The Clinical Relevance and Scientific Potential of Ultra High-Field-Strength MR Imaging
=======================================================================================

* Stephen Chan

As neuroradiologists, we are fortunate to work with the ever-advancing MR imaging technology. MR imaging has evolved as a robust and highly versatile clinical tool. In addition, MR imaging has caused dramatic changes in the clinical evaluation of a host of neurologic disorders; a well-recognized example is the role of diffusion-weighted imaging in acute stroke. Such advances typically stem from discoveries of phenomena rendered visible by the development of new MR imaging methods.

Unfortunately, not every new scientific observation or technologic development is associated with the impact and significance that diffusion-weighted imaging has had on the study of stroke. As critical readers of the scientific literature, radiologists should judge the technical accuracy and scientific soundness of a new observation or imaging method. This process involves critical thinking (1). As medical practitioners, we have to ascertain potential clinical relevance of new imaging technology. Then, we need to consider which patients would benefit most and what effect, if any, the new information may have on our future clinical practice. Finally, in cases in which a medical procedure or technology is associated with high cost or health risk, more refined judgment is required to integrate clinical benefits and scientific facts as well as to minimize cost and patient risk.

In this issue of the *AJNR*, Christoforidis et al present a means of direct visualization of abnormal microvascularity within glioblastoma multiforme by using 8-T MR imaging. By using a 2D gradient-recalled echo sequence to obtain 2-mm axial sections of the brain, these investigators were able to produce images of a brain tumor with an in-plane pixel size of 222 μ (by using a matrix of 900 × 900 and a 20-cm field of view). The resulting images displayed normal transmedullary veins that are invisible at conventional angiography. They also showed zones of increased microvascularity (corresponding to tumor blush on conventional angiograms) that are invisible at 1.5-T T2-weighted fast spin-echo MR imaging. In an attempt to underscore the significance of this finding, the authors cite animal models of tumor angiogenesis in which “apparent vessel density” on high-spatial-resolution MR images correlates with histopathologically identified density of microscopic blood vessels.

To address clinically driven questions, critical readers should note the following issues. First, the diagnostic significance of microvascularity alone in the histopathologic evaluation of glioblastoma multiforme is somewhat limited; there are four major grading criteria, of which neovascularity is only one. Second, the therapeutic significance of these findings is unclear; the elimination of neovascularity by various treatments does not necessarily make recurrence of glioblastoma multiforme less likely. In addition, the zones of microvascularity seem to consist mainly of small veins; therefore, the relevance to intraarterial chemotherapy is not clear. Third, the scientific significance seems to be related to angiogenesis, about which much has been written in recent years. However, further work needs to be conducted to determine the correlation of abnormal microvascularity with histopathologically established neovascularity, and to measure the success rate of ultra high-field-strength MR imaging in identifying microvascular changes in a series of glioblastoma cases (as opposed to those changes found in association with other brain tumors).

The issue of technical significance takes us from the clinically driven questions to those related to how a new technique should be applied. These technical issues raise some of the most interesting questions. Compared with a 1.5-T MR imaging system, an 8-T MR imaging system boosts the signal-to-noise ratio (S/N) by a factor between 3 and 5. (This factor depends on various technical parameters, including whether the 8-T MR imaging system uses stronger gradients and increased bandwidth, both of which may reduce S/N gains) (2). This technical advantage offers the possibility of performing various MR imaging experiments even beyond the pursuit of maximal spatial resolution. Unfortunately, technical problems can render whole-brain MR imaging much more problematic at 8 T than at 1.5 T. These problems are related to substantial artifacts due to B inhomogeneity, heightened magnetic susceptibility, inhomogeneous radio-frequency field (B1), and radio-frequency eddy currents (2, 3). For the sake of brevity, let us assume that engineering and scientific advances can solve these technical problems in the near future. Then, radiologic scientists will have the luxury of using the 8-T MR imaging system in many ways.

We may optimize MR pulse sequences that are adequate at 1.5 T but are constrained by physical or physiological limits in signal generation. Functional MR imaging at 1.5 T with blood oxygenation level–dependent effect from a specific task activation often results in a mere 1% to 2% signal intensity change. Compare that with a conservative estimate of 3% to 6% signal intensity change at 8 T. Similarly, diffusion and perfusion MR images are typically obtained at low spatial resolution to maximize acquisition speed. For these MR pulse sequences, the much higher field strength offers the opportunity to collect imaging data at even smaller time intervals (ie, greater temporal resolution) or with smaller imaging voxels (ie, greater spatial resolution).

Experiments can be performed at 8 T that would be extremely challenging at 1.5 T. This includes heteronuclear MR imaging, such as with phosphorus MR spectroscopy and sodium MR imaging, both of which are feasible at 1.5 T but generally require a prolonged pulse sequence to ensure satisfactory signal acquisition. Both of these technologies have been explored in animal models but have yet to make a notable impact in the clinical realm. As has been touted elsewhere, phosphorus MR spectroscopy may yield significant information regarding cellular energy metabolism, and sodium MR imaging may provide detail regarding the integrity (or lack thereof) of physiologically excitable cells, such as neurons.

