable RT, Evangelou N, Guttmann CR, Henry RG, et al. The central vein sign and its clinical evaluation for the diagnosis of multiple sclerosis: A consensus statement from the north american imaging in multiple sclerosis cooperative. Nat Rev Neurol. 2016;12:714–22. https://doi.org/10.1038/nrneurol.2016.166.
184. Absinta M, Sati P, Reich DS. Advanced MRI and staging of multiple sclerosis lesions. Nature Reviews Neurology. 2016;12:358.
185. Ng Kee Kwong KC, Mollison D, Meijboom R, York EN, Kampaite A, Thrippleton MJ, et al. The prevalence of paramagnetic rim lesions in multiple sclerosis: A systematic review and meta-analysis. PLoS One. 2021;16:e0256845. https://doi.org/10.1371/journal.pone.0256845.
186. Dula AN, Pawate S, Dethrage LM, Conrad BN, Dewey BE, Barry RL, et al. Chemical exchange saturation transfer of the cervical spinal cord at 7 t. NMR Biomed. 2016;29:1249–57. https://doi.org/10.1002/nbm.3581.
187. Winkler SA, Schmitt F, Landes H, Bever J de, Wade T, Alejski A, et al. Gradient and shim technologies for ultra high field MRI. Neuroimage. 2018;168:59–70. https://doi.org/10.1016/j.neuroimage.2016.11.033.
189. De Graaf RA, Brown PB, McIntyre S, Rothman DL, Nixon TW. Dynamic shim updating (DSU) for multislice signal acquisition. Magnetic Resonance in Medicine: An Official Journal of the International Society for Magnetic Resonance in Medicine. 2003;49:409–16.
190. Zecca C, Disanto G, Sormani MP, Riccitelli GC, Cianfoni A, Del Grande F, et al. Relevance of asymptomatic spinal MRI lesions in patients with multiple sclerosis. Mult Scler. 2016;22:782–91. https://doi.org/10.1177/1352458515599246.
191. Hamilton J, Franson D, Seiberlich N. Recent advances in parallel imaging for MRI. Prog Nucl Magn Reson Spectrosc. 2017;101:71–95. https://doi.org/10.1016/j.pnmrs.2017.04.002.
192. Barry RL, Vannesjo SJ, By S, Gore JC, Smith SA. Spinal cord MRI at 7T. Neuroimage. 2018;168:437–51. https://doi.org/10.1016/j.neuroimage.2017.07.003.
193. Vannesjo SJ, Clare S, Kasper L, Tracey I, Miller KL. A method for correcting breathing-induced field fluctuations in T2*-weighted spinal cord imaging using a respiratory trace. Magn Reson Med. 2019;81:3745–53. https://doi.org/10.1002/mrm.27664.
194. Heeswijk RB van, Bonanno G, Coppo S, Coristine A, Kober T, Stuber M. Motion compensation strategies in magnetic resonance imaging. Critical Reviews™ in Biomedical Engineering. 2012;40.
195. Sati P. Diagnosis of multiple sclerosis through the lens of ultra-high-field MRI. J Magn Reson. 2018;291:101–9. https://doi.org/10.1016/j.jmr.2018.01.022.
196. Smith SA, Dortch RD, Barry RL, Gore JC. Ultra-high field spinal cord imaging. In: Quantitative MRI of the spinal cord. 2014. p. 106–19. https://doi.org/10.1016/b978-0-12-396973-6.00008-3.
197. Fries P, Morelli JN, Lux F, Tillement O, Schneider G, Buecker A. The issues and tentative solutions for contrast-enhanced magnetic resonance imaging at ultra-high field strength. Wiley Interdiscip Rev Nanomed Nanobiotechnol. 2014;6:559–73. https://doi.org/10.1002/wnan.1291.
198. Hagberg GE, Scheffler K. Effect of r(1) and r(2) relaxivity of gadolinium-based contrast agents on the t(1)-weighted MR signal at increasing magnetic field strengths. Contrast Media Mol Imaging. 2013;8:456–65. https://doi.org/10.1002/cmmi.1565.
199. Harrison DM, Wang KY, Fiol J, Naunton K, Royal 3rd W., Hua J, et al. Leptomeningeal enhancement at 7T in multiple sclerosis: Frequency, morphology, and relationship to cortical volume. J Neuroimaging. 2017;27:461–8. https://doi.org/10.1111/jon.12444.
200. Heilmaier C, Theysohn JM, Maderwald S, Kraff O, Ladd ME, Ladd SC. A large-scale study on subjective perception of discomfort during 7 and 1.5 t MRI examinations. Bioelectromagnetics. 2011;32:610–9. https://doi.org/10.1002/bem.20680.
201. Friebe B, Wollrab A, Thormann M, Fischbach K, Ricke J, Grueschow M, et al. Sensory perceptions of individuals exposed to the static field of a 7T MRI: A controlled blinded study. J Magn Reson Imaging. 2015;41:1675–81. https://doi.org/10.1002/jmri.24748.
202. Dula AN, Pawate S, Dortch RD, Barry RL, George-Durrett KM, Lyttle BD, et al. Magnetic resonance imaging of the cervical spinal cord in multiple sclerosis at 7T. Mult Scler. 2016;22:320–8. https://doi.org/10.1177/1352458515591070.
203. Conrad BN, Barry RL, Rogers BP, Maki S, Mishra A, Thukral S, et al. Multiple sclerosis lesions affect intrinsic functional connectivity of the spinal cord. Brain. 2018;141:1650–64. https://doi.org/10.1093/brain/awy083.
204. Lefeuvre JA, Duan Q, De Zwart J, Gelderen P van, Lehéricy S, Jacobson SG, et al. MRI of the thoracic spinal cord in multiple sclerosis at 7T. In: ACTRIMS forum 2017. ACTRIMS.
205. Witt BR Atlee; Kirkland. 7T MRI shows enlarged anterior vein in the spinal cord of multiple sclerosis patients. ISMRM Proc. 2019.
206. Massire SL Aurélien; Demortiere. High-resolution multiparametric quantitative MRI of the cervical spinal cord at 7T: Preliminary results at the early stage of multiple sclerosis. ISMRM Proc. 2019.
207. Conrad BN, Lyttle BD, Pawate S, Barry RL, Landman BA, Smith SA. Measuring cross sectional area of the spinal cord at 7T: Validating fully automated segmentation. ISMRM Proc. 2015.
208. Rovira A, Wattjes MP, Tintore M, Tur C, Yousry TA, Sormani MP, et al. Evidence-based guidelines: MAGNIMS consensus guidelines on the use of MRI in multiple sclerosis-clinical implementation in the diagnostic process. Nat Rev Neurol. 2015;11:471–82. https://doi.org/10.1038/nrneurol.2015.106.
209. Geldschlager O, Bosch D, Avdievich NI, Henning A. Ultrahigh-resolution quantitative spinal cord MRI at 9.4T. Magn Reson Med. 2021;85:1013–27. https://doi.org/10.1002/mrm.28455.
210. Ineichen BV, Beck ES, Piccirelli M, Reich DS. New prospects for ultra-high-field magnetic resonance imaging in multiple sclerosis. Invest Radiol. 2021. https://doi.org/10.1097/RLI.0000000000000804.