Surface coil technology and ultra high-field-strength MR imaging could be combined to perform MR microscopy to detect otherwise invisible structural abnormalities in cerebral cortex. Although this could be used to visualize even finer microscopic detail within well-defined lesions, such as those in glioblastoma multiforme, another strategy is to study neurologic disorders associated with structural lesions that are difficult to visualize on conventional MR images. One example is the imaging of pediatric epilepsy; the use of optimized, high-spatial-resolution techniques at 1.5 T without surface coils (4) or with surface coils (5) has improved the detection of focal cortical dysplasia, a lesion that often is surgically treatable. It is likely that MR microscopy will further improve our ability to discern the subtle cortical abnormalities associated with focal cortical dysplasia. From a clinical viewpoint, MR microscopy offers greater promise in identifying hard-to-find structural lesions that are treatable than in demonstrating the fine ultrastructural detail of lesions that are not treatable; the latter remains merely academic, unless the discovery leads to a cure.

Another approach is to use the 8-T MR imaging system as a time-saving device that facilitates the construction of an MR imaging database that combines structural and functional neuroimaging data. Scientists could obtain a battery of imaging data from patients by performing a multiple-sequence MR imaging study. This approach is rendered feasible by using the increased S/N to save imaging time during the longer and more complex MR pulse sequences (eg, MR spectroscopy), thereby enabling an increase in the number and types of pulse sequences. For example, a single MR imaging examination could include structural imaging, diffusion and perfusion MR imaging, proton MR spectroscopy, visual functional MR imaging, and heteronuclear MR imaging. This imaging database could be useful in several contexts. One possibility is an in vivo study of neurobiologic changes throughout the human life cycle, coupled with the development of multidimensional, graphic displays of structural and functional imaging data sets. Comparison of new with old images could be facilitated by computer-driven video displays of time series imaging data sets.

Major scientific benefits are possible with use of the ultra high-field-strength MR imaging system. Developing the full potential of an 8-T MR imaging system is likely to require significant technical advances and continual problem solving, similar to what was needed in the early days of the 1.5-T MR imaging systems. In evaluating potential clinical benefits of 8 T, it should be noted that the technology currently available on commercial 1.5-T MR imaging systems is sophisticated and is much further developed than the most advanced 0.35-T MR imaging systems of the 1980s. Also, 3- and 4-T MR imaging systems are under development that will be less severely affected by technical artifacts related to higher field strength. Therefore, the incremental clinical benefits obtained in moving from a field strength of 1.5 T to 8 T are likely to be less dramatic than those achieved 15 to 20 years ago when the giant step from 0.35-T to 1.5-T MR imaging occurred. More research is required to establish the potential clinical benefits of the ultra high-field-strength MR imaging system; this work may be facilitated by the use of various imaging strategies that judiciously apply the system’s technical advantages to the solution of clinical problems that are potentially treatable.

## References

1.  Hillman BJ. **Noninterpretive skills for radiology residents: critical thinking: deciding whether to incorporate the recommendations of radiology publications and presentations into practice.** AJR Am J Roentgenol 2000;174:943–946
    
    [PubMed](http://www.ajnr.org/lookup/external-ref?access_num=10749227&link_type=MED&atom=%2Fajnr%2F23%2F9%2F1441.atom) 

2.  Glover GH. **Hardware for functional MRI.** In: Jezzard P, Mathews PM, Smith SM, eds. Functional MRI: Introduction to Methods. New York: Oxford University Press;2001 :109–122
    
    

3.  Truong TK, Clymer BD, Chakeres DW, Schmalbrock P. **Three-dimensional numerical simulations of susceptibility-induced magnetic field inhomogeneities in the human head at 8 Tesla.** Proc Intl Soc Mag Reson Med 2002;10:2323
    
    

4.  Chan S, Chin SS, Nordli DR, Goodman RR, DeLaPaz RL, Pedley TA. **Prospective magnetic resonance imaging identification of focal cortical dysplasia, including the non-balloon cell subtype.** Ann Neurol 1998;44:749–757
    
    [CrossRef](http://www.ajnr.org/lookup/external-ref?access_num=10.1002/ana.410440508&link_type=DOI) 
    
    [PubMed](http://www.ajnr.org/lookup/external-ref?access_num=9818930&link_type=MED&atom=%2Fajnr%2F23%2F9%2F1441.atom) 
    
    [Web of Science](http://www.ajnr.org/lookup/external-ref?access_num=000076880800007&link_type=ISI) 

5.  Grant PE, Barkovich AJ, Wald LL, Dillon WP, Laxer KD, Vigneron DB. **High-resolution surface-coil MR of cortical lesions in medically refractory epilepsy: a prospective study.** AJNR Am J Neuroradiol 1997;18:291–301
    
    [Abstract](http://www.ajnr.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiYWpuciI7czo1OiJyZXNpZCI7czo4OiIxOC8yLzI5MSI7czo0OiJhdG9tIjtzOjIwOiIvYWpuci8yMy85LzE0NDEuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

*   Copyright © American Society of Neuroradiology